Mutagenesis Techniques - Science

Sponsored by 

This booklet is brought to you by the AAAS/Science Business Office

Over 20 Years of Innovation 

As a global leader in the development of innovative technologies, 

our products simplify, accelerate, and improve drug discovery and 

drug development research. Since 1984, our reagents, kits, software, 

and instrumentation systems have been used in pharmaceutical and 

biotechnology laboratories worldwide to determine the molecular 

mechanisms of health and disease as well as to search for new drug 

therapies and diagnostic tests. 

To order a copy of our current product catalog, visit 

www.stratagene.com/catalog. 

AMPLIFICATION 

CELL BIOLOGY 

CLONING 

MICROARRAYS 

NUCLEIC ACID ANALYSIS 

PROTEIN FUNCTION & ANALYSIS 

QUANTITATIVE PCR 

SOFTWARE SOLUTIONS

Contents 

2 Introductions: 

Bringing About Change 

Sean Sanders 

The Power of Mutagenesis 

Patricia Sardina 

4 Bypassing a Kinase Activity with an ATP-Competitive Drug 

Feroz R. Papa, Chao Zhang, Kevan Shokat, Peter Walter 

Science 28 November 2003 302: 1533-1537 

12 Human Catechol-O-Methyltransferase Haplotypes Modulate 

Protein Expression by Altering mRNA Secondary Structure 

A. G. Nackley, S.A. Shabalina, I. E. Tchivileva, K. Satterfield, 

O. Korchynskyi, S. S. Makarov, W. Maixner, L. Diatchenko 

Science 22 December 2006 314: 1930-1933 

17 An mRNA Surveillance Mechanism That Eliminates Transcripts 

Lacking Termination Codons 

Pamela A. Frischmeyer, Ambro van Hoof, Kathryn O’Donnell, 

Anthony L. Guerrerio, Roy Parker, Harry C. Dietz 

Science 22 March 2002 295: 2258–2261 

24 Direct Demonstration of an Adaptive Constraint 

Stephen P. Miller, Mark Lunzer, Antony M. Dean 

Science 20 October 2006 314: 458–461 

30 Technical Notes: Large Insertions: Two Simple Steps Using 

Quikchange® II Site-Directed Mutagenesis Kits 

About the Cover: 

γ -Exonuclease binds to an end of double-stranded 

DNA and degrades one of the strands in a 

highly processive manner. The structure of the 

exonuclease consists of a toroidal trimer (~94 

angstroms in diameter) and is presumed to enclose 

its substrate in the manner illustrated. [Image from 

the cover of Science 277 (19 September 1997)] 

Design and Layout: Amy Hardcastle 

Copy Editor: Robert Buck 

© 2007 by The American Association for the Advancement of Science. 

All rights reserved. 

1

Introductions 

Bringing About Change 

This new method of mutagenesis has considerable potential in genetic studies.… [S]pecific 

mutation within protein structural genes will allow the production of proteins with specific amino 

acid changes. This will allow precise studies of structure-function relationships, for example 

the role of specific amino acids in enzyme active sites, and the more convenient production of 

proteins whose action is species specific, for example the conversion of a cloned gene coding for 

a lower vertebrate hormone to that for a human. 

— closing paragraph of the seminal paper from Michael Smith’s lab first describing 

site-directed mutagenesis 1 

The art of life lies in a constant readjustment to our surroundings. 

— Okakura Kakuzo, Japanese scholar 

It is a truism that, to survive, we must be able to change. This, too, is the essence of 

evolution on the most fundamental biological level: without the innate ability for 

DNA to undergo mutation, it would not be possible for organisms to adapt and survive 

changes in their environment. Since humankind achieved an understanding of this mutagenesis 

process and its power, we have been trying to control it and bend it to our will. 

First described in 1978 by Michael Smith at the University of British Columbia in 

Vancouver, 1 site-directed mutagenesis (SDM) has become an invaluable tool for molecular 

biologists. Originally named oligonucleotide-directed mutagenesis, the first experiments 

made use of short, 12-nucleotide strands of synthetic DNA containing a mismatch 

to an intrinsic reporter gene of the bacteriophage ΦX174. Using these oligonucleotides, 

together with DNA polymerase I from E. coli and T4 DNA ligase, Smith and his team 

demonstrated that it was possible to introduce a permanent mutation in the circular DNA 

of the phage, resulting in a phenotypic change. The prescient quote above shows the authors’ 

understanding of the manifold applications for the new methodology. 

Since its development, a number of modifications and improvements have been made 

to the SDM technique. The greatest advance came with the invention of the polymerase 

chain reaction (PCR) in 1983. Since then, SDM and PCR have been inextricably linked, 

a circumstance reflected in the 1993 Nobel in Chemistry being shared by Kary Mullis, 

the pioneer of PCR, and Michael Smith. This advance made SDM both quicker and more 

flexible and has enabled it to become a standard and necessary technique employed in 

labs worldwide. 

In this booklet, we bring together a collection of influential papers that all depended 

on SDM for their success. As one of the most used techniques in molecular biology, it 

plays an often unacknowledged–but essential–role in many research projects. There is little 

doubt that future refinements and improvements will be made to the technique, allowing 

for an even broader range of applications. As these modifications are commercialized 

and made available to all scientists, new and innovative uses for them will be discovered, 

facilitating enrichment of our knowledge in many areas of research, from basic cellular 

processes to complex diseases such as neurodegenerative disorders and cancer. 

Sean Sanders, Ph.D. 

Commercial Editor, Science 

1 C. A. Hutchison III, et al. Mutagenesis at a specific position in a DNA sequence. J. Biol. Chem. 253:6551-6560 (1978). 

2

The Power of Mutagenesis 

In vitro mutagenesis is a very powerful tool for studying protein structure-function 

relationships, altering protein activity, and for modifying vector sequences to incorporate 

affinity tags and correct frame shift errors. A variety of mutagenesis techniques 

have been developed over the past decade that allow researchers to generate point mutations, 

modify (insert, delete or replace) one or more codons, swap domains between 

related gene sequences, and create diverse collections of mutant clones. 

Mutagenesis strategies can be divided into two main types, random or site-directed. 

With random mutagenesis, point mutations are introduced at random positions in a geneof-interest, 

typically through PCR employing an error-prone DNA polymerase (errorprone 

PCR). Randomized sequences are then cloned into a suitable expression vector, 

and the resulting mutant libraries can be screened to identify mutants with altered or improved 

properties. By correlating changes in function to alterations in protein sequence, 

researchers can create a preliminary map of protein functional domains. 

Site-directed mutagenesis is the method of choice for altering a gene or vector sequence 

at a selected location. Point mutations, insertions, or deletions are introduced by 

incorporating primers containing the desired modification(s) with a DNA polymerase in 

an amplification reaction. In more complex protein engineering experiments, researchers 

can design mutagenic primers to incorporate degenerate codons (site-saturation mutagenesis). 

Stratagene, now an Agilent Technologies company, is a worldwide leader in developing 

innovative products and technologies for life science research. Our expertise in enzyme 

engineering has translated into the most efficient and easy-to-use mutagenesis kits 

available today. Our QuikChange ® Site-Directed Mutagenesis products provide the most 

reliable and accurate method for performing targeted mutagenesis, saving time and valuable 

resources. These products have been cited in thousands of publications, including 

the papers featured in this booklet. In addition, our GeneMorph ® Random Mutagenesis 

products offer an easy, robust method for creating diverse mutant collections. 

Our mutagenesis product offering has been substantially enhanced by our newest 

product introduction, the QuikChange ® Lightning Site-Directed Mutagenesis Kit. This 

kit offers significant time-savings by incorporating new, exclusive fast enzymes which 

allow researchers to shorten the duration of the thermal cycling and digestion steps while 

maintaining ultra-high fidelity and mutation efficiency. 

The efficacy of mutagenesis as a tool for studying protein function and structure may 

lead to new experimental therapeutic approaches to diseases such as cancer. Precise mapping 

of protein-protein interactions may ultimately lead to a greater understanding of 

disease mechanics. This Science Mutagenesis Techniques Collection booklet illustrates 

the utility of mutagenesis in a wide variety of research studies. We are pleased to have the 

opportunity to sponsor this publication. 

Patricia Sardina 

Product Manager 

Stratagene, An Agilent Technologies Company 

3

Secretory and transmembrane pro- functional domains (4). The most N-terteins 

traversing the endoplasmic minal domain, which resides in the ER 

reticulum (ER) during their biogen- lumen, senses elevated levels of unfolded 

esis fold to their native states in this com- ER proteins. Dissociation of ER chaperpartment 

(1). This process is facilitated by ones from Ire1 as they become engaged 

a plethora of ER-resident activities (the with unfolded proteins is thought to trig- 

protein folding machinery) (2). Insuffiger Ire1 activation (3). 

cient protein folding capacity causes an Ire1’s most C-terminal domain is a reg- 

accumulation of unfolded proteins in the ulated endoribonuclease (RNase), which 

ER lumen (a condition referred to as ER has a single known substrate in yeast: the 

stress) and triggers a transcriptional pro- HAC1 

gram called the UPR. In yeast, UPR targets 

include genes encoding chaperones, 

oxido-reductases, phospholipid biosynthetic 

enzymes, ER-associated degradation 

components, and proteins functioning 

downstream in the secretory pathway (3). 

Together, UPR target activities afford proteins 

passing through the ER an extended 

opportunity to fold and assemble properly, 

dispose of unsalvageable unfolded polypeptides, 

and increase the capacity for ER 

export. 

The yeast UPR is signaled through the 

ER stress sensor Ire1, a single-spanning 

ER transmembrane protein with three 

u mRNA (u stands for uninduced), 

which encodes the Hac1 transcriptional 

activator necessary for activation of UPR 

targets (5). HAC1u mRNA is constitutively 

transcribed, but not translated, because it 

contains a nonconventional translation-inhibitory 

intron (6). Upon activation, Ire1’s 

RNase cleaves HAC1u mRNA at two specific 

sites, excising the intron (7). The 5′ 

and 3′ exons are rejoined by tRNA ligase 

(8), resulting in spliced HAC1i Unfolded proteins in the endoplasmic reticulum cause trans-autophosphorylation 

of the bifunctional transmembrane kinase Ire1, which induces its endoribonuclease 

activity. The endoribonuclease initiates nonconventional splicing 

of HAC1 messenger RNA to trigger the unfolded-protein response (UPR). We 

explored the role of Ire1’s kinase domain by sensitizing it through site-directed 

mutagenesis to the ATP-competitive inhibitor 1NM-PP1. Paradoxically, rather 

than being inhibited by 1NM-PP1, drug-sensitized Ire1 mutants required 1NM- 

PP1 as a cofactor for activation. In the presence of 1NM-PP1, drug-sensitized 

Ire1 bypassed mutations that inactivate its kinase activity and induced a full 

UPR. Thus, rather than through phosphorylation per se, a conformational 

change in the kinase domain triggered by occupancy of the active site with a 

ligand leads to activation of all known downstream functions. 

Secretory and transmembrane proteins traversing 

the endoplasmic reticulum (ER) during 

their biogenesis fold to their native states 

in this compartment (1). This process is facilitated 

by a plethora of ER-resident activities 

(the protein folding machinery) (2). Insufficient 

protein folding capacity causes an 

accumulation of unfolded proteins in the ER 

lumen (a condition referred to as ER stress) 

and triggers a transcriptional program called 

the UPR. In yeast, UPR targets include genes 

encoding chaperones, oxido-reductases, 

phospholipid biosynthetic enzymes, ERassociated 

degradation components, and proteins 

functioning downstream in the secretory 

pathway (3). Together, UPR target activities 

afford proteins passing through the ER an 

extended opportunity to fold and assemble 

properly, dispose of unsalvageable unfolded 

polypeptides, and increase the capacity for 

ER export. 

The yeast UPR is signaled through the ER 

stress sensor Ire1, a single-spanning ER 

transmembrane protein with three functional 

domains (4). The most N-terminal domain, 

which resides in the ER lumen, senses elevated 

levels of unfolded ER proteins. Dissociation 

of ER chaperones from Ire1 as they 

mRNA 

become engaged with unfolded proteins is (i stands for induced), which is actively 

thought to trigger Ire1 activation (3). translated to produce the Hac1 transcrip- 

Ire1’s most C-terminal domain is a regutional 

activator, in turn upregulating UPR 

lated endoribonuclease (RNase), which has a 

single known substrate in yeast: the HAC1 target genes (5). 

A functional kinase domain precedes 

the RNase domain on the cytosolic side 

of the ER membrane (9). Activation of 

Ire1 leads to its oligomerization in the 

ER membrane, followed by trans-autophosphorylation 

(9, 10). Mutations of 

catalytically essential kinase active-site 

u 

mRNA (u stands for uninduced), which encodes 

the Hac1 transcriptional activator necessary 

for activation of UPR targets (5). 

HAC1u mRNA is constitutively transcribed, 

but not translated, because it contains a nonconventional 

translation-inhibitory intron 

(6). Upon activation, Ire1’s RNase cleaves 

HAC1u mRNA at two specific sites, excising 

the intron (7). The 5� and 3� exons are rejoined 

by tRNA ligase (8), resulting in 

spliced HAC1i mRNA (i stands for induced), 

which is actively translated to produce the 

Hac1 transcriptional activator, in turn upregulating 

UPR target genes (5). 

A functional kinase domain precedes the 

RNase domain on the cytosolic side of the ER 

membrane (9). Activation of Ire1 leads to its 

oligomerization in the ER membrane, followed 

by trans-autophosphorylation (9, 10). 

Mutations of catalytically essential kinase 

active-site residues or mutations of residues 

known to become phosphorylated prevent 

HAC1u mRNA splicing and abrogate UPR 

signaling, demonstrating that Ire1’s kinase 

phosphotransfer function is essential for 

RNase activation (4, 9, 11). Thus, Ire1 communicates 

an unfolded-protein signal from 

the ER to the cytosol using the lumenal domain 

as the sensor and the RNase domain as 

the effector. Although clearly important for 

the circuitry of this machine, it is unknown 

why and how Ire1’s kinase is required for 

activation of its RNase. To dissect the role of 

Ire1’s kinase function, we used a recently 

developed strategy that allows us to sensitize 

Ire1 to specific kinase inhibitors (12). 

Unexpected behavior of Ire1 mutants 

sensitized to an ATP-competitive drug. 

The mutation of Leu745 1 2 Department of Medicine, Department of Cellular 

and Molecular Pharmacology, 

—situated at a conserved 

position in the adenosine 5�-triphosphate (ATP)– 

binding site—to Ala or Gly is predicted to sen- 

3Department of Biochemistry 

and Biophysics, 4 Feroz R. Papa, 

Howard Hughes Medical 

Institute, University of California, San Francisco, CA 

94143–2200, USA. 

*To whom correspondence should be addressed. Email: 

frpapa@medicine.ucsf.edu 

1,3* Chao Zhang, 2 Kevan Shokat, 2 Peter Walter3,4 Unfolded proteins in the endoplasmic reticulum cause trans-autophosphorylation of the 

bifunctional transmembrane kinase Ire1, which induces its endoribonuclease activity. The 

endoribonuclease initiates nonconventional splicing of HAC1 messenger RNA to trigger 

the unfolded protein response (UPR). We explored the role of Ire1’s kinase domain by 

sensitizing it through site-directed mutagenesis to the ATP-competitive inhibitor 1NM- 

PP1. Paradoxically, rather than being inhibited by 1NM-PP1, drug-sensitized Ire1 mutants 

required 1NM-PP1 as a cofactor for activation. In the presence of 1NM-PP1, drugsensitized 

Ire1 bypassed mutations that inactivate its kinase activity and induced a full 

UPR. Thus, rather than through phosphorylation per se, a conformational change in the 

kinase domain triggered by occupancy of the active site with a ligand leads to activation 

of all known downstream functions. 

4 

Bypassing a Kinase Activity with an 

ATP-Competitive Drug 

Feroz R. Papa, 1,3 * Chao Zhang, 2 Kevan Shokat, 2 Peter Walter3,4 Bypassing a Kinase Activity with 

an ATP-Competitive Drug 

sitize Ire1 

butyl-3-nap 

d]pyrimidin 

an enlarged 

wild-type k 

fected by 

partially o 

Ire1, these 

UPR sign 

vivo repo 

creased k 

[Ire1(L745 

relative to 

Ire1(L745 

approachin 

(Fig. 1A). 

Unexpe 

to cells 

Ire1(L745A 

concentrat 

1NM-PP1– 

stead, 1NM 

slightly (F 

prise, 1NM 

stored sig 

Ire1(L745G 

presents a 

compound 

an inhibito 

tant enzym 

UPR ac 

by dithioth 

cin (18); 1 

indicating 

directly af 

over, the e 

other stru 

naphthylm 

and 1-naph 

ene group 

the naphth 

hibited wil 

Activat 

ATP-comp 

tion of wh 

sary in the 

this end, w 

mutation 

D828A, w 

due needed 

as the Ire 

was detect 

wild-type c 

detectable, 

and could 

Ire1(L745A 

the UPR r 

addition o 

levels appr 

tivation lev 

binding of 

Ire1 rende 

www.sciencemag.org SCIENCE VOL 302 28 NOVEMBER 2003

esidues or mutations of residues known 

to become phosphorylated prevent HAC1 u 

mRNA splicing and abrogate UPR signaling, 

demonstrating that Ire1’s kinase 

phosphotransfer function is essential for 

RNase activation (4, 9, 11). Thus, Ire1 

communicates an unfolded-protein signal 

from the ER to the cytosol using the lumenal 

domain as the sensor and the RNase 

domain as the effector. Although clearly 

important for the circuitry of this machine, 

it is unknown why and how Ire1’s kinase 

is required for activation of its RNase. To 

dissect the role of Ire1’s kinase function, 

we used a recently developed strategy that 

allows us to sensitize Ire1 to specific kinase 

inhibitors (12). 

Unexpected behavior of Ire1 mutants 

sensitized to an ATP-competitive 

drug. The mutation of Leu 745 —situated 

at a conserved position in the adenosine 

5′-triphosphate (ATP)–binding site—to 

Ala or Gly is predicted to sensitize Ire1 to 

the ATP-competitive drug 1-tert-butyl-3naphthalen-1-ylmethyl-1H-pyrazolo[3,4d]pyrimidin-4-ylemine 

(1NM-PP1) by 

creating an enlarged active-site pocket not 

found in any wild-type kinase (13). Most 

kinases are not affected by these mutations, 

but others are partially or severely 

impaired (14, 15). For Ire1, these substitutions 

markedly decreased UPR signaling 

(assayed through an in vivo reporter), 

which is indicative of decreased kinase activity. 

Ire1(L 745 →A 745 ) [Ire1(L745A) (16)] 

showed a 40% decrease relative to the 

wild-type kinase, whereas Ire1(L745G) 

showed a >90% decrease, approaching 

that of the Δire1 control (Fig. 1A) 

Unexpectedly, the addition of 1NM- 

PP1 to cells expressing partially active 

Ire1(L745A) caused no inhibition, even 

at concentrations that completely inhibit 

other 1NM-PP1–sensitized kinases (15, 

17). Instead, 1NM-PP1 increased reporter 

activity slightly (Fig. 1A, bar 4). To our 

further surprise, 1NM-PP1 provided significantly 

restored signaling to the severely 

crippled Ire1(L745G) (Fig. 1A, bar 6). 

This result presents a paradox: An ATPcompetitive 

compound that was expected 

to function as an inhibitor of the rationally 

engineered mutant enzyme instead permits 

activation. 

UPR activation strictly required induction 

by dithiothreitol (DTT) (Fig. 1) or 

tunicamycin (18); 1NM-PP1 by itself had 

no effect, indicating that it does not induce 

the UPR by directly affecting ER protein 

folding. Moreover, the effects of 1NM- 

PP1 were specific: other structurally similar 

compounds, 2-naphthylmethyl PP1, 

which is a regio-isomer, and 1-naphthyl 

PP1, which lacks the methylene group between 

the heterocyclic ring and the naphthyl 

group, neither activated nor inhibited 

wild-type or mutant Ire1 (18). 

Activation rather than inhibition by 

an ATP-competitive inhibitor raised the 

question of whether the kinase activity 

is necessary in the 1NM-PP1–sensitized 

mutants. To this end, we combined 

the L745A or L745G mutation with a 

“kinase-dead” variant D828A, which 

lacks an active-site Asp residue needed 

to chelate Mg 2+ (11, 19). Whereas the 

Ire1(L745A,D828A) double mutant was 

detectable in cells at levels near those of 

wild-type cells, Ire1(L745G,D828A) was 

undetectable, probably because of instability, 

and could not be assayed. As expected, 

Ire1(L745A,D828A) was unable to induce 

the UPR reporter with DTT alone, but 

the addition of 1NM-PP1 allowed activation 

to levels approaching 80% of the 

wild-type activation level (Fig. 1B). This 

suggests that the binding of 1NM-PP1 to 

1NM-PP1–sensitized Ire1 renders the kinase 

activity dispensable. 

Trans-autophosphorylation by Ire1 of 

two activation-segment Ser residues (S 840 

and S 841 ) is necessary for signaling (9), 

prompting the question of whether this 

requirement is also bypassed in 1NM- 

PP1–sensitized mutants. As expected, the 

unphosphorylatable Ire1 (S840A,S841A) 

mutant was inactive in the absence and 

presence of 1NM-PP1 (Fig. 1C). In 

5

of two 

nd S 841 ) 

ting the 

is also 

mutants. 

le Ire1 

the abig. 

1C). 

d Ire1 

(L745G, 

tly actisphoryld. 

Two 

lative to 

ignaling 

745 was 

to Gly 

ent with 

creasing 

n reducresence 

nsitized 

in the 

perhaps 

ding by 

ain was 

equires 

mRNA 

ation by 

wed the 

mRNA 

expectith 

DTT 

ealed by 

HAC1 i 

M-PP1 

Fig. 2A, 

h HAC1 

ted sig- 

P1–senmRNA 

a signifprovidlone 

did 

xpected, 

ng (e.g., 

M-PP1 

pressing 

t, which 

roduced 

xhibited 

duction 

(Fig. 2, 

re1 pronts,exbserved 

nsitized 

Fig. 1. (A to D) In vivo UPR assays using a UPR element–driven lacZ reporter. The relevant 

genotypes and the presence or absence of 1NM-PP1 and the UPR inducer DTT are indicated. For 

(A) to (C), DTT was used in all assays. wt, wild type; a.u., arbitrary units. 

contrast, the 1NM-PP1–sensitized Ire1 

(L745A,S840A,S841A) and Ire1(L745G, 

S840A,S841A) mutants were significantly 

activated by 1NM-PP1, indicating that 

phosphorylation of S 

6 

840 and S841 stress. Thus, in the genetic background of 

1NM-PP1–sensitized IRE1 alleles, 1NM-PP1 

is, in effect, a cofactor for signaling in the UPR. 

1NM-PP1 uncouples the kinase and 

RNase activities of 1NM-PP1–sensitized 

can be 

Ire1. To rule out indirect effects, we assayed 

bypassed. Two trends are apparent: With 

the activities of the Ire1 mutants described 

above DTT inalone, vitro. relative The cytosolic to the portion parent mutant of Ire1 

consisting Ire1(S840A,S841A), of the kinasesignaling domain and decreased RNase 

domain as the side (Ire1*) chain wasof expressed residue 745 andwas purified pro- 

from gressively Escherichia reduced coli from (7). Leu Using to Ala [�- to Gly 

(Fig. 1C, bars 3, 5, and 7). This is consistent 

with the results of Fig. 1A, which 

32P]-la beled ATP, we assayed wild-type Ire1* and 

Ire1*(L745G) for autophosphorylation in the 

presence or absence of 1NM-PP1. Wild-type 

Ire1* exhibited strong autophosphorylation, 

which was not inhibited by 1NM-PP1 (Fig. 3A, 

lanes 1 and 2). In contrast, Ire1*(L745G) ex- 

Fig. 2. In vivo HAC1 

mRNA splicing and 

Hac1 protein production. 

Northern blots 

showing (A) in vivo 

HAC1 mRNA splicing 

and (B) SCR1 ribosomal 

RNA (rRNA) as 

loading control. Western 

blots showing (C) 

Hac1 and (D) Ire1 protein 

levels. The arrow in 

(C), lane 6, calls attention 

to the small 

amount of Hac1 protein 

present in these cells. 

Ire1*(L745G) show an increasing is diminished sensitivity in its ability of the toacuse ATP tive-site as to substrate side-chain and reduction that 1NM-PP1 at residue is, as 

predicted, 745. Reciprocally, an ATP competitor. in the presence of 1NM- 

PP1, In contrast, activation when of provided 1NM-PP1–sensitized 

with [�- 

Ire1(S840A,S841A) mutants increased in 

the same order (Fig. 1C, bars 4, 6, and 8), 

perhaps because of progressively enhanced 

binding by 1NM-PP1 as the residue 745 

side chain was trimmed back 

1NM-PP1–sensitized Ire1 requires 

1NM-PP1 as a cofactor for HAC1 mRNA 

splicing. We next confirmed that activa- 

32P]-la beled N-6 benzyl ATP, an ATP analog with a 

bulky substituent (Fig. 3E), Ire1*(L745G) displayed 

enhanced autophosphorylation activity 

relative to wild-type Ire1* (Fig. 3B, lane 1 and 

2). Therefore, the L745G mutation does not destroy 

Ire1’s catalytic function, but rather alters its 

substrate specificity from ATP to ATP analogs 

having complementary features that permit binding 

in the expanded active site. 

To ask how the RNase activity of Ire1 is 

affected by manipulation of its kinase domain, 

we incubated wild-type Ire1* and Ire1*(L745G) 

with a truncated in vitro transcribed HAC1 RNA

tion by 1NM-PP1–sensitized Ire1 mutants 

followed the expected pathway of activating 

HAC1 mRNA splicing and Hac1 protein 

production. As expected, UPR induction 

of wild-type cells with DTT caused 

splicing of HAC1 mRNA, as revealed by 

near-complete conversion of HAC1 u to 

HAC1 i mRNA (Fig. 2A, lanes 1 and 2) (5); 

1NM-PP1 had no effect by itself or with 

DTT (Fig. 2A, lanes 3 and 4). In strict correlation 

with HAC1 mRNA splicing, Hac1 

protein accumulated significantly (Fig. 

2C, lanes 2 and 4) 

In contrast, cells expressing 1NM- 

PP1–sensitized Ire1(L745G) spliced HAC1 

mRNA weakly with DTT alone, but displayed 

a significant increase when 1NM- 

PP1 was also provided (Fig. 2A, lanes 5 to 

8). 1NM-PP1 alone did not activate HAC1 

mRNA splicing. As expected, Hac1 protein 

levels correlated with splicing (e.g., Fig. 

2C, lanes 6 and 8). Activation by 1NM- 

PP1 was even more pronounced in cells 

expressing the Ire1(L745A,D828A) double 

mutant, which neither spliced HAC1 

mRNA nor produced Hac1 protein with 

DTT alone, but exhibited robust splicing 

and Hac1 protein production when provided 

both 1NM-PP1and DTT (Fig. 2, A 

and C, compare lanes 10 and 12). 

Importantly, steady-state levels of Ire1 

proteins were comparable in all experiments, 

excluding as a trivial explanation 

for the observed effects the possibility 

that 1NM-PP1–sensitized Ire1 mutants 

are only stably expressed in the presence 

of 1NM-PP1 (Fig. 2D). Taken together, 

our data show that 1NM-PP1 permits but 

does not instruct 1NM-PP1–sensitized 

Ire1 mutants to splice HAC1 mRNA in response 

to ER stress. Thus, in the genetic 

background of 1NM-PP1–sensitized IRE1 

alleles, 1NM-PP1 is, in effect, a cofactor 

for signaling in the UPR 

1NM-PP1 uncouples the kinase and 

RNase activities of 1NM-PP1–sensitized 

Ire1. To rule out indirect effects, we 

assayed the activities of the Ire1 mutants 

described above in vitro. The cytosolic 

portion of Ire1 consisting of the kinase 

domain and RNase domain (Ire1*) was 

expressed and purified from Escherichia 

coli (7). Using γ- 32 P]-labeled ATP, we assayed 

wild-type Ire1* and Ire1*(L745G) 

for autophosphorylation in the presence or 

absence of 1NM-PP1. Wild-type Ire1* exhibited 

strong autophosphorylation, which 

was not inhibited by 1NM-PP1 (Fig. 3A, 

lanes 1 and 2). In contrast, Ire1*(L745G) 

exhibited markedly diminished autophosphorylation, 

consistent with poor signaling 

by Ire1(L745G) in vivo (Figs. 1 and 

2), which was completely extinguished by 

1NM-PP1 (Fig. 3A, lane 3 and 4). This 

suggested that Ire1*(L745G) is diminished 

in its ability to use ATP as substrate 

and that 1NM-PP1 is, as predicted, an ATP 

competitor. 

In contrast, when provided with [γ- 

32 P]-labeled N-6 benzyl ATP, an ATP 

analog with a bulky substituent (Fig. 3E), 

Ire1*(L745G) displayed enhanced autophosphorylation 

activity relative to wildtype 

Ire1* (Fig. 3B, lane 1 and 2). Therefore, 

the L745G mutation does not destroy 

Ire1’s catalytic function, but rather alters its 

substrate specificity from ATP to ATP analogs 

having complementary features that 

permit binding in the expanded active site. 

To ask how the RNase activity of Ire1 

is affected by manipulation of its kinase 

domain, we incubated wild-type Ire1* 

and Ire1*(L745G) with a truncated in 

vitro transcribed HAC1 RNA substrate 

(HAC1 600 ) and monitored the production 

of cleavage products (7). Wild-type Ire1* 

cleaved HAC1 600 RNA at both splice junctions, 

producing the expected fragments 

corresponding to the singly and doubly cut 

species (Fig. 3C, lane 2). The cleavage reaction 

was not affected by 1NM-PP1 (Fig. 

3C, lane 3). In contrast, Ire1*(L745G) was 

completely inactive, whereas the addition 

of 1NM-PP1, at the same concentration 

that was shown in Fig. 3A (lane 4) to completely 

inhibit the kinase activity, restored 

RNase activity significantly (Fig. 3C, 

lanes 4 and 5), strongly supporting the no- 

7

expressions for every mRNA betwe 

ified genotypes. 

The comparison of wild-ty 

expressing cells with �ire1 cells 

distinct set of up-regulated UPR tar 

wild-type IRE1-expressing cells 

genes that were up-regulated more t 

are shown in pink). These IRE1 

genes include previously described 

scriptional targets (21, 22). Notabl 

set of (pink-colored) UPR target 

conformational change up-regulated occurs to a similar ma 

IRE1(L745A,D828A)-expressing c 

in response to cofactor bind- 

to �ire1 cells (Fig. 5B), suggesting 

lane 2). The cleavage reaction was not affected ent velocity sedimentation. ing Ire1* by sedimented analyzing as nonical Ire1* UPRand was induced in the 

by 1NM-PP1 (Fig. 3C, lane 3). In contrast, a broad peak in the gradient Ire1*(L745G) (Fig. 4). Upon the through sensitized, glycerol- kinase-dead mutant. This 

Ire1*(L745G) was completely inactive, whereas addition of ADP (Fig. 4B), but not 1NM-PP1 was confirmed by the tight superim 

the addition of 1NM-PP1, at the same concen- (Fig. 4C), the peak of Ire1* gradient shifted as avelocity result of sedimentation. 

the scatter diagonals of both UPR 

tration that was shown in Fig. 3A (lane 4) to faster sedimentation. Conversely, Ire1* sedimented the peak of as the a broad rest ofpeak the transcriptome evid 

completely inhibit the kinase activity, restored Ire1*(L745G) shifted by a corresponding expression profiles of wild-type I 

tion that 1NM-PP1 uncouples the kinase in the gradient (Fig. 4). Upon the addition 

RNase activity significantly (Fig. 3C, lanes 4 and amount upon the addition of 1NM-PP1 (Fig. 4F) pared with IRE1(L745A,D828A) 

5), and strongly RNase supporting activities the notionof that1NM-PP1–sensi 1NM-PP1 but not upon the of addition ADP (Fig. of ADP 4B), (Fig. but 4E). This not 1NM-PP1 cells (Fig. 5C). (Fig. 

uncouples tized Ire1. the kinase and RNase activities of shift may be4C), indicative the peak of aof conformational Ire1* shifted as Building a result anof instructive UPR 

1NM-PP1–sensitized Ire1. 

change and/or change in the oligomeric state of 1NM-PP1–sensitized IRE1 mutants d 

We We have have previously previously shown thatshown cleavage that of the cleav- enzymes faster in response sedimentation. to the binding ofConversely, their far act the permissively, peak because activat 

HAC1 age mRNA of HAC1 by wild-type mRNA Ire1* by is stimulated wild-type respective Ire1* cofactors. of Ire1*(L745G) shifted by a both correspond- 1NM-PP1 and protein-misfold 

by adenosine 5�-diphosphate (ADP) (7). This is 1NM-PP1–sensitized, kinase-dead Ire1 This indicates that an unfolded-pr 

is stimulated by adenosine 5′-diphosphate ing amount upon the addition of 1NM-PP1 

consistent with the notion derived from in vivo signals a canonical UPR. To ask whether needs to be received by the ER lume 

studies (ADP) that Ire1 (7). requires This is anconsistent active kinasewith do- the bypassing no- the(Fig. Ire1 4F) kinasebut activity not upon with 1NM- the addition for activation. of ADP We asked if we could 

main,tion because derived we know from that in recombinant vivo studies wild- that PP1 Ire1 allows the (Fig. induction 4E). This of a full shift UPR, may we be requirement indicative using of a constitutively o 

type Ire1* is already phosphorylated, thereby profiled mRNA expression on yeast genomic IRE1 (called IRE1 

alleviating requires a necessity an active for de kinase novo phosphoryl- domain, because microarrays. a Toconformational this end, mRNA from change wild- and/or change 

ation we inknow vitro. ADP that recombinant efficiently stimulated wild-type the type Ire1* IRE1, �ire1, in the and oligomeric IRE1(L745A,D828A) state of the enzymes in 

RNase activity of Ire1* (Fig. 3D, lane 1) but not cells was isolated at different points in time 

is already phosphorylated, thereby allevi- response to the binding of their respective 

that of Ire1*(L745G) (Fig. 3D, lane 2). after the addition of 1NM-PP1 and DTT. cDating 

For wild-type a necessity Ire1 and 1NM-PP1–sensitized 

for de novo phosphory- NAs derivedcofactors. from these mRNA populations 

Ire1(L745G), lation in ADP vitro. andADP 1NM-PP1, efficiently respectively, stimulated were labeled with 1NM-PP1–sensitized, different fluorescent dyes, kinase-dead 

function as stimulatory cofactors. Therefore, we mixed pairwise, and hybridized to microarrays 

asked the whether RNase a conformational activity of Ire1* change (Fig. occurs 3D, (20). lane Scatter Ire1 plots of signals pairwise comparisons a canonical of UPR. To ask 

in 1) response but not to cofactor that of binding Ire1*(L745G) by analyzing(Fig. 

mRNA 3D, expression whether levels bypassing are shown the Ire1 in kinase activity 

Ire1* and Ire1*(L745G) through glycerol-gradi- Fig. 5; the x-y scatter plots represent relative 

lane 2). 

with 1NM-PP1 allows the induction of a 

For wild-type Ire1 and 1NM-PP1–sen- full UPR, we profiled mRNA expression 

sitized Ire1(L745G), ADP and 1NM-PP1, on yeast genomic microarrays. To this end, 

respectively, function as stimulatory co- mRNA from wildtype IRE1, Δire1, and 

factors. Therefore, we asked whether a IRE1(L745A,D828A) cells was isolated 

C ) isolated previou 

random mutagenesis of the ER lume 

(18, 21). IRE1C significantly induce 

without DTT, resulting in about 7 

maximal activity of wild-type IRE1 

(Fig. 1D, bars 2 and 5). We pred 

chimera of the IRE1C Fig. 3. In vitro assays of Ire1 mutant proteins. Ire1* autophos 

assays with [�and 

1NM-PP1 

kinase-dead IRE1(L745A,D828A 

would activate the UPR with 1NM 

even in the absence of DTT. The dat 

(bars 13 to 16) show that this pred 

32P] ATP (A) and [�-32P]-labeled N-6 benzyl AT 

phorylated Ire1* proteins (wild-type and L745G) are indicated. I 

assays (against in vitro transcribed HAC1600 RNA) with stimula 

tors 1NM-PP1 (C) and ADP (D). Icons in (C) and (D) indicate HA 

cleavage products, denoting all combinations of singly and doub 

(E) The chemical structures of 1NM-PP1 and N-6 benzyl ATP a 

expressions for every mRNA between the specified 

genotypes. 

Fig. 4. Glycerol-density The comparison of wild-type IRE1- 

gradient velocity expressingsedicells with �ire1 cells revealed a 

mentation of distinct wild-type set of up-regulated UPR target genes in 

and mutant wild-type Ire1*. Ire1* IRE1-expressing cells (Fig. 5A; 

proteins (A genes to C) that and were up-regulated more than twofold 

Ire1*(L745G) are proteins shown in pink). These IRE1-dependent 

(D to F) were genes incubated include previously described UPR tran- 

in the presence scriptional of ADP targets (21, 22). Notably, the same 

[(B) and (E)] or set1NM-PP1 of (pink-colored) UPR target genes was 

[(C) and (F)] and up-regulated subject- to a similar magnitude in 

ed to velocity IRE1(L745A,D828A)-expressing sedimen- 

cells relative 

tation analysis. to �ire1Sedi cells (Fig. 5B), suggesting that a ca- 

lane 2). The cleavage reaction was not affected ent velocity sedimentation. Ire1* sedimented mentation as isnonical from UPR left was induced in the 1NM-PP1– 

by 1NM-PP1 (Fig. 3C, lane 3). In contrast, a broad peak in the gradient (Fig. 4). to Upon right. the sensitized, kinase-dead mutant. This conclusion 

Ire1*(L745G) was completely inactive, whereas addition of ADP (Fig. 4B), but not 1NM-PP1 was confirmed by the tight superimposition of 

the addition of 1NM-PP1, at the same concen- (Fig. 4C), the peak of Ire1* shifted as a result of the scatter diagonals of both UPR targets and 

tration that was shown in Fig. 3A (lane 4) to faster sedimentation. Conversely, the peak of the rest of the transcriptome evident in the 

completely inhibit the kinase activity, restored Ire1*(L745G) shifted by a corresponding expression profiles of wild-type IRE1- com- 

RNase activity significantly (Fig. 3C, lanes 4 and amount upon the addition of 1NM-PP1 (Fig. 4F) pared with IRE1(L745A,D828A)-expressing 

5), strongly supporting the notion that 1NM-PP1 but not upon the addition of ADP (Fig. 4E). This cells (Fig. 5C). 

uncouples the kinase and RNase activities of shift may be indicative of a conformational Building an instructive UPR switch.The 


change and/or change in the oligomeric state of 1NM-PP1–sensitized IRE1 mutants described so 

We have previously shown that cleavage of the enzymes in response to the binding of their far act permissively, because activation requires 

HAC1 mRNA by wild-type Ire1* is stimulated respective cofactors. 

both 1NM-PP1 and protein-misfolding agents. 

by adenosine 5�-diphosphate (ADP) (7). This is 1NM-PP1–sensitized, kinase-dead Ire1 This indicates that an unfolded-protein signal 

consistent with the notion derived from in vivo signals a canonical UPR. To ask whether needs to be received by the ER lumenal domain 

studies that Ire1 requires an active kinase do- bypassing the Ire1 kinase activity with 1NM- for activation. We asked if we could bypass this 

main, because we know that recombinant wild- PP1 allows the induction of a full UPR, we requirement using a constitutively on version of 


alleviating a necessity for de novo phosphoryl- microarrays. To this end, mRNA from wildation 

in vitro. ADP efficiently stimulated the type IRE1, �ire1, and IRE1(L745A,D828A) 


that of Ire1*(L745G) (Fig. 3D, lane 2). after the addition of 1NM-PP1 and DTT. cD- 

For wild-type Ire1 and 1NM-PP1–sensitized NAs derived from these mRNA populations 

Ire1(L745G), ADP and 1NM-PP1, respectively, were labeled with different fluorescent dyes, 


asked whether a conformational change occurs (20). Scatter plots of pairwise comparisons of 

in response to cofactor binding by analyzing mRNA expression levels are shown in 


www.sciencemag.org SCIENCE VOL 302 28 NOVEMBER 2003 

C ) isolated previously through 

random mutagenesis of the ER lumenal domain 

(18, 21). IRE1C significantly induced the UPR 

without DTT, resulting in about 75% of the 

maximal activity of wild-type IRE1 with DTT 

(Fig. 1D, bars 2 and 5). We predicted that a 

chimera of the IRE1C assays with [�and 

1NM-PP1–sensitized, 

kinase-dead IRE1(L745A,D828A) mutants 

would activate the UPR with 1NM-PP1 alone, 

even in the absence of DTT. The data in Fig. 1D 

(bars 13 to 16) show that this prediction holds 

32P] ATP (A) and [�-32P]-labeled N-6 benzyl ATP (B). Phosphorylated 

Ire1* proteins (wild-type and L745G) are indicated. Ire1* RNase 

assays (against in vitro transcribed HAC1600 RNA) with stimulatory cofactors 

1NM-PP1 (C) and ADP (D). Icons in (C) and (D) indicate HAC1600 RNA 

cleavage products, denoting all combinations of singly and doubly cut RNA. 

(E) The chemical structures of 1NM-PP1 and N-6 benzyl ATP are shown. 

Fig. 4. Glycerol-density expressions for every mRNA between the spec- 

gradient velocity sediified genotypes. 

mentation of wild-type The comparison of wild-type IRE1- 

and mutant Ire1*. Ire1* 

expressing cells with �ire1 cells revealed a 

proteins (A to C) and 

Ire1*(L745G) proteins 

distinct set of up-regulated UPR target genes in 

(D to F) were incubated wild-type IRE1-expressing cells (Fig. 5A; 

in the presence of ADP genes that were up-regulated more than twofold 

[(B) and (E)] or 1NM-PP1 are shown in pink). These IRE1-dependent 

[(C) and (F)] and subject- genes include previously described UPR traned 

to velocity sedimenscriptional 

targets (21, 22). Notably, the same 

tation analysis. Sedimentation 

is from left 

set of (pink-colored) UPR target genes was 

to right. 

up-regulated to a similar magnitude in 

IRE1(L745A,D828A)-expressing cells relative 

to �ire1 cells (Fig. 5B), suggesting that a ca- 

lane 2). The cleavage reaction was not affected ent velocity sedimentation. Ire1* sedimented as nonical UPR was induced in the 1NM-PP1– 

by 1NM-PP1 (Fig. 3C, lane 3). In contrast, a broad peak in the gradient (Fig. 4). Upon the sensitized, kinase-dead mutant. This conclusion 

Ire1*(L745G) was completely inactive, whereas addition of ADP (Fig. 4B), but not 1NM-PP1 was confirmed by the tight superimposition of 

the addition of 1NM-PP1, at the same concen- (Fig. 4C), the peak of Ire1* shifted as a result of the scatter diagonals of both UPR targets and 

tration that was shown in Fig. 3A (lane 4) to faster sedimentation. Conversely, the peak of the rest of the transcriptome evident in the 

completely inhibit the kinase activity, restored Ire1*(L745G) shifted by a corresponding expression profiles of wild-type IRE1- com- 

RNase activity significantly (Fig. 3C, lanes 4 and amount upon the addition of 1NM-PP1 (Fig. 4F) pared with IRE1(L745A,D828A)-expressing 

5), strongly supporting the notion that 1NM-PP1 but not upon the addition of ADP (Fig. 4E). This cells (Fig. 5C). 

uncouples the kinase and RNase activities of shift may be indicative of a conformational Building an instructive UPR switch.The 


change and/or change in the oligomeric state of 1NM-PP1–sensitized IRE1 mutants described so 

We have previously shown that cleavage of the enzymes in response to the binding of their far act permissively, because activation requires 

HAC1 mRNA by wild-type Ire1* is stimulated respective cofactors. 

both 1NM-PP1 and protein-misfolding agents. 

by adenosine 5�-diphosphate (ADP) (7). This is 1NM-PP1–sensitized, kinase-dead Ire1 This indicates that an unfolded-protein signal 

consistent with the notion derived from in vivo signals a canonical UPR. To ask whether needs to be received by the ER lumenal domain 

studies that Ire1 requires an active kinase do- bypassing the Ire1 kinase activity with 1NM- for activation. We asked if we could bypass this 

main, because we know that recombinant wild- PP1 allows the induction of a full UPR, we requirement using a constitutively on version of 


alleviating a necessity for de novo phosphoryl- microarrays. To this end, mRNA from wildation 

in vitro. ADP efficiently stimulated the type IRE1, �ire1, and IRE1(L745A,D828A) 


that of Ire1*(L745G) (Fig. 3D, lane 2). after the addition of 1NM-PP1 and DTT. cD- 

For wild-type Ire1 and 1NM-PP1–sensitized NAs derived from these mRNA populations 

Ire1(L745G), ADP and 1NM-PP1, respectively, were labeled with different fluorescent dyes, 


asked whether a conformational change occurs (20). Scatter plots of pairwise comparisons of 

in response to cofactor binding by analyzing mRNA expression levels are shown in 


www.sciencemag.org SCIENCE VOL 302 28 NOVEMBER 2003 1535 

C ) isolated previously through 

random mutagenesis of the ER lumenal domain 

(18, 21). IRE1C significantly induced the UPR 

without DTT, resulting in about 75% of the 

maximal activity of wild-type IRE1 with DTT 

(Fig. 1D, bars 2 and 5). We predicted that a 

chimera of the IRE1C assays with [�and 

1NM-PP1–sensitized, 

kinase-dead IRE1(L745A,D828A) mutants 

would activate the UPR with 1NM-PP1 alone, 

even in the absence of DTT. The data in Fig. 1D 

(bars 13 to 16) show that this prediction holds 

32P] ATP (A) and [�-32 Fig. 3. In vitro assays P]-labeledof N-6Ire1 benzyl mutant ATP (B). Phosphorylated 

Ire1* proteins (wild-type and L745G) are indicated. Ire1* RNase 

proteins. Ire1* autophosphorylation 

assays (against in vitro transcribed HAC1600 RNA) with stimulatory cofactors 

1NM-PP1 (C) assays and ADP with (D). Icons [γ-in (C) and (D) indicate HAC1600 RNA 

cleavage products, denoting all combinations of singly and doubly cut RNA. 

(E) The chemical structures of 1NM-PP1 and N-6 benzyl ATP are shown. 

Fig. 4. Glycerol-density 

gradient velocity sedimentation 

of wild-type 

and mutant Ire1*. Ire1* 

proteins (A to C) and 

Ire1*(L745G) proteins 

(D to F) were incubated 

in the presence of ADP 

[(B) and (E)] or 1NM-PP1 

[(C) and (F)] and subjected 

to velocity sedimentation 

analysis. Sedimentation 

is from left 

to right. 

Fig. 4. Glycerol-density gradient velocity sedimentation 

of wild-type and mutant Ire1*. Ire1* proteins (A to C) 

and Ire1*(L745G) proteins (D to F) were incubated in the 

presence of ADP [(B) and(E)] or 1NM-PP1 [(C) and(F)] and 

subjected to velocity sedimentation analysis. Sedimentation 

is from left to right. 


32P] ATP (A) and[ γ- 

32P]-labeled N-6 benzyl ATP (B). 

Phosphorylated Ire1* proteins (wildtype 

and L745G) are indicated. 

Ire1* RNase assays (against in vitro 

transcribed HAC1 RNA) with 

600 

stimulatory cofactors 1NM-PP1 

(C) and ADP (D). Icons in (C) and 

(D) indicate HAC1 RNA cleavage 

600 

products, denoting all combinations 

of singly and doubly cut RNA. (E) The 

chemical structures of 1NM-PP1 and 

N-6 benzyl ATP are shown. 

8 

R E S E A R C H A 

R E S E A R C H A R T I C L E 

Fig. 3. In vitro assays of Ire1 mutant proteins. Ire1* autophosphorylation

1NM-PP1 alone (i.e., even in the absence of 

ER stress). 

identified. Here, we report that both the 

kinase activity and activation-segment phos- 

Fig. 5. (A to C) x-y scatter plot analysis of mRNA abundance. The plots (log 10 ) compare yeast 

genomic microarray analyses between the genotypes indicated. In each case, mRNA was purified 

from cells that were provided both DTT and 1NM-PP1. Pink dots represent genes with expression 

more than two times as high in wild-type cells as in �ire1 �ire1 cells. 

Fig. 6. Model of activation of 1NM-PP1–sensitized and wild-type Ire1. For both proteins, chaperone 

dissociation resulting from unfolded-protein accumulation causes oligomerization in the plane of 

the ER membrane, which is a precondition for further activation. For mutant Ire1 (A), (A), 1NM-PP1 

binding to the inactive kinase domain active site causes a conformational change, which activates 

the RNase. For wild-type Ire1 (B), (B), trans-autophosphorylation, with ATP, of the activation segment 

fully opens the kinase domain for binding of ADP or ATP, which activates the RNase. 

1536 

at different points in time after the 28 28addi NOVEMBER labeled 2003 with VOL different 302 SCIENCE fluorescent www.sciencemag.org 

dyes, 

tion of 1NM-PP1 and DTT. cDNAs de- mixed pairwise, and hybridized to mirived 

from these mRNA populations were croarrays (20). Scatter plots of pairwise 

9 

diated by 

RNase a 

Ire1, th 

mechani 

One m 

proposed 

model, 

Ire1 occ 

after the 

chaperon 

by latera 

membra 

ation of 

activatio 

(24, 25) 

allowing 

leading 

that acti 

In co 

the activ 

occurs e 

ation of 

sible tha 

PP1 can 

ment, w 

to the a 

ATP mo 

1NM-PP 

the unp 

opens tr 

frequenc 

rate of 

sensitize 

ADP and 

could b 

Based o 

drug bin 

rearrang 

the RNa 

phospho 

mimics 

phospho 

The m 

why 1NM 

1NM-PP1 

vate them 

by activa 

nal doma 

the obser 

may be e 

free muta 

well to c 

(or cons 

may be 

neighbori 

formation

comparisons of mRNA expression levels 

are shown in Fig. 5; the x-y scatter plots 

represent relative expressions for every 

mRNA between the specified genotypes. 

The comparison of wild-type IRE1- 

expressing cells with Δire1 cells revealed 

a distinct set of up-regulated UPR target 

genes in wild-type IRE1-expressing cells 

(Fig. 5A; genes that were up-regulated 

more than twofold are shown in pink). 

These IRE1-dependent genes include 

previously described UPR transcriptional 

targets (21, 22). Notably, the same set 

of (pink-colored) UPR target genes was 

up-regulated to a similar magnitude in 

IRE1(L745A,D828A)-expressing cells 

relative to Δire1 cells (Fig. 5B), suggesting 

that a canonical UPR was induced in 

the 1NM-PP1–sensitized, kinase-dead 

mutant. This conclusion was confirmed 

by the tight superimposition of the scatter 

diagonals of both UPR targets and the rest 

of the transcriptome evident in the expression 

profiles of wild-type IRE1- compared 

with IRE1(L745A,D828A)-expressing 

cells (Fig. 5C). 

Building an instructive UPR switch. 

The 1NM-PP1–sensitized IRE1 mutants 

described so far act permissively, because 

activation requires both 1NM-PP1 

and protein-misfolding agents. This indicates 

that an unfolded-protein signal 

needs to be received by the ER lumenal 

domain for activation. We asked if we 

could bypass this requirement using a 

constitutively on version of IRE1 (called 

IRE1 C ) isolated previously through random 

mutagenesis of the ER lumenal domain 

(18, 21). IRE1 C significantly induced 

the UPR without DTT, resulting in about 

75% of the maximal activity of wild-type 

IRE1 with DTT (Fig. 1D, bars 2 and 5). 

We predicted that a chimera of the IRE1 C 

and 1NM-PP1–sensitized, kinase-dead 

IRE1(L745A,D828A) mutants would activate 

the UPR with 1NM-PP1 alone, even 

in the absence of DTT. The data in Fig. 1D 

(bars 13 to 16) show that this prediction 

holds true. Whereas wild-type IRE1 re- 

10 

quires DTT for induction and is immune 

to 1NM-PP1 (Fig. 1D, bars 1 to 4), and 

IRE1(L745A,D828A) requires both DTT 

and 1NM-PP1 (Fig. 1D, bars 9 to 12), the 

chimeric IRE1 C (L745A,D828A) requires 

only 1NM-PP1 for activation, irrespective 

of the addition of DTT (Fig. 1D, bars 13 

to 16). Thus, IRE1 C (L745A,D828A) is an 

instructive switch that turns on the UPR 

upon the addition of 1NM-PP1 alone (i.e., 

even in the absence of ER stress). 

Summary and implications. Since 

our discovery that Ire1’s RNase initiates 

splicing of HAC1 mRNA, the function of 

its kinase domain has remained an enigma 

(7). Although mutational analyses have 

indicated a requirement for an enzymatically 

active kinase and have defined activation-segment 

phosphorylation sites (9), 

to date no targets besides Ire1 itself have 

been identified. Here, we report that both 

the kinase activity and activation-segment 

phosphorylation can be bypassed if the 

small ATP-mimic 1NM-PP1 is provided to 

a mutant enzyme to which it can bind. Instead 

of inhibiting the function served by 

ATP, 1NM-PP1 rectifies the signaling defect 

of 1NM-PP1–sensitized Ire1 mutants, 

leading to the induction of a canonical 

UPR (21). Thus, ligand binding to the kinase 

domain, rather than a phosphotransfer 

function mediated by the kinase activity, 

is required for RNase activation. The 

kinase module of Ire1, therefore, uses an 

unprecedented mechanism to propagate 

the UPR signal. 

One model that could explain our data 

is proposed in Fig. 6B. According to this 

model, autophosphorylation of wild-type 

Ire1 occurs in trans between Ire1 molecules 

after they have been released from 

their chaperone anchors and have oligomerized 

by lateral diffusion in the plane of 

the ER membrane (9, 23). Trans-autophosphorylation 

of the activation segment locks 

the activation segment in the “swung-out” 

state (24, 25), and fully opens the active 

site, allowing ADP or ATP to bind efficiently, 

leading in turn to conformational

changes that activate the RNase domain. 

In contrast, the binding of 1NM-PP1 

to the active site of 1NM-PP1–sensitized 

Ire1 occurs even in the absence of phosphorylation 

of the activation segment. It 

is plausible that because of its small size, 

1NMPP1 can bypass the “closed” activation 

segment, which is proposed to occlude 

access to the active site for the larger ADP 

and ATP molecules (Fig. 6A). Alternatively, 

1NM-PP1 may enter the active site 

when the unphosphorylated activation segment 

opens transiently, which occurs with 

low frequency in other kinases (26). If the 

offrate of 1NM-PP1 binding to 1NM-PP1– 

sensitized Ire1 is slow compared to that of 

ADP and ATP, most mutant Ire1 molecules 

could be trapped in a drug-bound state. 

Based on either scenario, we propose that 

drug binding alone causes conformational 

rearrangements in the kinase that activate 

the RNase. The nucleotide-bound state of 

phosphorylated wild-type Ire1, therefore, 

mimics the 1NM-PP1–bound state of unphosphorylated 


The model discussed so far does not 

explain why 1NM-PP1 should not bind 

to individual 1NM-PP1–sensitized Ire1 

molecules and activate them, even without 

oligomerization induced by activation of 

function. 

the UPR 

Currently, 

through 

our 

the 

data 

ER 

do 

lumenal 

not allow 

domain. 

us to 

distinguish between these possibilities. 

Two 

Our 

possibilities 

findings provoke 

could 

the 

account 

question 

for 

of 

the 

what 

ob- 

the servations: “natural” stimulatory (i) The binding ligandof of 1NM-PP1 Ire1’s kinase 

may domain be enhanced may be. in One oligomerized, likely candidate chap- is 

ADP, erone-free which mutants, is generated or (ii) by Ire1 1NM-PP1 from ATP may 

during bind equally the autophosphorylation well to chaperone-anchored 

reaction. In 

vitro, 

and 

ADP 

chaperone-free 

is a better activator 

(or constitutive-on) 

of Ire1 than 

ATP (7). Ire1 would therefore be “self priming,” 

Ire1, 

generating 

but oligomerization 

an efficient activator 

may be required 

already 

bound because to itsinteraction active site. between In many professional neighboring 

secretory molecules cells is required such as the for �-cells the conforma- of the 

endocrine tional change pancreas that (which activates contain the RNase high funcconcentrationstion. Currently, of Ire1), our data ADPdo levels not allow rise (and us to 

ATP levels decline) temporarily in proportion 

distinguish between these possibilities. 

to nutritional stress (27). ADP therefore is 

physiologically 

Our findings 

poised 

provoke 

to serve 

the 

as 

question 

a cofactor 

of 

that what could the “natural” signal a starvation stimulatory state. ligand It is of 

known Ire1’s that kinase protein domain folding may becomes be. One ineffi- likely 

cient candidate as the nutritional is ADP, which status of is cells generated declines, by 

triggering Ire1 from the ATP UPR during (28). the Thus, autophosphory- 

in the face of 

ATP depletion, ADP-mediated conformalation 

reaction. In vitro, ADP is a better 

tional changes might increase the dwell time 

of activated Ire1. As such, Ire1 could have 

evolved this regulatory mechanism to monitor 

the energy balance of the cell and couple 

activator of Ire1 than ATP (7). Ire1 would 

therefore be “self priming,” generating an 

efficient activator already bound to its active 

site. In many professional secretory 

cells such as the β-cells of the endocrine 

pancreas (which contain high concentrations 

of Ire1), ADP levels rise (and ATP 

levels decline) temporarily in proportion 

to nutritional stress (27). ADP therefore 

is physiologically poised to serve as a cofactor 

that could signal a starvation state. 

It is known that protein folding becomes 

inefficient as the nutritional status of cells 

declines, triggering the UPR (28). Thus, in 

the face of ATP depletion, ADP-mediated 

conformational changes might increase 

the dwell time of activated Ire1. As such, 

Ire1 could have evolved this regulatory 

mechanism to monitor the energy balance 

of the cell and couple this information to 

activation of the UPR. 

Although unprecedented in proteins 

containing active kinase domains, our 

findings are consistent with previous observations 

of RNase L, which is closely 

related to Ire1, bearing a kinase-like domain 

followed C-terminally by an RNase 

domain (29, 30). Similar to Ire1, RNase 

L is activated by dimerization (31). However, 

RNase L’s kinase domain is naturally 

Similar 

inactive 

to 

(29), 

Ire1, 

whereas 

RNase 

its 

L 

RNase 

is activated 

activity 

by 

is 

dimerization (31). However, RNase L’s kinase 

stimulated 

domain is 

by 

naturally 

adenosine 

inactive 

nucleotide 

(29), wherebindasing 

itsto RNase the kinase activitydomain. is stimulated Therefore, by adeno- like 

sine Ire1, nucleotide the ligand-occupied binding to the kinase domain. domain 

Therefore, of RNase like L serves Ire1, the as ligand-occupied a module that parkinaseticipates domain in activation of RNase Land serves regulation as a module of the 

that 

RNase 

participates 

function. 

in 

Insights 

activation 

gained 

and regulation 

from Ire1 

of the RNase function. Insights gained from 

Ire1 

and 

and 

RNase 

RNase 

L may 

L may 

extend 

extend 

to other 

to other 

proteins 

proteins 

containing containing kinase kinase or or enzymatically enzymaticallyinac inactivetive 

pseudokinase domains (32). 

References and Notes 

1. M. J. Gething, J. Sambrook, Nature 355, 33 (1992). 

2. F. J. Stevens, Y. Argon, Semin. Cell Dev. Biol. 10, 443 

(1999). 

3. C. Patil, P. Walter, Curr. Opin. Cell Biol. 13, 349 (2001). 

4. J. S. Cox, C. E. Shamu, P. Walter, Cell 73, 1197 (1993). 

5. J. S. Cox, P. Walter, Cell 87, 391 (1996). 

6. U. Ruegsegger, J. H. Leber, P. Walter, Cell 107, 103 

(2001). 

7. C. Sidrauski, P. Walter, Cell 90, 1031 (1997). 

8. C. Sidrauski, J. S. Cox, P. Walter, Cell 87, 405 (1996). 

9. C. E. Shamu, P. Walter, EMBO J. 15, 3028 (1996). 

10. A. Weiss, J. Schlessinger, Cell 94, 277 (1998). 

11. K. Mori, W. Ma, M. J. Gething, J. Sambrook, Cell 74, 

743 (1993). 

11 

12. K. Shah, Y. Liu, C. Deirmengian, K. M. Shokat, Proc. 

Natl. Acad. Sci. U.S.A. 94, 3565 (1997). 

18. F. R. Papa, 

19. M. Huse, 

20. J. L. DeRisi, 

21. K. J. Trave 

22. Materials 

material o 

23. F. Urano, A 

24. S. R. Hubb 

Chem. 27 

25. T. Schindl 

26. M. Porter 

Chem. 27 

27. F. Schuit, 

Chem. 27 

28. R. J. Kaufm 

(2002). 

29. B. Dong, M 

361 (200 

30. S. Naik, J. 

Res. 26, 1 

31. B. Dong, R. 

32. M. Kroihe 

(2001). 

33. We thank 

Ullrich fo 

tants, and 

Shokat la 

grants fro 

(AI44009 

Molecular 

support a 

Medical In 

Supporting O

12 

statistically and experimentally (1). How- 

ever, characterizing polymorphisms locat- 

esineffi- declines, 

e face of 

onforma- 

well time 

uld have 

to moni- 

d couple 

UPR. 

eins con- 

findings 

ations of 

to Ire1, 

owed C- 

(29, 30). 


(2001). 






743 (1993). 



13. A. C. Bishop et al., Curr. Biol. 8, 257 (1998). 

14. E. L. Weiss, A. C. Bishop, K. M. Shokat, D. G. Drubin, 

Nature Cell Biol. 2, 677 (2000). 

15. A. C. Bishop et al., Nature 407, 395 (2000). 

16. Single-letter abbreviations for the amino acid resi- 

dues are as follows: A, Ala; D, Asp; G, Gly; L, Leu; and 

S, Ser. 

17. A. S. Carroll, A. C. Bishop, J. L. DeRisi, K. M. Shokat, E. K. 

O’Shea, Proc. Natl. Acad. Sci. U.S.A. 98, 12578 (2001). 

33. We thank M. Stark for constructing the IRE1 allele, S. 

Ullrich for his initial effort in constructing Ire1 mu- 

tants, and J. Kuriyan and members of the Walter and 

Shokat labs for valuable discussions. Supported by 

grants from the NIH to P.W. (GM32384) and K.S. 

(AI44009), and postdoctoral support from the UCSF 

Molecular Medicine Program (to F.P.). P.W. receives 

support as an investigator of the Howard Hughes 

Medical Institute. 

Supporting Online Material 

www.sciencemag.org/cgi/content/full/1090031/DC1 

Materials and Methods 

Microarray Excel Spreadsheets S1 to S3 


4 August 2003; accepted 1 October 2003 

Published online 16 October 2003; 

10.1126/science.1090031 

Include this information when citing this paper. 

REPORTS 

ion of the Inverse 

ppler Effect 

ddon* and T. Bearpark 

ervation of an inverse Doppler shift, in which the 

ased on reflection from a receding boundary. This 

een produced by reflecting a wave from a moving 

transmission line. Doppler shifts produced by this 

eproducible manner by electronic control of the 

ically five orders of magnitude greater than those 

with kinematic velocities. Potential applications 

tunable and multifrequency radiation sources. 

wn phe- 

a wave is 

locity of 

the source and the observer (1, 2). Our con- 

ventional understanding of the Doppler ef- 

fect, from the schoolroom to everyday expe- 

rience of passing vehicles, is that increased 

frequencies are measured when a source and 

observer approach each other. Applications 

of the effect are widely established and in- 

clude radar, laser vibrometry, blood flow 

measurement, and the search for new astro- 

nomical objects. The inverse Doppler effect 

refers to frequency shifts that are in the op- 

posite sense to those described above; for 

example, increased frequencies would be 

measured on reflection of waves from a re- 

ceding boundary. Demonstration of this 

counterintuitive phenomenon requires a fun- 

damental change in the way that radiation is 

reflected from a moving boundary, and, de- 

spite a wide range of theoretical work that 

spans the past 60 years, the effect has not 

previously been verified experimentally. 

Although inverse Doppler shifts have 

been predicted to occur in particular disper- 

sive media (3–8) and in the near zone of 

three-dimensional dipoles (9–11), these 

schemes are difficult to implement experi- 

mentally and have not yet been realized. 

However, recent advances in the design of 

composite condensed media (metamaterials) 

offers new and exciting possibilities for the 

control of radiation. In particular, materials 

with a negative refractive index (NRI) (12– 

14) and the use of shock discontinuities in 

photonic crystals (15) and transmission lines 

, Advanced 

ffice Box 5, 

dressed. E- 


ivated by 

e L’s ki- 

), where- 

yadeno- e domain. 

upied ki- 

a module 

egulation 

ined from 

ther pro- 

tically in- 

33 (1992). 

iol. 10, 443 

349 (2001). 

197 (1993). 

. 

ell 107, 103 

97). 

405 (1996). 

8 (1996). 

998). 

ok, Cell 74, 

hokat, Proc. 

98). 

. G. Drubin, 

00). 

o acid resi- 

; L, Leu; and 

Shokat, E. K. 

78 (2001). 

18. F. R. Papa, C. Zhang, K. Shokat, P. Walter, data not shown. 

19. M. Huse, J. Kuriyan, Cell 109, 275 (2002). 

20. J. L. DeRisi, V. R. Iyer, P. O. Brown, Science 278, 680 (1997). 

21. K. J. Travers et al., Cell 101, 249 (2000). 

22. Materials and methods are available as supporting 

material on Science Online. 

23. F. Urano, A. Bertolotti, D. Ron, J. Cell Sci. 113, 3697 (2000). 

24. S. R. Hubbard, M. Mohammadi, J. Schlessinger, J. Biol. 

Chem. 273, 11987 (1998). 

25. T. Schindler et al., Mol. Cell 3, 639 (1999). 

26. M. Porter, T. Schindler, J. Kuriyan, W. T. Miller, J. Biol. 

Chem. 275, 2721 (2000). 

27. F. Schuit, K. Moens, H. Heimberg, D. Pipeleers, J. Biol. 

Chem. 274, 32803 (1999). 

28. R. J. Kaufman et al., Nature Rev. Mol. Cell Biol. 3, 411 

(2002). 

29. B. Dong, M. Niwa, P. Walter, R. H. Silverman, RNA 7, 

361 (2001). 

30. S. Naik, J. M. Paranjape, R. H. Silverman, Nucleic Acids 

Res. 26, 1522 (1998). 

31. B. Dong, R. H. Silverman, Nucleic Acids Res. 27, 439 (1999). 

32. M. Kroiher, M. A. Miller, R. E. Steele, Bioessays 23, 69 

(2001). 

33. We thank M. Stark for constructing the IRE1 C allele, S. 
















10.1126/science.1090031 


REPORTS 

he 

his 

ing 











R E S E A R C H A R T I C L E 

Ire1 than 

elf prim- 

r already 

fessional 

s of the 

high con- 

rise (and 

roportion 

refore is 

cofactor 

te. It is 

es ineffi- 

declines, 

e face of 

onforma- 

well time 

uld have 

to moni- 

d couple 

UPR. 

eins con- 

findings 

ations of 

to Ire1, 

owed C- 

(29, 30). 

that participates in activation and regulation 


Ire1 and RNase L may extend to other pro- 

teins containing kinase or enzymatically in- 

active pseudokinase domains (32). 




(1999). 





(2001). 






743 (1993). 









S, Ser. 





Chem. 275, 2721 (2000). 


Chem. 274, 32803 (1999). 


(2002). 


361 (2001). 


Res. 26, 1522 (1998). 



(2001). 

















10.1126/science.1090031 


REPORTS 


ppler Effect 










wn phe- 

a wave is 

locity of 



































, Advanced 

fice Box 5, 

dressed. E- 


n squirrels 

pared with 

od supple- 

reds have 

size or a 

ce that the 

bserved in 

increased 

). We hy- 

resource- 

investment 

ounts, re- 

ure fitness 

roduction, 

ailable re- 

augmenta- 

show that 

rger litters 

cause off- 

ver, when 

juveniles is 

offspring, 

population 

n of young 

come with 

ause over- 

r years of 

for Amer- 

e previous 

s surviving 

t 15 = –0.7, 

portion of 

ersus adult 

.37 ± 0.27, 

amp-and- 

tion, then 

ion growth 

y the pred- 

tive output 

current but 

ggest that 

ms is more 

an present 

seed mast- 

vivalpros- e resource 

allel in re- 

predators 

28, 863 

Univ. of 

tal Care 

). 

May, Ed. 

don Ser. B 

M. Schauber, 

, 232 (2000). 

9. L. M. Curran, M. Leighton, Ecol. Monogr. 70, 101 (2000). 

10. W. D. Koenig, J. M. H. Knops, Am. Sci. 93, 340 (2005). 

11. D. H. Janzen, Annu. Rev. Ecol. Syst. 2, 465 (1971). 

12. A. F. Hedlin, For. Sci. 10, 124 (1964). 

13. M. J. McKone, D. Kelly, A. L. Harrison, J. J. Sullivan, 

A. J. Cone, N.Z. J. Zool. 28, 89 (2001). 

14. P. R. Wilson, B. J. Karl, R. J. Toft, J. R. Beggs, Biol. Conserv. 

83, 175 (1998). 

15. T. Ruf, J. Fietz, W. Schlund, C. Bieber, Ecology 87, 372 

(2006). 

16. F. H. Bronson, Reproductive Biology (Univ. of Chicago 

Press, Chicago, 1989). 

17. Materials and methods are available on Science Online. 

18. J. O. Wolff, J. Mammal. 77, 850 (1996). 

19. A. S. McNeilly, Reprod. Fertil. Dev. 13, 583 (2001). 

20. J. D. Ligon, Nature 250, 80 (1974). 

21. P. J. Berger, N. C. Negus, E. H. Sanders, P. D. Gardner, 

Science 214, 69 (1981). 

22. S. Eis, J. Inkster, Can. J. For. Res. 2, 460 (1972). 

23. B. M. Fitzgerald, M. J. Efford, B. J. Karl, N.Z. J. Zool. 31, 

167 (2004). 

24. M. C. Smith, J. Wildl. Manag. 32, 305 (1968). 

25. D. A. Rusch, W. G. Reeder, Ecology 59, 400 (1978). 

26. L. A. Wauters, C. Swinnen, A. A. Dhondt, J. Zool. 227, 71 

(1992). 

27. L. A. Wauters, A. A. Dhondt, J. Zool. 217, 93 (1989). 

28. C. D. Becker, Can. J. Zool. 71, 1326 (1993). 

29. K. W. Larsen, C. D. Becker, S. Boutin, M. Blower, 

J. Mammal. 78, 192 (1997). 

30. M. M. Humphries, S. Boutin, Ecology 81, 2867 

(2000). 

31. A. G. McAdam, S. Boutin, Evolution 57, 1689 (2003). 

32. L. A. Wauters, E. Matthysen, F. Adriaensen, G. Tosi, 

J. Anim. Ecol. 73, 11 (2004). 

33. We thank J. Herbers, C. Aumann, J. LaMontagne, 

M. Andruskiw, J. Lane, R. Boonstra, K. McCann, 

B. Thomas, Kluane Red Squirrel Project 2005 annual 

meeting participants, and three anonymous reviewers for 

very useful comments. We are indebted to the many field 

workers who helped collect the data and to A. Sykes and 

E. Anderson for management of the long-term data. 

Financed by Natural Sciences and Engineering Research 

Council of Canada, NSF (American reds), a Concerted 

Action of the Belgian Ministry of Education, and a 

Training and Mobility of Researchers grant from the 

Commission of the European Union (Eurasian reds). This 

is publication 30 of the Kluane Red Squirrel Project. 


www.sciencemag.org/cgi/content/full/314/5807/1928/DC1 


SOM Text 

Fig. S1 

Table S1 

References 

25 September 2006; accepted 15 November 2006 

10.1126/science.1135520 

Human Catechol-O-Methyltransferase 

Haplotypes Modulate Protein Expression 

by Altering mRNA Secondary Structure 

A. G. Nackley, 1 S. A. Shabalina, 2 I. E. Tchivileva, 1 K. Satterfield, 1 O. Korchynskyi, 3 

S. S. Makarov, 4 W. Maixner, 1 L. Diatchenko 1 * 

Catechol-O-methyltransferase (COMT) is a key regulator of pain perception, cognitive function, and 

affective mood. Three common haplotypes of the human COMT gene, divergent in two synonymous 

and one nonsynonymous position, code for differences in COMT enzymatic activity and are 

associated with pain sensitivity. Haplotypes divergent in synonymous changes exhibited the largest 

difference in COMT enzymatic activity, due to a reduced amount of translated protein. The major 

COMT haplotypes varied with respect to messenger RNA local stem-loop structures, such that the 

most stable structure was associated with the lowest protein levels and enzymatic activity. Site- 

directed mutagenesis that eliminated the stable structure restored the amount of translated 

protein. These data highlight the functional significance of synonymous variations and suggest the 

importance of haplotypes over single-nucleotide polymorphisms for analysis of genetic variations. 

T 

he ability to predict the downstream 

effects of genetic variation is critically 

important for understanding both the 

evolution of the genome and the molecular 

basis of human disease. The effects of non- 

synonymous polymorphisms have been wide- 

ly characterized; because these variations 

directly influence protein function, they are 

relatively easy to study statistically and 

experimentally (1). However, characterizing 

polymorphisms located in regulatory regions, 

which are much more common, has proved to 

be problematic (2). Here, we focus on the 

mechanism whereby polymorphisms of the 

cathechol-O-methyltransferase (COMT ) gene 

regulate gene expression. 

COMT is an enzyme responsible for de- 

grading catecholamines and thus represents a 

critical component of homeostasis maintenance 

(3). The human COMT gene encodes two 

distinct proteins: soluble COMT (S-COMT) 

and membrane-bound COMT (MB-COMT) 

1 Center for Neurosensory Disorders, University of North 

Carolina, Chapel Hill, NC 27599, USA. 2 National Center for 

Biotechnology Information, National Institutes of Health, 

Bethesda, MD 20894, USA. 3 Thurston Arthritis Center, 

University of North Carolina, Chapel Hill, NC 27599, USA. 

4 Attagene, Inc., 7030 Kit Creek Road, Research Triangle Park, 

NC 27560, USA. 

The ability to predict the downstream 

effects of genetic variation is criti- 

cally important for understanding 

both the evolution of the genome and the 

molecular basis of human disease. The ef- 

fects of nonsynonymous polymorphisms 

have been widely characterized; because 

these variations directly influence protein 

function, they are relatively easy to study 

feeding on seed (26, 27). Lastly, food supple- 

mentation experiments of American reds have 

failed to produce increases in litter size or a 

second litter, providing strong evidence that the 

increased reproductive investment observed in 

our study cannot be triggered by increased 

energy availability alone (17, 28, 29). We hy- 

pothesize that, rather than following a resource- 

tracking strategy where reproductive investment 

is determined by current resource amounts, re- 

productive rates are driven by future fitness 

payoffs. During years of low seed production, 

competition among juveniles for available re- 

sources is intense, and, although litter augmenta- 

tion experiments in American reds show that 

females are capable of supporting larger litters 

(30), they refrain from doing so because off- 

spring recruitment is low (30). However, when 

mast years occur, competition among juveniles is 

reduced, and females produce more offspring, 

which successfully recruit into the population 

(31, 32). Further, the increased production of young 

by females in mast years does not come with 

any obvious cost to the female, because over- 

winter survival is not reduced after years of 

increased reproductive investment (for Amer- 

ican reds, offspring production in the previous 

year versus proportion of adult females surviving 

to spring has slope = –0.023 ± 0.034, t 15 = –0.7, 

and P = 0.51; for Eurasian reds, proportion of 

estrous females in the previous year versus adult 

female survival to spring has slope = 0.37 ± 0.27, 

t 15 = 1.4, and P = 0.19). 

If masting has evolved as a swamp-and- 

starve adaptation against seed predation, then 

anticipatory reproduction and population growth 

represent a potent counteradaptation by the pred- 

ators. Given that increased reproductive output 

in these systems coincides with low current but 

high future resources, our results suggest that 

reproductive investment in these systems is more 

responsive to future fitness prospects than present 

energetic constraints. The evolution of seed mast- 

ing in trees is also driven by the survival pros- 

pects for progeny rather than simple resource 

tracking, suggesting an intriguing parallel in re- 

productive strategies of trees and the predators 

that consume their seed. 


1. J. L. Gittleman, S. D. Thompson, Am. Zool. 28, 863 

(1988). 

2. T. H. Clutton-Brock, Reproductive Success (Univ. of 

Chicago Press, Chicago, 1988). 

3. T. H. Clutton-Brock, The Evolution of Parental Care 

(Princeton Univ. Press, Princeton, NJ, 1999). 

4. G. Caughley, in Theoretical Ecology, R. M. May, Ed. 

(Blackwell, Oxford, 1981), ch. 6. 

5. P. Bayliss, D. Choquenot, Proc. R. Soc. London Ser. B 

357, 1233 (2002). 

6. C. G. Jones, R. S. Ostfeld, M. P. Richard, E. M. Schauber, 

J. O. Wolff, Science 279, 1023 (1998). 

7. R. S. Ostfeld, F. Keesing, Trends Ecol. Evol. 15, 232 (2000). 

8. D. Kelly, V. L. Sork, Annu. Rev. Ecol. Syst. 33, 427 

(2002). 

12. A. F. Hedlin, For. Sci. 10, 124 (1964). 

13. M. J. McKone, D. Kelly, A. L. Harrison, J. J. Sullivan, 

A. J. Cone, N.Z. J. Zool. 28, 89 (2001). 

14. P. R. Wilson, B. J. Karl, R. J. Toft, J. R. Beggs, Biol. Conserv. 

83, 175 (1998). 

15. T. Ruf, J. Fietz, W. Schlund, C. Bieber, Ecology 87, 372 

(2006). 

16. F. H. Bronson, Reproductive Biology (Univ. of Chicago 

Press, Chicago, 1989). 

17. Materials and methods are available on Science Online. 

18. J. O. Wolff, J. Mammal. 77, 850 (1996). 

19. A. S. McNeilly, Reprod. Fertil. Dev. 13, 583 (2001). 

20. J. D. Ligon, Nature 250, 80 (1974). 

21. P. J. Berger, N. C. Negus, E. H. Sanders, P. D. Gardner, 

Science 214, 69 (1981). 

22. S. Eis, J. Inkster, Can. J. For. Res. 2, 460 (1972). 

23. B. M. Fitzgerald, M. J. Efford, B. J. Karl, N.Z. J. Zool. 31, 

167 (2004). 

24. M. C. Smith, J. Wildl. Manag. 32, 305 (1968). 

25. D. A. Rusch, W. G. Reeder, Ecology 59, 400 (1978). 

26. L. A. Wauters, C. Swinnen, A. A. Dhondt, J. Zool. 227, 71 

(1992). 

27. L. A. Wauters, A. A. Dhondt, J. Zool. 217, 93 (1989). 

28. C. D. Becker, Can. J. Zool. 71, 1326 (1993). 

29. K. W. Larsen, C. D. Becker, S. Boutin, M. Blower, 

J. Mammal. 78, 192 (1997). 

32. L. A. Wa 

J. Anim. 

33. We than 

M. Andru 

B. Thom 

meeting 

very use 

workers 

E. Ander 

Financed 

Council 

Action o 

Training 

Commiss 

is public 

Supporting 

www.sciencem 

Materials and 

SOM Text 

Fig. S1 

Table S1 

References 

25 September 

10.1126/scien 

Human Catechol-O-Methy 

Haplotypes Modulate Prot 

by Altering mRNA Secon 

A. G. Nackley, 1 S. A. Shabalina, 2 I. E. Tchivileva, 1 K. Satte 

S. S. Makarov, 4 W. Maixner, 1 L. Diatchenko 1 * 

Catechol-O-methyltransferase (COMT) is a key regulator of pai 

affective mood. Three common haplotypes of the human COM 

and one nonsynonymous position, code for differences in CO 

associated with pain sensitivity. Haplotypes divergent in synon 

difference in COMT enzymatic activity, due to a reduced amo 

COMT haplotypes varied with respect to messenger RNA loca 

most stable structure was associated with the lowest protein 

directed mutagenesis that eliminated the stable structure res 

protein. These data highlight the functional significance of sy 

importance of haplotypes over single-nucleotide polymorphis 

T 

he ability to predict the downstream 

effects of genetic variation is critically 

important for understanding both the 

evolution of the genome and the molecular 

basis of human disease. The effects of non- 

synonymous polymorphisms have been wide- 

ly characterized; because these variations 

directly in 

relatively 

experimen 

polymorph 

which are 

be proble 

mechanism 

cathechol- 

regulate g 

COMT 

grading ca 

critical com 

(3). The 

distinct pr 

and mem 

through th 

1 Center for Neurosensory Disorders, University of North 

Carolina, Chapel Hill, NC 27599, USA. 2 National Center for 

Biotechnology Information, National Institutes of Health, 

Bethesda, MD 20894, USA. 3 Thurston Arthritis Center, 

University of North Carolina, Chapel Hill, NC 27599, USA. 

4 Attagene, Inc., 7030 Kit Creek Road, Research Triangle Park, 

NC 27560, USA. 

*To whom correspondence should be addressed. E-mail: 

lbdiatch@email.unc.edu 

22 DECEMBER 2006 VOL 314 SCIENCE www.sciencemag.org 

1930 

Ire1 than 

elf prim- 

r already 

fessional 

s of the 

high con- 

rise (and 

roportion 

refore is 

cofactor 

te. It is 

es ineffi- 

declines, 

e face of 

onforma- 

well time 

uld have 

to moni- 

d couple 

UPR. 

eins con- 

findings 

ations of 

to Ire1, 

owed C- 

(29, 30). 









(1999). 





(2001). 






743 (1993). 









S, Ser. 





Chem. 275, 2721 (2000). 


Chem. 274, 32803 (1999). 


(2002). 


361 (2001). 


Res. 26, 1522 (1998). 



(2001). 

















10.1126/science.1090031 


REPORTS 


ppler Effect 










wn phe- 

a wave is 

locity of 



































, Advanced 

fice Box 5, 

dressed. E- 


Ire1 than 

elf prim- 

r already 

fessional 

s of the 

high con- 

rise (and 

roportion 

refore is 

cofactor 

te. It is 

es ineffi- 

declines, 

e face of 

onforma- 

well time 

uld have 

to moni- 

d couple 

UPR. 

eins con- 

findings 

ations of 

to Ire1, 

owed C- 

(29, 30). 









(1999). 





(2001). 






743 (1993). 









S, Ser. 





Chem. 275, 2721 (2000). 


Chem. 274, 32803 (1999). 


(2002). 


361 (2001). 


Res. 26, 1522 (1998). 



(2001). 

















10.1126/science.1090031 


REPORTS 


ppler Effect 










wn phe- 

a wave is 

locity of 



































, Advanced 

fice Box 5, 

dressed. E- 

www.sciencemag.org SCIENCE VOL 302 28 NOVEMBER 2003 1537

A common SNP in codon 158 (val met) notype. Because variation in the S-COMT for the s 

has been associated with pain ratings and promoter region does not contribute to pain structure, 

longest, m 

in the va 

Fig. 1. Common haplotypes 

of the human COMT 

gene differ with respect 

to mRNA secondary structure 

and enzymatic activity. 

(A) A schematic 

diagram illustrates COMT 

MB-COM 

loop struc 

~17 kcal/m 

LPS haplo 

(Fig. 1B a 

porting p 

obtained 

genomic organization and 

ondary str 

SNP composition for the 

of COMT 

three haplotypes. Per- 

species (f 

cent frequency of each 

structures 

haplotype in a cohort of 

healthy Caucasian females, 

and percent independent 

SNP contribution 

to pain sensitivity, are 

indicated. (B) The local 

stem-loop structures associated 

with each of 

the three haplotypes are 

shown. Relative to the 

LPS and APS haplotypes, 

the HPS local stem-loop 

structure had a higher 

folding potential. (C and 

D) The LPS haplotype exhibited 

the highest, while 

the HPS haplotype exhib- 

highly sta 

gous to t 

stantial d 

observed 

notable f 

studies w 

modeling 

We co 

cDNA cl 

tors that 

correspon 

lotypes ( 

were tran 

six constr 

tein expr 

measured 

ited the lowest enzymatic 

HPS hap 

activity and protein lev- 

reduction 

els in cells expressing 

MB-COM 

COMT. ***P < 0.001, ≠ 

LPS. 

and fig. S 

ited mark 

protein ex 

APS hapl 

fold redu 

MB-COM 

+++ A common SNP in codon 158 (val met) notype. Because variation in the S-COMT for the 

has been associated with pain ratings and promoter region does not contribute to pain structur 

longest 

in the 

Fig. 1. Common haplotypes 

of the human COMT 

gene differ with respect 

to mRNA secondary structure 

and enzymatic activity. 

(A) A schematic 

diagram illustrates COMT 

MB-CO 

loop stru 

~17 kca 

LPS hap 

(Fig. 1B 

porting 

obtained 

genomic organization and 

ondary 

SNP composition for the 

of COM 

three haplotypes. Per- 

species 

cent frequency of each 

structur 

haplotype in a cohort of 

healthy Caucasian females, 

and percent independent 

SNP contribution 

to pain sensitivity, are 

indicated. (B) The local 

stem-loop structures associated 

with each of 

the three haplotypes are 

shown. Relative to the 

LPS and APS haplotypes, 

the HPS local stem-loop 

structure had a higher 

folding potential. (C and 

D) The LPS haplotype exhibited 

the highest, while 

the HPS haplotype exhib- 

highly 

gous to 

stantial 

observe 

notable 

studies 

modelin 

We c 

cDNA 

tors th 

corresp 

lotypes 

were tr 

six con 

tein ex 

measure 

ited the lowest enzymatic 

HPS h 

activity and protein lev- 

reductio 

els in cells expressing 

MB-CO 

COMT. ***P < 0.001, ≠ 

LPS. P < 0.001, ≠ APS. 

and fig 

ited ma 

protein 

APS ha 

fold red 

MB-CO 

+++ P < 0.001, ≠ APS. 

group demonstrated that three common 

haplotypes of the human COMT gene are 

associated with pain sensitivity and the 

likelihood of developing temporomandibular 

joint disorder (TMJD), a common 

chronic musculoskeletal pain condition 

(4). Three major haplotypes are formed 

by four single-nucleotide polymorphisms 

(SNPs): one located in the S-COMT promoter 

region (A/G; rs6269) and three in 

the S- and MB-COMT coding region at 

codons his62his (C/T; rs4633), leu136leu (C/ 

G; rs4818), and val158 ed in regulatory regions, which www.sciencemag.org are www.sciencemag.org much 

SCIENCE VOL VOL314 314 22 22 DECEMBER DECEMBER200 20 

more common, has proved to be problematic 

(2). Here, we focus on the mechanism 

whereby polymorphisms of the cathechol- 

O-methyltransferase (COMT) gene regulate 

gene expression. 

COMT is an enzyme responsible 

for degrading catecholamines and thus 

represents a critical component of homeostasis 

maintenance (3). The human 

COMT gene encodes two distinct proteins: 

soluble COMT (S-COMT) and 

membrane-bound COMT (MB-COMT) 

met (A/G; rs4680) 

through the use of alternative translation (Fig. 1A). On the basis of subjects’ pain 

initiation sites and promoters (3). Re- responsiveness, haplotypes were designatcently, 

COMT has been implicated in the ed as low (LPS; GCGG), average (APS; 

modulation of persistent pain (4–7). Our ATCA), or high (HPS; ACCG) pain sen- 

13

1932 

ture and converts convertsit it into a LPS haplotype–like 

structure (HPS Lsm). Double mutation of 

Fig. 2. Site-directed mutagenesis 

that destroys 

the stable stem-loop structure 

corresponding to to the 

HPS haplotype restores 

COMT enzymatic activity 

and protein expression. 

(A) The mRNA structure 

corresponding to to the HPS 

haplotype was converted 

to to an LPS haplotype-like 

structure (HPS Lsm) by 

single mutation of of 403C 

to to G. The original HPS 

haplotype structure (HPS 

dm) was restored by double 

mutation of of interacting 

nucleotides 403C 403Cto to G 

and 479G 479Gto toC. C. (B and C) 

The HPS Lsm exhibited 

COMT enzymatic activity 

and protein levels equivalentalentto 

to those thoseof of the LPS 

haplotype, whereas the 

HPS dm exhibited reduced 

enzymatic activity. 

***P < 0.001, ≠ LPS. 

sitive. Individuals 

carrying HPS/APS 

or APS/APS diplotypes 

were nearly 

2.5 times as likely 

to develop TMJD. 

Previous data further 

suggest that 

COMT haplotypes code for differences in 

COMT enzymatic activity (4); however, 

the molecular mechanisms involved have 

remained unknown. 

A common SNP in codon 158 (val158met) has been associated with pain ratings and 

μ-opioid system responses (7) as well as 

addiction, cognition, and common affective 

disorders (3, 8–10). The substitution 

of valine (Val) by methionine (Met) results 

in reduced thermostability and activity of 

the enzyme (3). It is generally accepted 

that val158met is the main source of individual 

variation in human COMT activity; 

numerous studies have identified associations 

between the low-activity met allele 

and several complex phenotypes (3, 8, 10). 

14 

verified by an alternate approach of modifying 

mRNA secondary structure (fig. S5). 

However, observed associations between 

these conditions and the met allele are often 

modest and occasionally inconsistent 

(3). This suggests that additional SNPs in 

the COMT gene modulate COMT activity. 

Indeed, we found that haplotype rather 

than an individual SNP better accounts for 

variability in pain sensitivity (4). The HPS 

and LPS haplotypes that both code for the 

stable val 158 variant were associated with 

the two extreme-pain phenotypes; thus, 

the effect of haplotype on pain sensitivity 

in our study cannot be explained by 

the sum of the effects of functional SNPs. 

Instead, the val 158 met SNP interacts with 

other SNPs to determine phenotype. Because 

variation in the S-COMT promoter 

Refere Refer 

1. 1. L. L. Y. Y. Ya Y 

Hum. M 

2. 2. J. J. C. C. Kn K 

3. 3. P. P. T. T. M 

(1999) (1999 

4. 4. L. L. Diat Dia 

(2005) (2005 

5. 5. J. J. J. J. Ma M 

(1976) (1976 

6. 6. T. T. T. T. Ra R 

7. 7. J. J. K. K. Zu Z 

8. 8. G. G. Win 

134 (2 (2 

9. 9. B. B. Funk Fun 

10. G. G. Oros Oro 

(2004) (2004 

11. J. J. Duan Dua 

(2003) (2003 

12. L. L. X. X. Sh S 

Acad. S 

13. I. I. Puga Pug 

14. K. K. Mita Mit 

203, 9 

15. T. T. D. D. S 

H. H. J. J. Le L 

(1994) (1994 

16. A. A. Shal Sha 

(2002) (2002 

17. Materia Materi 

materia materi 

18. D. D. H. H. M 

Biol. 2 

19. M. M. Zuk 

20. A. A. Y. Y. O 

M. M. A. A. R 

21. Sequen Seque 

CA4894 CA489 

haploty haplot 

three threeh 

site sitewe we 

enzyme enzym 

three threeS 

through throug 

(BI835 

ends endsas a 

22. S. S. A. A. Sh S 

Acids AcidsR 

22 DECEMBER 2006 VOL 314 SCIENCE www.sciencemag.or 

www.sciencemag.o

egion does not contribute to pain phenotype 

(Fig. 1A), we suggest that the rate of 

mRNA degradation or protein synthesis 

is affected by the structural properties of 

the haplotypes, such as haplotype-specific 

mRNA secondary structure. 

Previous reports have shown that 

polymorphic alleles can markedly affect 

mRNA secondary structure (11, 12), which 

can then have functional consequences on 

the rate of mRNA degradation (11, 13). 

It is also plausible that polymorphic alleles 

directly modulate protein translation 

through alterations in mRNA secondary 

structure, because protein translation efficiency 

is affected by mRNA secondary 

structure (14–16). To test these possibilities, 

we evaluated the effect of LPS, APS, 

and HPS haplotypes on the stability of the 

corresponding mRNA secondary structures 

(17). 

Secondary structures of the full-length 

LPS, APS, and HPS mRNA transcripts 

were predicted by means of the RNA 

Mfold (18, 19) and Afold (20) programs. 

The mRNA folding analyses demonstrated 

that the major COMT haplotypes 

differ with respect to mRNA secondary 

structure. The LPS haplotype codes for 

the shortest, least stable local stem-loop 

structure, and the HPS haplotype codes 

for the longest, most stable local stemloop 

structure in the val 158 region for both 

S-COMT and MB-COMT. Gibbs free energy 

(ΔG) for the stem-loop structure associated 

with the HPS haplotype is ~17 

kcal/mol less than that associated with 

the LPS haplotype for both S-COMT and 

MB-COMT (Fig. 1B and fig. S1A). Additional 

evidence supporting predicted 

RNA folding structures was obtained by 

generating consensus RNA secondary 

structures based on comparative analysis 

of COMT sequences from eight mammalian 

species (fig. S2). The consensus RNA 

folding structures were LPS-like and did 

not contain highly stable local stem-loop 

structures analogous to the human HPSlike 

form. Thus, substantial deviation 

from consensus structure, as observed for 

the HPS haplotype, should have notable 

functional consequences. Additional studies 

were conducted to test this molecular 

modeling. 

We constructed full-length S- and MB- 

COMT cDNA clones in mammalian expression 

vectors that differed only in three 

nucleotides corresponding to the LPS, APS, 

and HPS haplotypes (17, 21). Rat adrenal 

(PC-12) cells were transiently transfected 

with each of these six constructs. COMT 

enzymatic activity, protein expression, and 

mRNA abundance were measured. Relative 

to the LPS haplotype, the HPS haplotype 

showed a 25- and 18-fold reduction in 

enzymatic activity for S- and MB-COMT 

constructs, respectively (Fig. 1C and fig. 

S1B). The HPS haplotype also exhibited 

marked reductions in S- and MB-COMT 

protein expression (Fig. 1D and fig. S1C). 

The APS haplotype displayed a moderate 

2.5- and 3-fold reduction in enzymatic 

activity for S- and MB-COMT constructs, 

respectively, while protein expression levels 

did not differ. The moderate reduction 

in enzymatic activity produced by the APS 

haplotype is most likely due to the previously 

reported decrease in protein thermostability 

coded by the met 158 allele (3). 

These data illustrate that the reduced enzymatic 

activity corresponding to the HPS 

haplotype is paralleled by reduced protein 

levels, an effect that could be mediated by 

local mRNA secondary structure at the level 

of protein synthesis and/or mRNA degradation. 

Because total RNA abundance 

and RNA degradation rates did not parallel 

COMT protein levels (fig. S2), differences 

in protein translation efficiency likely results 

from differences in the local secondary 

structure of corresponding mRNAs 

To directly assess this hypothesis, we 

performed site-directed mutagenesis (17). 

The stable stem-loop structure of S- and 

MB-COMT mRNA corresponding to the 

HPS haplotype is supported by base pairs 

between several critical nucleotides, including 

403C and 479G in S-COMT and 

15

n thermo- 

3). These 

ic activity 

paralleled 

t could be 

tructure at 

r mRNA 

dance and 

el COMT 

n protein 

m differe 

of cor- 

esis, we 

sis (17). 

f S- and 

the HPS 

between 

ng 403C 

701G in 

Mutation 

to G in 

op structype–like 

tation of 

625C and 701G in MB-COMT (Fig. 2A 

and fig. S3A). Mutation of 403C to G 

in S-COMT or 625C to G in MB-COMT 

destroys the stable stem-loop structure 

and converts it into a LPS haplotype–like 

structure (HPS Lsm). Double mutation of 

mRNA in position 403C to G and 479G to 

C in S-COMT or 625C to G and 701G to 

C in MB-COMT reconstructs the original 

long stem-loop structure (HPS dm). The 

single- and double-nucleotide HPS mutants 

(HPS Lsm and HPS dm, respectively) 

were transiently transfected to PC-12 

cells. As predicted by the mRNA secondary-structure 

folding analyses, the HPS 

Lsm exhibited increased COMT enzymatic 

activity and protein levels equivalent to 

those of the LPS haplotype, whereas the 

HPS dm exhibited reduced enzymatic 

activity and protein levels equivalent to 

those of the original HPS haplotype (Fig. 

2, B and C; fig. S3, B and C). These data 

rule out the involvement of RNA sequence 

recognition motifs or codon usage in the 

regulation of translation. In contrast to 

the HPS haplotype, protein levels did not 

parallel COMT enzymatic activity for the 

APS haplotype and site-directed mutagenesis 

confirmed that the met158 structure (HPS dm). The single- and doublenucleotide 

HPS mutants (HPS Lsm and HPS 

dm, respectively) were transiently transfected to 

PC-12 cells. As predicted by the mRNA 

secondary-structure folding analyses, the HPS 

Lsm exhibited increased COMT enzymatic 

activity and protein levels equivalent to those 

of the LPS haplotype, whereas the HPS dm 

exhibited reduced enzymatic activity and protein 

levels equivalent to those of the original 

HPS haplotype (Fig. 2, B and C; fig. S3, B and 

C). These data rule out the involvement of RNA 

sequence recognition motifs or codon usage in 

the regulation of translation. In contrast to the 

HPS haplotype, protein levels did not parallel 

COMT enzymatic activity for the APS haplotype 

and site-directed mutagenesis confirmed that the 

met 

allele, 

not a more stable mRNA secondary structure, 

drives the reduced enzymatic activity 

observed for the APS haplotype (fig. 

S4). This difference is moderate relative 

to the mRNA structure–dependent difference 

coded by LPS and HPS haplotypes. 

These results were verified by an alternate 

approach of modifying mRNA secondary 

structure (fig. S5). 

Our data have very broad evolutionary 

and medical implications for the analysis of 

variants common in the human population. 

The fact that alterations in mRNA secondary 

structure resulting from synonymous 

changes have such a pronounced effect on 

the level of protein expression emphasizes 

the critical role of synonymous nucleotide 

positions in maintaining mRNA secondary 

structure and suggests that the mRNA 

secondary structure, rather than indepen- 

158 allele, not a more stable mRNA secondary 

structure, drives the reduced enzymatic 

activity observed for the APS haplotype (fig. 

S4). This difference is moderate relative to the 

mRNA structure–dependent difference coded by 

LPS and HPS haplotypes. These results were 

verified by an alternate approach of modifying 

mRNA secondary structure (fig. S5). 

2 DECEMBER 162006 

VOL 314 SCIENCE www.sciencemag.org 

The fact that alterations in mRNA secondary 

structure resulting from synonymous changes 

have such a pronounced effect on the level of 

protein expression emphasizes the critical role 

of synonymous nucleotide positions in maintaining 

dent nucleotides mRNA secondary in the synonymous structure andposi suggeststions, 

that should theundergo mRNA substantial secondary selective structure, 

rather pressure than (22). independent Furthermore, nucleotides our data in stress the 

synonymous the importance positions, of synonymous should undergo SNPs sub- as 

stantial 

potential 

selective 

functional 

pressure 

variants 

(22). 

in 

Furthermore, 

the area of 

our data stress the importance of synonymous 

SNPs 

human 

as 

medical 

potential 

genetics. 

functional 

Although 

variants in 

non- 

the 

area synonymous of humanSNPs medical are genetics. believed Although to have 

nonsynonymous the strongest impact SNPs on arevariation believed in to gene have 

the function, strongest our data impact clearly on variation demonstrate in gene that 

function, haplotypic ourvariants data clearly of common demonstrate synony- that 

haplotypic variants of common synonymous 

mous SNPs can have stronger effects on 

SNPs can have stronger effects on gene function 

gene than function nonsynonymous than nonsynonymous variations varia- and 

play tions an and important play an important role in disease role in onset disease and 

progression. onset and progression. 


1. L. Y. Yampolsky, F. A. Kondrashov, A. S. Kondrashov, 

Hum. Mol. Genet. 14, 3191 (2005). 

2. J. C. Knight, J. Mol. Med. 83, 97 (2005). 

3. P. T. Mannisto, S. Kaakkola, Pharmacol. Rev. 51, 593 

(1999). 

4. L. Diatchenko et al., Hum. Mol. Genet. 14, 135 

(2005). 

5. J. J. Marbach, M. Levitt, J. Dent. Res. 55, 711 

(1976). 

6. T. T. Rakvag et al., Pain 116, 73 (2005). 

7. J. K. Zubieta et al., Science 299, 1240 (2003). 

8. G. Winterer, D. Goldman, Brain Res. Brain Res. Rev. 43, 

134 (2003). 

9. B. Funke et al., Behav. Brain Funct. 1, 19 (2005). 

10. G. Oroszi, D. Goldman, Pharmacogenomics 5, 1037 

(2004). 

11. J. Duan et al., Hum. Mol. Genet. 12, 205 

(2003). 

12. L. X. Shen, J. P. Basilion, V. P. Stanton Jr., Proc. Natl. 

Acad. Sci. U.S.A. 96, 7871 (1999). 

13. I. Puga et al., Endocrinology 146, 2210 (2005). 

14. K. Mita, S. Ichimura, M. Zama, T. C. James, J. Mol. Biol. 

203, 917 (1988). 

15. T. D. Schmittgen, K. D. Danenberg, T. Horikoshi, 

H. J. Lenz, P. V. Danenberg, J. Biol. Chem. 269, 16269 

(1994). 

16. A. Shalev et al., Endocrinology 143, 2541 

(2002). 

17. Materials and methods are available as supporting 

material on Science Online. 

18. D. H. Mathews, J. Sabina, M. Zuker, D. H. Turner, J. Mol. 

Biol. 288, 911 (1999). 

19. M. Zuker, Nucleic Acids Res. 31, 3406 (2003). 

20. A. Y. Ogurtsov, S. A. Shabalina, A. S. Kondrashov, 

M. A. Roytberg, Bioinformatics 22, 1317 (2006). 

21. Sequence accession numbers: S-COMT clones BG290167, 

CA489448, and BF037202 represent LPS, APS, and HPS 

haplotypes, respectively. Clones corresponding to all 

three haplotypes and including the transcriptional start 

site were constructed by use of the unique restriction 

enzyme Bsp MI. The gene coding region containing all 

three SNPs was cut and inserted into the plasmids 

through use of the S-COMT (BG290167) and MB-COMT 

(BI835796) clones containing the entire COMT 5′ and 3′ 

ends as a backbone vector. 

22. S. A. Shabalina, A. Y. Ogurtsov, N. A. Spiridonov, Nucleic 

Acids Res. 34, 2428 (2006).

e of Child 

the NIH, National Center for Biotechnology 

SOM Text 

nal Institute of Information. 

Figs. S1 to S6 

National 

References 

arch at the NIH Supporting Online Material 

esearch was www.sciencemag.org/cgi/content/full/314/5807/1930/DC1 13 June 2006; accepted 16 November 2006 

rch Program of Materials and Methods 

10.1126/science.1131262 

cidophilic 

aled by Community 

ysis 

ichard I. Webb, 

17 

3 Judith Flanagan, 2 * 

, 2 ‡ Jillian F. Banfield 1,2 motes further pyrite dissolution, leading to AMD 

generation. AMD solutions forming underground 

in the Richmond Mine at Iron Mountain, 

California, are warm (30° to 59°C), acidic (pH 

~0.5 to 1.5), metal-rich [submolar Fe 

§ 

cies can be easily overlooked in standard polymerase chain 

sed community genomic data obtained without PCR or 

ents bearing unusual 16S ribosomal RNA (rRNA) and proteinging 

to novel archaeal lineages. The organisms are minor 

in pH 0.5 to 1.5 solutions within the Richmond Mine, 

RNA showed that the fraction less than 0.45 micrometers in 

anisms. Transmission electron microscope images revealed that 

ual folded membrane protrusions and have apparent volumes 

variety of derived from multispecies consortia (4–6) and 

late natural whole environments (7–9) have provided new 

ed by the information about diversity and metabolic 

in reaction potential. However, PCR-based methods have 

nt methods limited ability to detect organisms whose genes 

enes (1–3). are significantly divergent relative to gene 

e fragments sequences in databases, and most cultivationindependent 

genomic sequencing approaches 

are relatively insensitive to organisms that 

occur at low abundance. Consequently, it is 

likely that low-abundance microorganisms distantly 

related to known species will be undetected 

members of natural consortia, even in low 

complexity systems such as acid mine drainage 

(AMD) (10). 

An important way in which microorganisms 

affect geochemical cycles is by accelerating the 

dissolution of minerals. For example, microorganisms 

can derive metabolic energy by oxidizing 

iron released by the dissolution of 

pyrite (FeS2). The ferric iron by-product pro- 

2+ REPORTS 

f Child 


SOM Text 

l Institute of Information. 


tional 

References 

h at the NIH Supporting Online Material 

earch was www.sciencemag.org/cgi/content/full/314/5807/1930/DC1 13 June 2006; accepted 16 November 2006 

Program of Materials and Methods 

10.1126/science.1131262 

idophilic 

and micromolar 

As and Cu (11)] and host active 

led by Community 

microbial communities. Extensive cultivationindependent 

sequence analysis of functional and 

sis 

rRNA genes (11, 12) revealed that biofilms contain 

a significant number of Archaea, but the 

hard I. Webb, diversity reported to date has been limited to the 

order Thermoplasmatales (10). Current models 

for AMD generation thus include only these 

species. 

The genomes of the five dominant members 

of one biofilm community from the “5-way” 

region of the Richmond Mine (fig. S1) were 

largely reconstructed through the assembly of 

76 Mb of shotgun genomic sequence (4). Previously 

unreported is a genome fragment that 

encodes part of the 16S rRNA gene of a novel 

archaeal lineage: Archaeal Richmond Mine 

Acidophilic Nanoorganism (ARMAN-1). Using 

an expanded data set that now comprises more 

than 100 Mb of genomic sequence, we reconstructed 

a contiguous 4.2-kb fragment adjacent 

to this gene. A second 13.2-kb genome fragment 

encoding a 16S rRNA gene from an organism 

that is related to ARMAN-1 (ARMAN-2) was 

reconstructed from 117 Mb of community ge- 

s, University of 

Environmental 

nomic sequence derived from a biofilm from the 

y of California, 

A drift (fig. S1). Within the data sets from each 

icroscopy and 

site, results to date indicate that each ARMAN 

ogy and Para- 

population is near-clonal. 

isbane 4072, 

Comparison of the ARMAN-1 and -2 DNA 

San Francisco, 

fragments revealed some gene rearrangements, 

insertions, and deletions (Fig. 1). Genes present 

Joint Genome 

in both organisms encode putative inorganic 

pyrophosphatases, a transcription regulator, and 

Diego, La Jolla, 

a gene shown to be an arsenate reductase (13). 

essed. E-mail: 

Comparative analysis of these genes with sequences 

in the public databases consistently 

ARMAN-1 

pyrophosphatase insert 

1 

4,226 

31% 69% 85% 90% 97% 

13,236 

3 Judith Flanagan, 2 * 

‡ Jillian F. Banfield 1,2 motes further pyrite dissolution, leading to AMD 

generation. AMD solutions forming underground 

in the Richmond Mine at Iron Mountain, 

California, are warm (30° to 59°C), acidic (pH 

~0.5 to 1.5), metal-rich [submolar Fe 

§ 

ies can be easily overlooked in standard polymerase chain 

d community genomic data obtained without PCR or 

nts bearing unusual 16S ribosomal RNA (rRNA) and proteining 

to novel archaeal lineages. The organisms are minor 

pH 0.5 to 1.5 solutions within the Richmond Mine, 

NA showed that the fraction less than 0.45 micrometers in 

isms. Transmission electron microscope images revealed that 

al folded membrane protrusions and have apparent volumes 

ariety of derived from multispecies consortia (4–6) and 

te natural whole environments (7–9) have provided new 

d by the information about diversity and metabolic 

reaction potential. However, PCR-based methods have 

t methods limited ability to detect organisms whose genes 

nes (1–3). are significantly divergent relative to gene 

fragments sequences in databases, and most cultivationindependent 

genomic sequencing approaches 

are relatively insensitive to organisms that 

occur at low abundance. Consequently, it is 

likely that low-abundance microorganisms distantly 

related to known species will be undetected 

members of natural consortia, even in low 

complexity systems such as acid mine drainage 

(AMD) (10). 

An important way in which microorganisms 

affect geochemical cycles is by accelerating the 

dissolution of minerals. For example, microorganisms 

can derive metabolic energy by oxidizing 

iron released by the dissolution of 

pyrite (FeS2). The ferric iron by-product pro- 

2+ 23. We are grateful to the National Institute of Child 


SOM Text 

Health and Human Development, National Institute of Information. 


Neurological Disorders and Stroke, and National 

References 

Institute of Dental and Craniofacial Research at the NIH Supporting Online Material 

for financial support of this work. This research was www.sciencemag.org/cgi/content/full/314/5807/1930/DC1 REPORTS 13 June 2006 

also supported by the Intramural Research Program of Materials and Methods 

10.1126/scien 

Lineages of Acidophilic 

Archaea Revealed by Community 

Genomic Analysis 

Brett J. Baker, 

and micromolar 

As and Cu (11)] and host active 

microbial communities. Extensive cultivationindependent 

sequence analysis of functional and 

rRNA genes (11, 12) revealed that biofilms contain 

a significant number of Archaea, but the 

diversity reported to date has been limited to the 

order Thermoplasmatales (10). Current models 

for AMD generation thus include only these 

species. 

The genomes of the five dominant members 

of one biofilm community from the “5-way” 

region of the Richmond Mine (fig. S1) were 

largely reconstructed through the assembly of 

76 Mb of shotgun genomic sequence (4). Previously 

unreported is a genome fragment that 

encodes part of the 16S rRNA gene of a novel 

archaeal lineage: Archaeal Richmond Mine 

Acidophilic Nanoorganism (ARMAN-1). Using 

an expanded data set that now comprises more 

than 100 Mb of genomic sequence, we reconstructed 

a contiguous 4.2-kb fragment adjacent 

to this gene. A second 13.2-kb genome fragment 

encoding a 16S rRNA gene from an organism 

that is related to ARMAN-1 (ARMAN-2) was 

reconstructed from 117 Mb of community ge- 

University of 

vironmental 

nomic sequence derived from a biofilm from the 

f California, 

A drift (fig. S1). Within the data sets from each 

roscopy and 

site, results to date indicate that each ARMAN 

y and Para- 

population is near-clonal. 

ane 4072, 

Comparison of the ARMAN-1 and -2 DNA 

n Francisco, 

fragments revealed some gene rearrangements, 

insertions, and deletions (Fig. 1). Genes present 

int Genome 

in both organisms encode putative inorganic 

pyrophosphatases, a transcription regulator, and 

ego, La Jolla, 

a gene shown to be an arsenate reductase (13). 

sed. E-mail: 

Comparative analysis of these genes with sequences 

in the public databases consistently 

ARMAN-1 

pyrophosphatase insert 

1 

4,226 

31% 69% 85% 90% 97% 

1 Gene W. Tyson, 2 Richard I. Webb, 3 Judith Flanagan, 2 * 

Philip Hugenholtz, 2 † Eric E. Allen, 2 ‡ Jillian F. Banfield 1,2 motes furth 

generation 

ground in t 

California, 

~0.5 to 1.5 

cromolar 

microbial 

independen 

rRNA gene 

§ 

tain a sign 

diversity re 

Novel, low-abundance microbial species can be easily overlooked in standard polymerase chain order Ther 

reaction (PCR)–based surveys. We used community genomic data obtained without PCR or for AMD 

cultivation to reconstruct DNA fragments bearing unusual 16S ribosomal RNA (rRNA) and protein- species. 

coding genes from organisms belonging to novel archaeal lineages. The organisms are minor The ge 

components of all biofilms growing in pH 0.5 to 1.5 solutions within the Richmond Mine, of one bio 

California. Probes specific for 16S rRNA showed that the fraction less than 0.45 micrometers in region of 

diameter is dominated by these organisms. Transmission electron microscope images revealed that largely rec 

the cells are pleomorphic with unusual folded membrane protrusions and have apparent volumes 76 Mb of 

of

of decay from NMD (Fig. 1). 

The turnover of normal mRNAs requires 

deadenylation followed by Dcp1pmediated 

decapping and degradation by 

the major 5′-to-3′ exonuclease Xrn1p (6, 

7). NMD is distinguished in that these 

events occur without prior deadenylation 

(8, 9). Nonstop-PGK1 transcripts showed 

rapid decay in strains lacking Xrn1p or 

Dcp1p activity (Fig. 1), providing further 

evidence that they are not subject to 

NMD. This result was surprising since 

both of these factors are also required 

for the turnover of normal mRNAs after 

deadenylation by the major deadenylase 

Ccr4p (10). Nonstop decay was also un- 

Fig. 1. Nonstop-PGK1 transcripts are rapidly 

degraded by a novel mechanism. Half-lives 

(t 1/2 ) of WT-PGK1, PTC(22)-PGK1, nonstop- 

PGK1, and Ter-poly(A)-PGK1 transcripts 

are shown (27). Half-lives were performed 

in a wild-type yeast strain ( WT ), strains 

deleted for Upf1p (upf1Δ), Xrn1p (xrn1Δ), or 

Ccr4p (ccr4Δ), a strain lacking activity of the 

decapping enzyme Dcp1p (dcp1-2), and a WT 

yeast strain treated for 1 hour with 100 μg/ml 

of cycloheximide (+CHX ). 

18 

altered in a strain lacking Ccr4p (Fig. 1). 

Therefore, degradation of nonstop-PGK1 

transcripts requires none of the factors involved 

in the pathway for degradation of 

wild-type or nonsense mRNAs. 

Additional experiments were performed 

to assess the role of translation in nonstop 

decay. Treatment with cycloheximide 

(CHX) or depletion of charged tRNAs in 

yeast harboring the conditional cca1-1 allele 

(11) grown at the nonpermissive temperature 

substantially increased the stability 

of nonstop-PGK1 transcripts (Fig. 1) 

(12). It has been shown that translation 

into the 3′ UTR of selected transcripts can 

displace bound trans-factors that are posi- 

tive determinants of message stability (13, 

14). To determine whether this mechanism 

is relevant to nonstop decay, we examined 

into the 3′ UTR of selected transcripts can

major substantial (threefold) stabilization (Fig. 1). 

decay Unlike nonstop-PGK1 transcripts, the half- 

Ccr4p life of Ter-poly(A)-PGK1 was increased in 

onstop- the absence of Xrn1p (Fig. 1), indicating 

hefac- that tive this determinants transcript is of degraded message stability by the path- (13, 

egrada-way 14). for To determine normal mRNA whether and this not mechanism by the 

As. nonstop pathway. These data suggest that 

is relevant to nonstop decay, we examined 

formed the instability of nonstop-PGK1 transcripts 

onstop cannot 

the performance 

fully be explained 

of transcript 

by ribosomal 

Ter-poly(A)dis 

ximide placement PGK1 which of factors contains bound a termination to the 3� codon UTR. 

NAs in inserted There is one ancodon emerging upstream viewof that the the site 5� of 

1-1 al- and polyadenylate 3� ends of mRNAs [poly(A)] interact addition to form (15) in a 

etem- closed transcript loopnonstop-PGK1. and that this conformation The addition is of 

stabili- required a termination for efficient codon translation at the 3′ end initiation of the 

Fig. 1) and normal mRNA stability. Participants in 

3′ UTR of a nonstop transcript resulted in 

on into the interaction include components of the 

an dis- translation 

substantial 

initiation 

(threefold) 

complex 

stabilization 

eIF4F as well 

(Fig. 

ositive as1). Pab1p. Unlike The nonstop-PGK1 absence of transcripts, a termination the 

3, 14). codon half-life in nonstop-PGK1 of Ter-poly(A)-PGK1 is predicted to was allow in- 

ism is the creased ribosome in the to continue absence translating of Xrn1p (Fig. through 1), 

ned the the indicating 3� UTR and that poly(A) this transcript tail, potentially is degraded reoly(A)sulting 

by the inpathway displacement for normal of Pab1p, mRNA disruption and not 

codon of normal mRNP structure, and consequently 

by the nonstop pathway. These data suggest 

that the instability of nonstop-PGK1 

transcripts cannot fully be explained by 

ribosomal displacement of factors bound 

to the 3′ UTR. 

There is an emerging view that the 

5′and 3′ ends of mRNAs interact to form 

a closed loop and that this conformation is 

required for efficient translation initiation 

and normal mRNA stability. Participants 

in the interaction include components of 

the translation initiation complex eIF4F as 

well as Pab1p. The absence of a termination 

codon in nonstop-PGK1 is predicted 

to allow the ribosome to continue translating 

through the 3′ UTR and poly(A) 

tail, potentially resulting in displacement 

of Pab1p, disruption of normal mRNP 

structure, and consequently accelerated 

degradation of the transcript. However, 

while mRNAs in pab1Δ strains do undergo 

accelerated decay, the rapid turnover is 

a result of premature decapping and can 

be suppressed by mutations in XRN1 (16), 

allowing distinction from nonstop decay. 

rapidly Lability of nonstop transcripts might also 

alf-lives be a consequence of ribosomal stalling at 

onstop- the 3′ end of the transcript. It is tempting 

pts are 

d in a to speculate that 3′-to-5′ exonucleolytic 

eted for activity underlies the decapping- and 5′- 

Ccr4p to-3′ exonuclease–independent, transla- 

decap- 

T yeast tion-dependent accelerated decay of non- 

ml of cycloheximide stop transcripts. (�CHX ). An appealing candidate 

lowing distinction from nonstop decay. Lability 

of nonstop transcripts might also be a 

consequence of ribosomal stalling at the 3� 

end of the transcript. It is tempting to speculate 

is the that exosome, 3�-to-5�a collection exonucleolytic of proteins activity 

underlies with 3′-to-5′ the exoribonuclease decapping- and 5�-to-3� activity that exonuclease–independent, 

translation-dependent 

functions in the processing of 5.8S RNA, 

accelerated decay of nonstop transcripts. An 

appealing 

rRNA, small 

candidate 

nucleolar 

is the 

RNA 

exosome, 

(snoRNA), 

a collection 

small of proteins nuclear with RNA 3�-to-5� (snRNA), exoribonuclease and other 

activity transcripts. thatData functions presented in thein processing van Hoof et of 

5.8S al. (17) RNA, validate rRNA, this prediction. small nucleolar RNA 

Fig. 2. Nonstop decay is conserved in mammalian 

cells. (A) Steady-state abundance of �-glucuronidase 

transcripts derived from wild type– 

( WT ), nonsense-, nonstop-, and Ter-poly(A)- 

�gluc minigene constructs after transient 

transfection into HeLa cells (28). Zeocin (zeo) 

transcripts served as a control for loading and 

transfection efficiency. For each construct, the 

�gluc/Zeo transcript ratio is expressed relative 

to that observed for WT-�gluc. (B) Steadystate 

abundance of WT-�gluc, nonstop-�gluc, 

and nonsense-�gluc transcripts were determined 

in the nuclear (N) and cytoplasmic (C) 

fractions of transfected cells (28). The ratio of 

normalized nonstop- or nonsense-�gluc transcript 

levels to WT-�gluc transcript levels are 

shown for each cellular compartment. Reported 

results are the average of three independent 

experiments. 

www.sciencemag.org SCIENCE VOL 295 22 MARCH 2002 19 2259 

Downloaded from 

www.sciencemag.org on June 29, 2007

In order to determine whether nonstop 

decay functions in mammals, we assessed 

the performance of transcripts derived 

from β-glucuronidase (βgluc) minigene 

constructs containing exons 1, 10, 11, and 

12 separated by introns derived from the 

endogenous gene (18). The abundance of 

transcripts derived from a nonstop-βgluc 

version of this minigene was significantly 

reduced relative to WT-βgluc, suggesting 

that a termination codon is also essential 

for normal mRNA stability in mammalian 

cells (Fig. 2A). Addition of a stop 

codon one codon upstream from the site 

of poly(A) addition in nonstop-βgluc 

[Ter-poly(A)-βgluc (15)] increased the 

abundance of this transcript to near–wildtype 

levels (Fig. 2A). Thus, all functional 

characteristics of nonstop transcripts in 

yeast appear to be relevant to mammalian 

systems. 

Both the abundance and stability of 

most nonsense transcripts, including nonsense-βgluc 

(Fig. 2B), are reduced in the 

nuclear fraction of mammalian cells (19, 

20). This has been interpreted to suggest 

that translation and NMD initiate while 

the mRNA is still associated with (if 

not within) the nuclear compartment. If 

translation is initiated on nucleus-associated 

transcripts, so might nonstop decay. 

Subcellular fractionation studies localized 

mammalian nonstop decay to the cytoplasmic 

compartment, providing further 

distinction from NMD (Fig. 2B). 

The conservation of nonstop decay 

in yeast and mammals suggests that the 

pathway serves an important biologic 

role. There are many potential physiologic 

sources of nonstop transcripts that warrant 

consideration. Mutations in bona fide termination 

codons would not routinely initiate 

nonstop decay due to the frequent occurrence 

of in-frame termination codons in 

the 3′ UTR and could not plausibly provide 

the evolutionary pressure for maintenance 

of nonstop decay. In contrast, alternative 

use of 3′-end processing signals embedded 

in coding sequence has been documented 

20 

in many genes including CBP1 in yeast 

and the growth hormone receptor (GHR) 

gene in fowl (21–23). Moreover, a computer 

search of the human mRNA and S. 

cerevisiae open reading frame (ORF) databases 

revealed many additional genes that 

contain a strict consensus sequence for 3′ 

end cleavage and polyadenylation within 

their coding region (Fig. 3A). Utilization 

of these premature signals would direct 

formation of truncated transcripts that 

might be substrates for the nonstop decay 

pathway. Indeed, analysis of 3425 random 

yeast cDNA clones sequenced from the 3′ 

end (i.e., 3′ ESTs) revealed that 40 showed 

apparent premature polyadenylation within 

the coding region (24). 

The truncated GHR transcript in fowl 

is apparently translated, as evidenced by 

its association with polysomes (22). As 

predicted for a substrate for nonstop decay, 

the ratio of truncated–to–full-length 

GHR transcripts was dramatically increased 

after treatment of cultured chicken 

hepatocellular carcinoma cells with the 

translational inhibitor emetine (Fig. 3B). 

Other truncated transcripts, both bigger 

and smaller than the predicted 0.7-kb nonstop 

transcript, did not show a dramatic 

increase in steady-state abundance upon 

translational arrest suggesting that they 

are derived from other mRNA processing 

events, perhaps alternative splicing (Fig. 

3B) (12). To directly test whether the nonstop 

mRNA pathway degrades prematurely 

polyadenylated mRNAs, we analyzed 

CBP1 transcripts in yeast. The CBP1 gene 

produces a 2.2-kb full-length mRNA and 

a 1.2-kb species with premature 3′ end 

processing and polyadenylation within the 

coding region (25). As expected from the 

observation that nonstop decay requires 

the exosome in yeast (17), deletion of the 

gene encoding the exosome component 

Ski7p stabilized the prematurely polyadenylated 

mRNA but had no effect on the 

stability of the full-length mRNA (Fig. 

3C). These data indicate that physiologic 

transcripts arising from premature poly-

Fig. 3. Many physiologic transcripts are 

substrates for nonstop decay. (A) 239 

human mRNAs contain a polyadenylation 

signal consisting of either of the two most 

common words for the positioning, cleavage, 

and downstream elements (29), separated by 

optimal distances (30), where the cleavage 

site occurs between the start and stop 

codons as determined from the cDNA coding 

sequence (CDS). Similarly 52 S. cerevisiae ORFs 

contained the yeast polyadenylation signal 

consisting of optimal upstream, positioning, 

and cleavage signals followed by any of nine 

common “U-rich” signals (24) with optimal 

spacing of the elements. Computer search 

was performed using the human mRNA 

database and the S. cerevisiae ORF database 

and an analysis program written in PERL that 

is available upon request. (B) Abundance of 

nonstop and full-length cGHR transcripts in 

chicken hepatocellular carcinoma cells (CRL- 

2117 purchased from the American Type 

Culture Collection, Manassas, VA) treated 

with 100 μg/ml of emetine for 6 hours. Northern analysis was performed as described (22). The 

relative ratio of each transcript in untreated and treated cells is reported. (C) Half-lives (t 1/2 ) of 

the full-length and nonstop CBP1 mRNAs measured in wild-type–(WT) and SKI7-deleted (ski7Δ) 

yeast strains carrying plasmid pG::-26 (25, 31). (D) Half-lives of WT-PGK1 and Ter-poly(A)-PGK1 

transcripts in wild-type–(WT ) and SKI7-deleted (ski7Δ) yeast strains treated with the indicated 

dose of paromomycin (mg/ml) (PM) for 20 hours. For cycloheximide (CHX ) experiments, yeast 

were treated identically except 100 μg/ml of CHX was added during the last hour before half-life 

determination. Half-lives were determined as described (27). 

21

adenylation are subject to nonstop mRNA 

decay. The cytoplasmic localization of 

ski7p is consistent with our observation 

that nonstop decay occurs within the cytoplasm 

(Fig. 2B). 

Any event that diminishes translational 

fidelity and promotes readthrough of termination 

codons could plausibly result in 

the generation of substrates for nonstop 

decay. In view of recent attempts to treat 

genetic disorders resulting from PTCs 

with long-term and high-dose aminoglycoside 

regimens, this may achieve medical 

significance. As a proof-of-concept experiment, 

we examined the performance of 

transcripts with one [Ter-poly(A)-PGK1] 

or multiple (WT-PGK1) in-frame termination 

codons (including those in the 3′ 

prematurely polyadenylated mRNA but had 

no UTR) effect in yeast on thestrains stability after of treatment the full-length with 

mRNA paromomycin, (Fig. 3C). which Theseinduces data indicate ribosomal that 

physiologic readthrough. transcripts Remarkably, arising both from transcripts premature 

showed polyadenylation a dose-dependent are subject decrease to nonstop in sta- 

mRNA bility that decay. could Thebe cytoplasmic reversed by localization inhibiting 

of ski7p is consistent with our observation 

translational elongation with CHX or pre- 

that nonstop decay occurs within the cytoplasmvented 

(Fig. by deletion 2B). of the gene encoding 

Ski7p Any(Fig. event3D). that These diminishes data suggest translational that 

fidelity nonstop and decay promotes can limit readthrough the efficiency of termi- of 

nation therapeutic codons strategies could plausibly aimed at result enhancing in the 

generation nonsense suppression of substratesand for nonstop might contrib- decay. 

In view of recent attempts to treat genetic 

ute to the toxicity associated with amino- 

disorders resulting from PTCs with longtermglycoside 

and high-dose therapy. aminoglycoside regimens, 

The this degradation may achieve of nonstop medical transcripts significance. 

may be As regulated. a proof-of-concept The relative experiment, expression 

we level examined of truncated the performance GHR transcripts of transcripts com- 

with pared one to full-length [Ter-poly(A)-PGK1] transcripts or varies multiple in a 

(WT-PGK1) in-frame termination codons 

tissue-, gender-, and developmental stage- 

(including those in the 3� UTR) in yeast 

strains specific after manner treatment (22) and with the paromomycin, 

relative abun- 

which dance of induces truncated ribosomal CBP1 readthrough. transcripts varies Remarkably, 

with growth both condition transcripts (21). showed Data present- a dosedependented 

here warrant decrease the hypothesis in stability that thatregula could 

be tion reversed may occur by inhibiting at the level translational of nonstop elontrangation with CHX or prevented by deletion 

script stability rather than production. The 

of the gene encoding Ski7p (Fig. 3D). 

These conservation data suggest of the that GHR nonstop coding decay sequence can 

limit 3′- end theprocessing efficiency signal of therapeutic throughout strategies avian 

aimed phylogeny at enhancing and conservation nonsenseof suppression premature 

and polyadenylation might contribute of the to yeast the toxicity RNA14 associtranatedscript with and aminoglycoside its fruitfly homolog therapy. su(f ) sup- 

The degradation of nonstop transcripts may 

port speculation that nonstop transcripts 

be regulated. The relative expression level of 

truncated or derived GHR protein transcripts products compared serve essential to fulllength 

transcripts varies in a tissue-, gender-, 

and 22 developmental stage-specific manner (22) 

and the relative abundance of truncated CBP1 

developmental and/or homeostatic functions 

that are regulated by nonstop decay 

(21, 26). 

Many processes contribute to the precise 

control of gene expression including 

transcriptional and translational control 

mechanisms. In recent years, mRNA stability 

has emerged as a major determinant 

of both the magnitude and fidelity of gene 

expression. Perhaps the most striking and 

comprehensively studied example is the 

accelerated decay of transcripts harboring 

PTCs by the NMD pathway. Nonstop decay 

now serves as an additional example 

of the critical role that translation plays 

in monitoring the fidelity of gene expression, 

the stability R E Pof Oaberrant R T S 

or atypical 

transcripts, and hence the abundance of 

gene expression, the stability of aberrant or 

atypical 

truncated 

transcripts, 

proteins. 

and hence the abundance 

of truncated proteins. 


1. A. W. Karzai, E. D. Roche, R. T. Sauer, Nature Struct. 

Biol. 7, 449 (2000). 

2. K. C. Keiler, P. R. Waller, R. T. Sauer, Science 271, 990 

(1996). 

3. The nonstop-PGK1 construct was generated from 

WT-PGK1 [pRP469 (8)] by creating three point mutations, 

using site-directed mutagenesis (Quik- 

Change Site-Directed Mutagenesis Kit, Stratagene) 

which eliminated the bona fide termination codon 

and all in-frame termination codons in the 3� UTR. 

The Ter-poly(A)-PGK1 construct was created from 

the nonstop-PGK1 construct, using site-directed mutagenesis 

(Quik-Change Site-Directed Mutagenesis 

Kit, Stratagene). All changes were confirmed by sequencing. 

Primer sequences are available upon request. 

The PTC(22)-PGK1 construct is described in (8) 

(pRP609). 

4. P. Leeds, J. M. Wood, B. S. Lee, M. R. Culbertson, Mol. 

Cell. Biol. 12, 2165 (1992). 

5. Y. Cui, K. W. Hagan, S. Zhang, S. W. Peltz, Genes Dev. 

9, 423 (1995). 

6. D. Muhlrad, C. J. Decker, R. Parker, Genes Dev. 8, 855 

(1994). 

7. C. A. Beelman et al., Nature 382, 642 (1996). 

8. D. Muhlrad, R. Parker, Nature 370, 578 (1994). 

9. G. Caponigro, R. Parker, Microbiol. Rev. 60, 233 

(1996). 

10. M. Tucker et al., Cell 104, 377 (2001). 

11. C. L. Wolfe, A. K. Hopper, N. C. Martin, J. Biol. Chem. 

271, 4679 (1996). 

12. P. A. Frischmeyer, H. C. Dietz, data not shown. 

13. I. M. Weiss, S. A. Liebhaber, Mol. Cell. Biol. 14, 8123 

(1994). 

14. X. Wang, M. Kiledjian, I. M. Weiss, S. A. Liebhaber, 

Mol. Cell. Biol. 15, 1769 (1995). 

15. We performed 3� rapid amplification of cDNA ends 

(RACE) using the Marathon cDNA Amplification Kit 

(Clontech) and the Advantage 2 Polymerase Mix 

(Clontech) according to the manufacturer’s instructions. 

Poly(A) RNA isolated from cells (Oligotex 

mRNA midi kit, Qiagen) transiently transfected with 

WT-�gluc or from a wild-type yeast strain carrying a 

plasmid expressing WT-PGK1 was used as template. 

A single RACE product was generated, cloned ( TOPO 

TA 2.1 kit, Invitrogen), and sequenced. 

16. G. Caponigro, R. Parker, Genes Dev. 9, 2421 (1995). 

21. K. A. Spa 

4676 (19 

22. E. R. Ol 

Endocrin 

23. C. Hilge 

J. Virol. 

24. J. H. Gr 

Nucleic 

25. K. A. Spa 

Biol. 17, 

26. A. Audib 

U.S.A. 9 

27. All PGK1 

the GAL1 

were gro 

media–u 

scription 

SC-ura m 

off) to a 

removed 

10 �g of 

od (4) w 

dehyde g 

Screen P 

nucleotid 

cally det 

transcrip 

the perce 

log plot. 

for the c 

formed a 

1 hour] e 

28. HeLa ce 

bovine s 

confluen 

RPMI 16 

dithiothr 

and cells 

Electrop 

a capaci 

kit, Qiag 

tion, wa 

gel, tran 

with a po 

prising e 

had bee 

(Random 

heim). T 

with boi 

labeled 

quantita 

subcellu 

HeLa cel

ty associ- 15. We performed 3� rapid amplification of cDNA ends 

. 


(Clontech) and the Advantage 2 Polymerase Mix 

ripts may (Clontech) according to the manufacturer’s instruc- 

n level of tions. Poly(A) RNA isolated from cells (Oligotex 

d to full- mRNA midi kit, Qiagen) transiently transfected with 

WT-�gluc or from a wild-type yeast strain carrying a 

, gender-, plasmid expressing WT-PGK1 was used as template. 

nner (22) A single RACE product was generated, cloned ( TOPO 

ted CBP1 TA 2.1 kit, Invitrogen), and sequenced. 

16. G. Caponigro, R. Parker, Genes Dev. 9, 2421 (1995). 

ition (21). 17. A. van Hoof, P. A. Frischmeyer, H. C. Dietz, R. Parker, 

thesis that Science 295, 2262 (2002). 

f nonstop 18. To create the WT-�gluc minigene construct, mouse 

genomic DNA was used as template for amplifying 

ction. The 

three different portions of the �-glucuronidase gene: 

uence 3�- from nucleotide 2313 to 3030 encompassing exon 1 

nphylog- and part of intron 1 (including the splice donor selyadenylquence), 

from nucleotide 11285 to 13555 encompassing 

part of intron 9 (including the splice acceptor sept 

and its quence ), exon 10, and part of intron 10 (including the 

lation that splice donor sequence), and from 13556 to 15813 

products which included the remainder of intron 10, exon 11, 

intron 11, and exon 12. All three fragments were TA 

r homeo- cloned ( Topo TA 2.1 kit, Invitrogen), sequenced, and 

y nonstop then cloned into pZeoSV2 (Invitrogen). A single basepair 

deletion corresponding to the gus 

e precise 

ing tranlmechability 

has 

f both the 

pression. 

mprehenerateddes 

by the 

serves as 

l role that 

fidelity of 

23 

mps tagenesis (Quik-Change Site-Directed Mutagenesis 

gel, transferred to a nylon membrane, and hybridized 

ide regi- Kit, Stratagene). All changes were confirmed by se- 

with a polymerase chain reaction (PCR) product comsignifiquencing. 

Primer sequences are available upon reprising 

exons 10 through 12 of the �-gluc gene that 

quest. The PTC(22)-PGK1 construct is described in (8) 

eriment, 

had been radioactively labeled by nick translation 

(pRP609). 

(Random Primed DNA Labeling Kit, Boehringer Mann- 

anscripts 4. P. Leeds, J. M. Wood, B. S. Lee, M. R. Culbertson, Mol. 

heim). The membrane was subsequently stripped 

multiple Cell. Biol. 12, 2165 (1992). 

with boiling 0.5% SDS and probed with radioactively 

5. Y. Cui, K. W. Hagan, S. Zhang, S. W. Peltz, Genes Dev. 

codons 

labeled zeocin cDNA. Northern blot results were 

9, 423 (1995). 

quantitated using an Instant Imager (Packard). For 

in yeast 6. D. Muhlrad, C. J. Decker, R. Parker, Genes Dev. 8, 855 subcellular fractionation, transiently transfected 

omycin, (1994). 

HeLa cells were trypsinized, washed in cold 1� phos- 


ugh. Rephate 

buffered saline, and pelleted by centrifuging at 


2040g for 5 min at 4°C. Cells were then resuspended 

adose- 9. G. Caponigro, R. Parker, Microbiol. Rev. 60, 233 in 200 �l of 140 mM NaCl, 1.5 mM MgCl2 , 10 mM 

at could (1996). 

Tris-HCl (pH 8.6), 0.5% NP-40, and 1 mM DTT, 

nal elon- 10. M. Tucker et al., Cell 104, 377 (2001). 

vortexed, and incubated on ice for 5 min. Nuclei were 

11. C. L. Wolfe, A. K. Hopper, N. C. Martin, J. Biol. Chem. 

deletion 

pelleted by centrifuging at 12,000g for 45 s at 4°C 

271, 4679 (1996). 

and subsequently separated from the aqueous (cyto- 

ig. 3D). 12. P. A. Frischmeyer, H. C. Dietz, data not shown. 

plasmic) phase. 

ecay can 13. I. M. Weiss, S. A. Liebhaber, Mol. Cell. Biol. 14, 8123 29. J. H. Graber, C. R. Cantor, S. C. Mohr, T. F. Smith, Proc. 

(1994). 

trategies 


14. X. Wang, M. Kiledjian, I. M. Weiss, S. A. Liebhaber, 30. F. Chen, C. C. MacDonald, J. Wilusz, Nucleic Acids Res. 

pression Mol. Cell. Biol. 15, 1769 (1995). 

23, 2614 (1995). 

y associ- 15. We performed 3� rapid amplification of cDNA ends 31. Plasmid pG-26 (25) containing the CBP1 gene un- 

. 

(RACE) using the Marathon cDNA Amplification Kit der the control of the GAL promoter and the LEU2 

(Clontech) and the Advantage 2 Polymerase Mix selectable marker was introduced into wild-type 

ripts may (Clontech) according to the manufacturer’sallele instruc- was and SKI7-deleted strains. Yeast were grown in me- 

level of tions. generated Poly(A) by site-directed RNA isolated mutagenesis from cells (Quik-Change (Oligotex dia lacking leucine and containing 2% galactose at 

d to full- mRNA Site-Directed midi kit, Mutagenesis Qiagen) transiently Kit, Stratagene) transfected of thewith WT- 30°C to an OD600 of 0.5. Cells were pelleted and 

WT-�gluc �gluc construct or fromto a create wild-type nonsense-�gluc. yeast strain carrying Nonstop- a resuspended in media lacking a carbon source. 

, gender-, plasmid �gluc was expressing generatedWT-PGK1 from WT-�gluc was used by two as template. rounds of Glucose was then added to a final concentration of 

nner (22) Asite-directed single RACEmutagenesis product was(Quik-Change generated, cloned Site-Directed ( TOPO 2% at time 0 and time points were taken. RNA was 

ted CBP1 TA Mutagenesis 2.1 kit, Invitrogen), Kit, Stratagene), and sequenced. which removed the bona extracted and analyzed by northern blotting with a 

16. G. fide Caponigro, termination R. Parker, codon Genes and allDev. in-frame 9, 2421 termination (1995). 

tion (21). 

labeled oligo probe oRP1067 (5�-CTCGGTCCTG- 

17. A. codons van Hoof, in theP. 3�A. UTR. Frischmeyer, Ter-poly(A)-�gluc H. C. Dietz, was R. generated Parker, TACCGAACGAGACGAGG-3�). 

thesis that Science from nonstop-�gluc 295, 2262 (2002). via a single point mutation, which 32. We thank N. Martin, A. Atkin, C. Dieckmann, and M. 

f nonstop 18. To created create a stop the one WT-�gluc codon minigene upstream of construct, the poly(A) mouse tail. Sands for the provision of valuable reagents. This 

genomic All changes DNA were was confirmed used as by template sequencing. for amplifying 

tion. The 

Primer se- work was supported by a grant from NIH (GM55239) 

three quences different are available portions upon of the request. �-glucuronidase gene: (H.C.D.), the Howard Hughes Medical Institute 

uence 3�- 19. from P. Belgrader, nucleotide J. Cheng, 2313 to X. 3030 Zhou, encompassing L. S. Stephenson, exonL. 1E. 

(H.C.D, R.P.), and the Medical Scientist Training Pronphylog- 

and Maquat, part Mol. of intron Cell. 1 Biol. (including 14, 8219 the(1994). splice donor segram (P.A.F.). 

lyadenyl 20. quence), O. Kessler, fromL. nucleotide A. Chasin, 11285 Mol. toCell. 13555 Biol. encompass- 16, 4426 

ing (1996). part of intron 9 (including the splice acceptor se- 22 October 2001; accepted 31 January 2002 

t and its 

berrant or 21. quence K. A. Sparks, ), exonC. 10, L. and Dieckmann, part of intron Nucleic 10 Acids (including Res. the 26, 

ation that splice 4676 (1998). donor sequence), and from 13556 to 15813 

bundance 

products www.sciencemag.org 22. which E. R. Oldham, includedSCIENCE the B. Bingham, remainder VOL W. of intron R. 295 Baumbach, 10, 22 exon MARCH Mol. 11, 2002 2261 

intron Endocrinol. 11, and 7, exon 1379 12. (1993). All three fragments were TA 

r homeo- 

23. cloned C. Hilger, ( Topo I. Velhagen, TA 2.1 kit, H. Invitrogen), Zentgraf, sequenced, C. H. Schroder, and 

nonstop then J. Virol. cloned 65, into 4284 pZeoSV2 (1991). (Invitrogen). A single base- 

24. pair J. H. deletion Graber, corresponding C. R. Cantor, to S. the C. Mohr, gus T. F. Smith, 

ature Struct. 

e precise Nucleic Acids Res. 27, 888 (1999). 

25. K. A. Sparks, S. A. Mayer, C. L. Dieckmann, Mol. Cell. 

ing ce 271, tran- 990 Biol. 17, 4199 (1997). 

lmecha- 26. A. Audibert, M. Simonelig, Proc. Natl. Acad. Sci. 

erated from 

bility has U.S.A. 95, 14302 (1998). 

e point mu- 27. All PGK1 minigene constructs are under the control of 

esis both(Quik- the 

the GAL1 upstream activation sequence (UAS). Cultures 

pression. Stratagene) were grown to mid-log phase in synthetic complete 

ation codon mprehen- media–uracil (SC-ura) containing 2% galactose (tran- 

the 3� UTR. scription on). Cells were pelleted and resuspended in 

reated rated from de- 

SC-ura media. Glucose was then added (transcription 

sirected by the mu- off) to a final concentration of 2% and aliquots were 

Mutagenesis serves as removed at the indicated time points. Approximately 

rmed by se- 10 �g of total RNA isolated with the hot phenol methlerole 

upon that reod (4) was electrophoresed on a 1.2% agarose formalscribed 

idelityin of (8) dehyde gel, transferred to a nylon membrane (Gene- 

Screen Plus, NEN), and hybridized with a 23 bp oligo- 

ertson, Mol. nucleotide end-labeled radioactive probe that specifically 

detects the polyG tract in the 3� UTR of the PGK1 

, Genes Dev. transcripts (8). Half-lives were determined by plotting 

the percent of mRNA remaining versus time on a semi- 

Dev. 8, 855 log plot. All half-lives were determined at 25°C except 

for the ccr4� (performed at 30°C) and dcp1-2 [per- 

996). 

(1994). 

v. 60, 233 

formed after shift to the nonpermissive temp (37°C) for 

1 hour] experiments. 

28. HeLa cells [maintained in DMEM with 10% fetal 

bovine serum (FBS)] were grown to �80 to 90% 

confluency, trypsinized, and resuspended in 300 �l of 

. Biol. Chem. RPMI 1640 without FBS, 10 mM glucose and 0.1 mM 

dithiothreitol (DTT). Plasmid DNA (10 �g) was added 

shown. 

ol. 14, 8123 

and cells were electroporated (Bio-Rad, Gene Pulser II 

Electroporation System) at a voltage of 0.300 kV and 

. Liebhaber, 

a capacitance of 500 �F. Poly(A) RNA (Oligotex midi 

kit, Qiagen, 2 �g), isolated 36 hours after transfection, 

was separated on a 1.6% agarose formaldehyde 

mps t genetic The Ter-poly(A)-PGK1 construct was created from removed at the indicated time points. Approximately 

the nonstop-PGK1 construct, using site-directed mu- 10 �g of total RNA isolated with the hot phenol meth- 

ith longtagenesis (Quik-Change Site-Directed Mutagenesis od (4) was electrophoresed on a 1.2% agarose formalideregi- 

Kit, Stratagene). All changes were confirmed by sedehyde gel, transferred to a nylon membrane (Genelsignifiquencing. 

Primer sequences are available upon re- Screen Plus, NEN), and hybridized with a 23 bp oligoquest. 

The PTC(22)-PGK1 construct is described in (8) nucleotide end-labeled radioactive probe that specifieriment, 

(pRP609). 

cally detects the polyG tract in the 3� UTR of the PGK1 

anscripts 4. P. Leeds, J. M. Wood, B. S. Lee, M. R. Culbertson, Mol. transcripts (8). Half-lives were determined by plotting 

multiple Cell. Biol. 12, 2165 (1992). 

the percent of mRNA remaining versus time on a semi- 

5. Y. Cui, K. W. Hagan, S. Zhang, S. W. Peltz, Genes Dev. log plot. All half-lives were determined at 25°C except 

codons 9, 423 (1995). 

for the ccr4� (performed at 30°C) and dcp1-2 [per- 

in yeast 6. D. Muhlrad, C. J. Decker, R. Parker, Genes Dev. 8, 855 formed after shift to the nonpermissive temp (37°C) for 

omycin, (1994). 



ugh. Re- 




adose- 9. G. Caponigro, R. Parker, Microbiol. Rev. 60, 233 confluency, trypsinized, and resuspended in 300 �l of 

at could (1996). 

RPMI 1640 without FBS, 10 mM glucose and 0.1 mM 

nal elon- 10. M. Tucker et al., Cell 104, 377 (2001). 


11. C. L. Wolfe, A. K. Hopper, N. C. Martin, J. Biol. Chem. and cells were electroporated (Bio-Rad, Gene Pulser II 

deletion 271, 4679 (1996). 


ig. 3D). 12. P. A. Frischmeyer, H. C. Dietz, data not shown. 


ecay can 13. I. M. Weiss, S. A. Liebhaber, Mol. Cell. Biol. 14, 8123 kit, Qiagen, 2 �g), isolated 36 hours after transfec- 

(1994). 

tion, was separated on a 1.6% agarose formaldehyde 

trategies 

14. X. Wang, M. Kiledjian, I. M. Weiss, S. A. Liebhaber, gel, transferred to a nylon membrane, and hybridized 

pression Mol. Cell. Biol. 15, 1769 (1995). 

with a polymerase chain reaction (PCR) product com- 

y associ- 15. We performed 3� rapid amplification of cDNA ends prising exons 10 through 12 of the �-gluc gene that 


. 


(Clontech) and the Advantage 2 Polymerase Mix (Random Primed DNA Labeling Kit, Boehringer Mann- 

ripts may (Clontech) according to the manufacturer’s instrucheim). The membrane was subsequently stripped 

level of tions. Poly(A) RNA isolated from cells (Oligotex with boiling 0.5% SDS and probed with radioactively 


d to full- mRNA midi kit, Qiagen) transiently transfected with 

WT-�gluc or from a wild-type yeast strain carrying a quantitated using an Instant Imager (Packard). For 

, gender-, plasmid expressing WT-PGK1 was used as template. subcellular fractionation, transiently transfected 

nner (22) A single RACE product was generated, cloned ( TOPO HeLa cells were trypsinized, washed in cold 1� phos- 

ted CBP1 TA 2.1 kit, Invitrogen), and sequenced. 

phate buffered saline, and pelleted by centrifuging at 

16. G. Caponigro, R. Parker, Genes Dev. 9, 2421 (1995). 2040g for 5 min at 4°C. Cells were then resuspended 

tion (21). 17. A. van Hoof, P. A. Frischmeyer, H. C. Dietz, R. Parker, in 200 �l of 140 mM NaCl, 1.5 mM MgCl2 , 10 mM 

thesis that Science 295, 2262 (2002). 


f nonstop 18. To create the WT-�gluc minigene construct, mouse vortexed, and incubated on ice for 5 min. Nuclei were 

genomic DNA was used as template for amplifying pelleted by centrifuging at 12,000g for 45 s at 4°C 

tion. The 

three different portions of the �-glucuronidase gene: and subsequently separated from the aqueous (cyto- 

uence 3�- from nucleotide 2313 to 3030 encompassing exon 1 plasmic) phase. 

nphylog- and part of intron 1 (including the splice donor se- 29. J. H. Graber, C. R. Cantor, S. C. Mohr, T. F. Smith, Proc. 


lyadenylquence), from nucleotide 11285 to 13555 encompassing 

part of intron 9 (including the splice acceptor se- 30. F. Chen, C. C. MacDonald, J. Wilusz, Nucleic Acids Res. 

t and its quence ), exon 10, and part of intron 10 (including the 23, 2614 (1995). 

lation that splice donor sequence), and from 13556 to 15813 31. Plasmid pG-26 (25) containing the CBP1 gene under 

the control of the GAL promoter and the LEU2 

products which included the remainder of intron 10, exon 11, 

intron 11, and exon 12. All three fragments were TA selectable marker was introduced into wild-type 

r homeo- cloned ( Topo TA 2.1 kit, Invitrogen), sequenced, allele was and and SKI7-deleted strains. Yeast were grown in menonstop 

generated then cloned byinto site-directed pZeoSV2 mutagenesis (Invitrogen). A(Quik-Change single basedia lacking leucine and containing 2% galactose at 

Site-Directed pair deletion Mutagenesis corresponding Kit, toStratagene) the gus of the WT- 30°C to an OD600 of 0.5. Cells were pelleted and 

�gluc construct to create nonsense-�gluc. Nonstop- resuspended in media lacking a carbon source. 

e precise �gluc was generated from WT-�gluc by two rounds of Glucose was then added to a final concentration of 

ing tran- site-directed mutagenesis (Quik-Change Site-Directed 2% at time 0 and time points were taken. RNA was 

Mutagenesis Kit, Stratagene), which removed the bona 

lmecha- extracted and analyzed by northern blotting with a 

fide termination codon and all in-frame termination labeled oligo probe oRP1067 (5�-CTCGGTCCTGbility 

has codons in the 3� UTR. Ter-poly(A)-�gluc was generated TACCGAACGAGACGAGG-3�). 

both the from nonstop-�gluc via a single point mutation, which 32. We thank N. Martin, A. Atkin, C. Dieckmann, and M. 

pression. created a stop one codon upstream of the poly(A) tail. Sands for the provision of valuable reagents. This 

All changes were confirmed by sequencing. Primer se- work was supported by a grant from NIH (GM55239) 

mprehenquences are available upon request. 

(H.C.D.), the Howard Hughes Medical Institute 

ratedde- 19. P. Belgrader, J. Cheng, X. Zhou, L. S. Stephenson, L. E. (H.C.D, R.P.), and the Medical Scientist Training Pros 

by the Maquat, Mol. Cell. Biol. 14, 8219 (1994). 

gram (P.A.F.). 

20. O. Kessler, L. A. Chasin, Mol. Cell. Biol. 16, 4426 

serves as (1996). 

22 October 2001; accepted 31 January 2002 

l role that 

idelity of 

www.sciencemag.org SCIENCE VOL 295 22 MARCH 2002 2261 

mps SC-ura media. Glucose was then added (transcription 

off) to a final concentration of 2% and aliquots were 

removed at the indicated time points. Approximately 

10 �g of total RNA isolated with the hot phenol method 

(4) was electrophoresed on a 1.2% agarose formaldehyde 

gel, transferred to a nylon membrane (Gene- 

Screen Plus, NEN), and hybridized with a 23 bp oligonucleotide 

end-labeled radioactive probe that specifically 

detects the polyG tract in the 3� UTR of the PGK1 

transcripts (8). Half-lives were determined by plotting 

the percent of mRNA remaining versus time on a semilog 

plot. All half-lives were determined at 25°C except 

for the ccr4� (performed at 30°C) and dcp1-2 [performed 

after shift to the nonpermissive temp (37°C) for 




confluency, trypsinized, and resuspended in 300 �l of 

RPMI 1640 without FBS, 10 mM glucose and 0.1 mM 


and cells were electroporated (Bio-Rad, Gene Pulser II 



kit, Qiagen, 2 �g), isolated 36 hours after transfection, 

was separated on a 1.6% agarose formaldehyde 

gel, transferred to a nylon membrane, and hybridized 

with a polymerase chain reaction (PCR) product comprising 

exons 10 through 12 of the �-gluc gene that 


(Random Primed DNA Labeling Kit, Boehringer Mannheim). 

The membrane was subsequently stripped 

with boiling 0.5% SDS and probed with radioactively 


quantitated using an Instant Imager (Packard). For 

subcellular fractionation, transiently transfected 

HeLa cells were trypsinized, washed in cold 1� phosphate 

buffered saline, and pelleted by centrifuging at 

2040g for 5 min at 4°C. Cells were then resuspended 

in 200 �l of 140 mM NaCl, 1.5 mM MgCl2 , 10 mM 


vortexed, and incubated on ice for 5 min. Nuclei were 

pelleted by centrifuging at 12,000g for 45 s at 4°C 

and subsequently separated from the aqueous (cytoplasmic) 

phase. 

29. J. H. Graber, C. R. Cantor, S. C. Mohr, T. F. Smith, Proc. 


30. F. Chen, C. C. MacDonald, J. Wilusz, Nucleic Acids Res. 

23, 2614 (1995). 

31. Plasmid pG-26 (25) containing the CBP1 gene under 

the control of the GAL promoter and the LEU2 

selectable marker was introduced into wild-type 

allele was and SKI7-deleted strains. Yeast were grown in me- 

generated by site-directed mutagenesis (Quik-Change dia lacking leucine and containing 2% galactose at 

Site-Directed Mutagenesis Kit, Stratagene) of the WT- 30°C to an OD600 of 0.5. Cells were pelleted and 

�gluc construct to create nonsense-�gluc. Nonstop- resuspended in media lacking a carbon source. 

�gluc was generated from WT-�gluc by two rounds of Glucose was then added to a final concentration of 

site-directed mutagenesis (Quik-Change Site-Directed 2% at time 0 and time points were taken. RNA was 

Mutagenesis Kit, Stratagene), which removed the bona extracted and analyzed by northern blotting with a 

fide termination codon and all in-frame termination labeled oligo probe oRP1067 (5�-CTCGGTCCTGcodons 

in the 3� UTR. Ter-poly(A)-�gluc was generated TACCGAACGAGACGAGG-3�). 

from nonstop-�gluc via a single point mutation, which 32. We thank N. Martin, A. Atkin, C. Dieckmann, and M. 

created a stop one codon upstream of the poly(A) tail. Sands for the provision of valuable reagents. This 

All changes were confirmed by sequencing. Primer se- work was supported by a grant from NIH (GM55239) 

quences are available upon request. 

(H.C.D.), the Howard Hughes Medical Institute 

19. P. Belgrader, J. Cheng, X. Zhou, L. S. Stephenson, L. E. (H.C.D, R.P.), and the Medical Scientist Training Pro- 

Maquat, Mol. Cell. Biol. 14, 8219 (1994). 

gram (P.A.F.). 

20. O. Kessler, L. A. Chasin, Mol. Cell. Biol. 16, 4426 

(1996). 

22 October 2001; accepted 31 January 2002 

www.sciencemag.org SCIENCE VOL 295 22 MARCH 2002 2261 

iencemag.org on June 29, 2007 

Downloaded from 

from www.sciencemag.org 

on June on Downloaded June 29, 2007 29, 2007 from www.sc

The old notion of natural selection are usually opaque, and a lack of response 

as an omnipotent force in biologi- to selection may reflect nothing more than 

cal evolution has given way to one a lack of heritable variation (4, 7). If the 

where adaptive processes are constrained cause of a constraint is to be elucidated, 

by physical, chemical, and biological exi- it must be for a simple phenotype whose 

gencies (1–4). Whether constraint and/or relationship to fitness is understood. 

stabilizing selection explain phenotypic Coenzyme use by β-isopropylmalate 

stasis, in the fossil record and in phylog- dehydrogenase (IMDH) is a simple pheenies, 

remains an open question (5). Direct notype whose etiology and relationship 

experimental tests of constraint are scarce to fitness are understood (10, 11). IMDHs 

(6–9). Even tight correlations among catalyze the oxidative decarboxylation of 

traits, at once suggestive of constraint, can β-isopropylmalate to α-ketoisocaproate 

be broken by artificial selection to produce during the biosynthesis of leucine, an es- 

new phenotypic combinations (8, 9). Desential amino acid. All IMDHs use nicospite 

all circumstantial evidence, results tinamide adenine dinucleotide (NAD) as a 

from direct experimental tests imply that coenzyme (cosubstrate). This invariance of 

selection is largely unconstrained. function among IMDHs hints at the pres- 

The direct experimental test for conence of ancient constraints, even though 

straint is conceptually simple. A pheno- some related isocitrate dehydrogenases 

type is subjected to selection (natural or (IDHs) use NADP instead (12, 13). 

artificial) in an attempt to break the pos- Structural comparisons with related 

tulated constraint (6–9). A response to se- NADP-using IDHs identify amino aclection 

indicates a lack of constraint. No ids controlling coenzyme use (14–16) 

response to selection indicates the presence 

of a constraint. However, the cause of 

a constraint is rarely specified because the 

etiologies of most phenotypes are not well 

(Fig. 1A). Introducing five replacements 

(Asp 

understood, their relationships to fitness 

236 → Arg, Asp289 → Lys, Ile290→ Tyr, 

Ala296→ Val, and Gly337→ Tyr) into the 

coenzyme-binding pocket of Escherichia 

coli leuB–encoded IMDH by site-directed 

mutagenesis causes a complete reversal 

in specificity (10, 11): NAD performance 

(k NAD 

cat /K NAD 

m ,where K is the Michae- 

m 

lis constant) is reduced by a factor of 

340, from 68 × 103 M –1 s –1 to 0.2 × 103 (16), and controls the activation and differentiation 

of T cells (17). Our finding that a mycobacterial 

3. Y. Wang et al., J. Immunol. 169, 2422 (2002). 


5. P. Srivastava, Annu. Rev. Immunol. 20, 395 (2002). 

26 May 2006; 

10.1126/scien 

Direct Demonstration of an 

Adaptive Constraint 

Stephen P. Miller, 1 Mark Lunzer, 1 Antony M. Dean 1,2 than a lack 

cause of a 

be for a sim 

fitness is u 

Coenzym 

* 

genase (IM 

etiology and 

The role of constraint in adaptive evolution is an open question. Directed evolution of an engineered (10, 11). IM 

b-isopropylmalate dehydrogenase (IMDH), with coenzyme specificity switched from nicotinamide ation of b 

adenine dinucleotide (NAD) to nicotinamide adenine dinucleotide phosphate (NADP), always produces during the 

mutants with lower affinities for NADP. This result is the correlated response to selection for relief amino acid. 

from inhibition by NADPH (the reduced form of NADP) expected of an adaptive landscape subject to dinucleotide 

three enzymatic constraints: an upper limit to the rate of maximum turnover (kcat ), a correlation in This invaria 

NADP and NADPH affinities, and a trade-off between NAD and NADP usage. Two additional constraints, at the presen 

high intracellular NADPH abundance and the cost of compensatory protein synthesis, have ensured some relate 

the conserved use of NAD by IMDH throughout evolution. Our results show that selective mechanisms use NADP 

and evolutionary constraints are to be understood in terms of underlying adaptive landscapes. 

Structur 

using IDHs 

he old notion of natural selection as an constraint, can be broken by artificial selection to enzyme use 

omnipotent force in biological evolution produce new phenotypic combinations (8, 9). De- replacemen 

Thas 

given way to one where adaptive prospite all circumstantial evidence, results from Ile 

cesses are constrained by physical, chemical, and direct experimental tests imply that selection is 

biological exigencies (1–4). Whether constraint largely unconstrained. 

and/or stabilizing selection explain phenotypic The direct experimental test for constraint is 

stasis, in the fossil record and in phylogenies, re- conceptually simple. A phenotype is subjected to 

mains an open question (5). Direct experimental selection (natural or artificial) in an attempt to 

tests of constraint are scarce (6–9). Even tight break the postulated constraint (6–9). A response 

correlations among traits, at once suggestive of to selection indicates a lack of constraint. No response 

to selection indicates the presence of a 

constraint. However, the cause of a constraint is 

rarely specified because the etiologies of most 

phenotypes are not well understood, their relationships 

to fitness are usually opaque, and a lack of 

response to selection may reflect nothing more 

290 Y Ty 

into the coe 

coli leuB–e 

genesis cau 

(10, 11): N 

Km is the 

factor of 34 

103 M –1 

iDCs (Fig. 4C). Furthermore, pretreatment of 

iDCs with pertussis toxin or TAK-779 inhibited 

the ability of peptide-loaded myHsp70 to stimu- 

further connection between the innate and adaptive 

immune responses during mycobacterial infection. 

The cellular aggregation induced by myHsp70 

K. T. Jean 

9. J. E. Hildr 

Eur. J. Im 

late DCs to generate influenza peptide–specific signaling through CCR5 may play an important 10. P. Revy, M 

Nat. Imm 

CTLs (Fig. 4D). To examine the effect of CCR5- role in the formation of granulomas, the hallmark 

11. M. L. Dus 

mediated signaling in mycobacterial infection, we of mycobacterial infection. 


363 (200 

used the model pathogen M. bovis BCG-lux (12). Microbial-induced DC responses need to be 12. B. Kampm 

As reported for murine DCs (13), human DCs highly regulated, reflecting a balance between a 13. K. A. Bod 

69, 800 

from Adaptive immune people wereConstraint unable to kill inter- rapid and appropriate response to invading mi- 

14. J. G. Cyst 

nalized mycobacteria, even in the presence of crobes and the inducement of immunopathology 15. M. Nieto 

autologous T cells. TAK-779 inhibition of CCR5 (2)—a particular problem in mycobacterial in- 16. F. Castelli 

led to a dose-dependent enhancement of intrafection. An increasing number of human patho- 17. S. A. Luth 

18. T. Dragic 

cellular mycobacterial replication (Fig. 4E and fig. gens, including HIV (18), toxoplasma (19), and 

19. J. Aliberti 

S6A) at all concentrations of mycobacteria tested M. tuberculosis (as described here), target the 20. E. de Silva 

(fig. S6B), suggesting an important role for this CCR5 receptor. This role of CCR5 as a pattern- 21. We are gr 

receptor in controlling mycobacterial infection. recognition receptor for myHsp70 may, at least Spectroph 

The identification of CCR5 as the critical re- in part, be responsible for the maintenance of the the Wellco 

(R.A.F.), th 

ceptor for myHsp70-mediated DC stimulation has high CCR5D32 allele frequency (10 to 15%) in FEBS (E.H 

implications for both mycobacterial infection and Northern European populations (20) and may 

the therapeutic use of myHsp70. CCR5 is impor- alter the pattern of disease seen in people with Supporting O 

tant in immune cell cross talk. Interaction with its the CCR5D32 allele. 

www.sciencema 

naturally occurring ligand MIP-1b promotes the 

Materials and M 

recruitment of cells to sites of inflammation (14), References and Notes 


facilitates immune synapse formation (15), orches- 1. J. L. Flynn, J. Chan, Trends Microbiol. 13, 98 (2005). 

Table S1 

Videos S1 to S 

2. J. L. Flynn, J. Chan, Annu. Rev. Immunol. 19, 93 (2001). 

trates T cell interactions within lymph nodes 

3. Y. Wang et al., J. Immunol. 169, 2422 (2002). 

References 

(16), and controls the activation and differentiation 4. M. J. Gething, J. Sambrook, Nature 355, 33 (1992). 26 May 2006; 

of T cells (17). Our finding that a mycobacterial 5. P. Srivastava, Annu. Rev. Immunol. 20, 395 (2002). 10.1126/scienc 


Adaptive Constraint 

(kNADP cat /Km NA 

1 2 from 0.49 � 

Stephen BioTechnology P. Miller, Institute, Department of Ecology, Evolution 

and Behavior, University of Minnesota, St. Paul, MN 

The engine 

55108, USA. 

final R re 


wild-type E 

deanx024@umn.edu 

cific toward 

458 

20 OCTOBER 2006 VOL 314 SCIENCE www.sciencemag.org 

1 Mark Lunzer, 1 Antony M. Dean 1,2 than a lack 

cause of a c 

be for a sim 

fitness is un 

Coenzym 

* 

genase (IM 

etiology and 

The role of constraint in adaptive evolution is an open question. Directed evolution of an engineered (10, 11). IM 

b-isopropylmalate dehydrogenase (IMDH), with coenzyme specificity switched from nicotinamide ation of b- 

adenine dinucleotide (NAD) to nicotinamide adenine dinucleotide phosphate (NADP), always produces during the 

mutants with lower affinities for NADP. This result is the correlated response to selection for relief amino acid. 

from inhibition by NADPH (the reduced form of NADP) expected of an adaptive landscape subject to dinucleotide 

three enzymatic constraints: an upper limit to the rate of maximum turnover (kcat ), a correlation in This invaria 

NADP and NADPH affinities, and a trade-off between NAD and NADP usage. Two additional constraints, at the presen 

high intracellular NADPH abundance and the cost of compensatory protein synthesis, have ensured some relate 

the conserved use of NAD by IMDH throughout evolution. Our results show that selective mechanisms use NADP 

and evolutionary constraints are to be understood in terms of underlying adaptive landscapes. 

Structura 

using IDHs 

he old notion of natural selection as an constraint, can be broken by artificial selection to enzyme use 

omnipotent force in biological evolution produce new phenotypic combinations (8, 9). De- replacemen 

Thas 

given way to one where adaptive prospite all circumstantial evidence, results from Ile 

cesses are constrained by physical, chemical, and direct experimental tests imply that selection is 

biological exigencies (1–4). Whether constraint 

and/or stabilizing selection explain phenotypic 

stasis, in the fossil record and in phylogenies, remains 

an open question (5). Direct experimental 

tests of constraint are scarce (6–9). Even tight 

largely unconstrained. 

The direct experimental test for constraint is 

conceptually simple. A phenotype is subjected to 

selection (natural or artificial) in an attempt to 

break the postulated constraint (6–9). A response 

correlations among traits, at once suggestive of to selection indicates a lack of constraint. No response 

to selection indicates the presence of a 

constraint. However, the cause of a constraint is 

rarely specified because the etiologies of most 

phenotypes are not well understood, their relationships 

to fitness are usually opaque, and a lack of 

response to selection may reflect nothing more 

290 Y Ty 

into the coe 

coli leuB–e 

genesis cau 

(10, 11): N 

Km is the 

factor of 34 

103 M –1 s 

1 2 

BioTechnology Institute, Department of Ecology, Evolution 

and Behavior, University of Minnesota, St. Paul, MN 

55108, USA. 


deanx024@umn.edu 

(kNADP cat /Km NA 

from 0.49 � 

The engine 

final R rep 

wild-type E 

cific toward 

458 

24 

20 OCTOBER 2006 VOL 314 SCIENCE www.sciencemag.org

M –1 s –1 , whereas NADP performance 

(k NADP 

cat /K NADP 

m ) is increased by a factor of 

70, from 0.49 × 103 M –1 s –1 to 34 × 103 M –1 

s –1 . The engineered LeuB[RKYVYR] mutant 

(the final R represents Arg341 , already 

present in wild-type E. coli IMDH) is as 

active and as specific toward NADP as the 

wild-type enzyme is toward NAD. Evidently, 

protein architecture has not constrained 

IMDH to use NAD since the last 

common ancestor. 

Despite similar in vitro performances, 

the NADP-specific LeuB[RKYVYR] mu- 

tant is less fit than the NAD-specific wild 

type (11). IMDHs, wild-type and mutant 

alike, display a factor of 30 higher affinity 

for the reduced form of NADP (NADPH) 

(coproduct) than for NADP (Fig. 1B). We 

suggested the LeuB[RKYVYR] mutant is 

subject to intense inhibition by intracellular 

NADPH, which is far more abundant 

in vivo than is NADP (11, 17). The inhibition 

slows leucine biosynthesis, reduces 

growth rate, and lowers Darwinian fitness 

(Fig. 1C). The wild type retains high fitness 

because its affinity for NADPH is 

REPORTS 

Fig. 1. The molecular anatomy of an adaptive constraint. (A) Structural and NADPH (K NADPH 

phosphorus) with labels designating i the ) seen with engin 

alignment of the coenzyme binding pockets of E. coli amino IMDH acid (14) and (brown site number in the in screened IMDH followed mutants (dots). (C) Th 

main chain) with bound NAD modeled from Thermus thermophilus by the amino IMDH acid in by IDH. IMDH Coenzyme (11) showing use the phenotype 

(15) and E. coli IDH (16) (green main chain) with bound is determined NADP. Only by keyH 

bonds in chemostat to NAD competition (brown and descr 

residues are shown (gray, carbon; red, oxygen; blue, lines) nitrogen; and to yellow, the 2’-phosphate predicted (2’P) distribution of NADP of performanc 

phosphorus) with labels designating the amino acid and (green sitelines). number (B) Mutations in binding in the pocket coenzyme (gray dots), with wil 

IMDH followed by the amino acid in IDH. Coenzyme use binding is determined pocket by that H dot), stabilize andNADP single also amino acid repla 

bonds to NAD (brown lines) and to the 2-phosphate (2P) stabilize of NADP NADPH. (greenThe 

binding ensuing pocket correlation (pink dots). The fitne 

lines). (B) Mutations in the coenzyme binding pocket that in affinities stabilize for NADP NADP be (K lower than in the wild type bec 

also stabilize NADPH. The ensuing correlation in affinities for NADP (KNADP m ) intracellular NADPH (11). 

toward NAD. Evidently, protein architecture has 

not constrained IMDH to use NAD since the last 

common ancestor. 

Despite similar in vitro performances, the 

NADP-specific LeuBERKYVYR^ mutant is less 

fit than the NAD-specific wild type (11). IMDHs, 

PH (K NADPH 

i wild-type 

) seen Fig. 1. with The and 

engineered molecular mutant alike, 

mutants anatomy display 

(circles) of a factor an (11) adaptive of 

is 

30 

retained 

reened mutants 

higher constraint. (dots). 

affinity 

(C) (A) The 

for 

adaptive Structural the reduced 

landscape alignment form 

for 

of 

coenzyme 

NADP of the use 

(11) showing 

(NADPH) coenzyme the phenotype-fitness 

(coproduct) binding pockets map 

than for of (blue 

NADP E. coli surface) IMDH determined 

(Fig. 1B). (14) 

ostat competition 

We (brown and 

suggested main described 

thechain) by 

LeuBERKYVYR^ with equation bound S5 NAD (11, 18) and the 

mutant modeled is sub- 

d distribution of performances for 512 mutants in the coenzyme 

pocket (gray 

ject from 

dots), 

to intense Thermus 

with wild 

inhibition thermophilus 

type (red 

by intracellular IMDH (15) 

dot), LeuB[RKYVYR] 

NADPH, and E. 

(black 

d single amino 

which coli is IDH 

acid 

far(16) more(green replacements 

abundant main 

in 

in vivo chain) 

LeuB[RKYVYR] 

than with is NADP bound 

coenzyme 

pocket (pink(11, NADP. 

dots). 17). The The 

Only 

fitness inhibition 

key residues 

of LeuB[RKYVYR] slows leucine 

are shown 

biosynthesis, 

(gray, 

carbon; red, oxygen; blue, nitrogen; is hypothesized yellow, to 

r than in the reduces wild type growth because rate, and of lowers strong Darwinian inhibition fitness by abundant 

ular NADPH (Fig. (11). 1C). The wild type retains high fitness because Fig. 2. The phenotypic basis of the genetic screen. Lowered expr 

its affinity for NADPH is low, whereas inhibition by brings IMDH to saturation with isopropylmalate, 25 increasing co 

NADH, which is far less abundant than NAD in expression in the chemostat-derived adaptive landscape [lower con 

NADP 

m ) and NADPH 

(K NADPH 

i ) seen with engineered mutants 

(circles) (11) is retained in the screened 

mutants (dots). (C) The adaptive landscape 

for coenzyme use by IMDH (11) showing 

the phenotype-fitness map (blue surface) 

determined in chemostat competition and 

described by equation S5 (11, 18) and the 

predicted distribution of performances 

for 512 mutants in the coenzyme binding 

pocket (gray dots), with wild type (red dot), 

LeuB[RKYVYR] (black dot), and single amino 

acid replacements in LeuB[RKYVYR] coenzyme 

binding pocket (pink dots). The fitness of 

LeuB[RKYVYR] is hypothesized to be lower than 

in the wild type because of strong inhibition by 

abundant intracellular NADPH (11)

. Coenzyme use is determined by H 

2-phosphate (2P) of NADP (green 

inding pocket that stabilize NADP 

ation in affinities for NADP (K m NADP ) 

cture has 

e the last 

ces, the 

nt is less 

IMDHs, 

tor of 30 

f NADP 

Fig. 1B). 

nt is sub- 

NADPH, 

is NADP 

ynthesis, 

an fitness 

s because 

ibition by 

NAD in 

uence of 

ure (18), 

DPH inuse. 

rained to 

ssociated 

, identifynhibition) 

causes of 

P (maxireak 

the 

ties (Fig. 

oenzyme 

nefit the 

romising 

therefore 

use NAD 

at ADP , an 

f NADP 

de-off in 

low, whereas inhibition by NADH, which 

is far less abundant than NAD in vivo, is 

ineffective. Perhaps as a consequence of 

differences in Michaelis complex structure 

(18), IDH is not subject to such intense 

NADPH inhibition and hence could 

evolve NADP use. 

We hypothesize that IMDH is constrained 

to use NAD because the strong 

inhibition associated with NADP use reduces 

fitness. However, identifying the 

mechanism of selection (NADPH inhibition) 

is not synonymous with identifying 

the causes of constraint. Mutations that 

increase k NADP 

cat (maximum rate of NADP 

turnover), that break the correlation in 

NADP and NADPH affinities (Fig. 1B), 

or that eliminate the trade-off in coenzyme 

performances (Fig. 1C) could each 

benefit the LeuB[RKYVYR] mutant without 

compromising its performance with 

NADP (18). We therefore hypothesize 

that IMDH is constrained to use NAD by 

three causes: an upper limit to kNADP ing) (19–21) to test whether NADP-specific 

IMDHs with higher fitness could 

be isolated. Random substitutions were 

introduced into leuB[RKYVYR] by means 

of error-prone polymerase chain reaction 

(18). Mutated alleles were ligated downstream 

of the T7 promoter in pETcoco (a 

stable single-copy vector) and transformed 

into strain RFS[DE3] (leuA 

, an 

cat 

+ BamC + We used directed evolution (targeted random 

mutagenesis and selective screening) (19–21) to 

test whether NADP-specific IMDHs with higher 

fitness could be isolated. Random substitutions 

were introduced into leuB[ RKYVYR] by means 

of error-prone polymerase chain reaction (18). 

Mutated alleles were ligated downstream of the 

T7 promoter in pETcoco (a stable single-copy 

vector) and transformed into strain RFSEDE3^ 

(leuA , with T7 

RNA polymerase expressed from a chromosomal 

lacUV5 promoter). Sequencing 

unscreened plasmids revealed that, on average, 

each 1110–base pair leuB[RKYVYR] 

received two nucleotide substitutions. 

From the pattern of base substitutions in 

þBamCþ assuming Poisson statistics, we estimate that only 

10.5 (0.4%) of the 2431 possible amino acid replacements 

were missing in our mutant library (18). 

Our experimental design used decreased 

IMDH expression to provide a simple selective 

screen for mutations in leuB[ RKYVYR] that increase 

growth and/or fitness. As predicted from 

the phenotype-fitness map (Fig. 2A), selection 

against leuB[ RKYVYR] intensified as IMDH 

, with T7 RNA polymerase ex- expression was lowered in the presence of excess 

pressed from a chromosomal lacUV5 promoter). glucose (Fig. 2B). Using 40 mM isopropyl-b-D- 

Sequencing unscreened plasmids revealed that, thiogalactopyranoside (IPTG) to induce a low 

on average, each 1110–base pair leuB[ RKYVYR] level of IMDH expression, we found that cells 

received two nucleotide substitutions. From the harboring leuB[ WT] formed large colonies at 24 

pattern of base substitutions in these mutants and hours, whereas cells harboring leuB[ RKYVYR] 

ww.sciencemag.org SCIENCE VOL 314 20 OCTOBER 2006 459 

unbreakable correlation in the affinities of 

NADP and NADPH, and an inescapable 

trade-off in coenzyme performance. 

We used directed evolution (targeted 

random mutagenesis and selective screen- 

26 

dot), and single amino acid replacements in LeuB[RKYVYR] coenzyme 

binding pocket (pink dots). The fitness of LeuB[RKYVYR] is hypothesized to 

be lower than in the wild type because of strong inhibition by abundant 

intracellular NADPH (11). 

Fig. 2. The phenotypic basis of the genetic screen. Lowered expression in the presence of excess glucose 

brings IMDH to saturation with isopropylmalate, increasing coenzyme affinities. (A) Lowering IMDH 

expression in the chemostat-derived adaptive landscape [lower concentration of E in equation S5 (11, 18)] is 

predicted to reduce the fitness of LeuB[RKYVYR] (white sphere) far more than that of the wild type (red 

sphere). (B) IPTG-controlled expression of IMDHs ligated downstream of the T7 promoter in pETcoco in strain 

RFS[DE3] (leuA þ B am C þ , with T7 RNA polymerase expressed from a chromosomal lacUV5 promoter) confirms 

that lower expression affects growth in minimal glucose medium of the LeuB[RKYVYR] (white) more than in 

the wild type (red). Plasmids lacking LeuB (black) are incapable of growth except in the presence of leucine. 

these mutants and assuming Poisson statistics, 

we estimate that only 10.5 (0.4%) of 

the 2431 possible amino acid replacements 

were missing in our mutant library (18). 

Our experimental design used decreased 

IMDH expression to provide a 

simple selective screen for mutations 

in leuB[RKYVYR] that increase growth 

and/or fitness. As predicted from the 

phenotype-fitness map (Fig. 2A), selection 

against leuB[RKYVYR] intensified 

as IMDH expression was lowered in the 

presence of excess glucose (Fig. 2B). 

Using 40 μM isopropyl-β-D-thiogalacto

eplacement in the coenzyme binding pocket, or 

both. Upstream substitutions occurred at three 

nucleotide positions: –3, –9, and –14 relative to 

eliminate H bonds to the 2-phosphate of NADP 

to cause striking reductions in NADP performance 

(Table 1). Although some mutants have improved 

Table 1. Kinetic effects of amino acid replacements in LeuB[RKYVYR] isolated at 48 hours growth. 

Standard errors are G13% of estimates. Single-letter abbreviations for amino acid residues: A, Ala; 

C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, 

Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr. 

NADP NAD 

Enzyme 

Performance 

(A) 

Performance 

(B) 

Preference 

(A/B) 

K NADPH 

i 

(mM) 

Km (mM) 

kcat (s –1 ) 

kcat /Km (mM –1 s –1 ) 

Km (mM) 

kcat (s –1 ) 

kcat /Km (mM –1 s –1 ) 

Site 

DDIAGR (wild type) 8400 4.10 0.4900 101 6.90 68.3200 0.007 254.30 

RKYVYR 183 6.20 33.8800 4108 0.83 0.2000 169.400 9.90 

Beneficial replacements 

K235N 3919 5.95 1.5200 6356 0.50 0.0800 19.000 129.00 

K235N,–3(M),A60V 1744 5.06 2.9000 5395 0.44 0.0800 35.700 56.00 

K235R 554 6.39 11.5000 3292 0.36 0.1100 106.100 26.00 

K235R,E63V 658 6.75 10.3000 3748 0.46 0.1200 84.100 28.00 

K289E* 3449 2.71 0.7900 4897 1.98 0.4000 2.000 133.50 

K289E*,D317E 3554 3.62 1.0200 5580 2.09 0.3700 2.800 112.00 

K289E*,D87Y,S182F 2551 2.54 1.0000 5750 3.65 0.6300 1.600 105.20 

K289E*,F102L 3891 3.66 0.9400 4823 1.10 0.2300 4.100 131.90 

K289M 2216 4.13 1.8600 4034 0.48 0.1200 15.500 78.60 

K289M,F170L 1945 2.50 1.2900 2947 0.48 0.1600 8.100 87.20 

K289N* 1141 6.32 5.5400 4997 1.14 0.2300 24.300 44.00 

K289N*,R187H 1146 5.83 5.0900 3069 0.79 0.2600 19.600 44.70 

K289N*,H367Q 1248 5.27 4.2200 4555 1.05 0.2300 18.300 57.80 

K289T 1596 6.18 3.8700 4056 0.78 0.1900 20.400 71.60 

K289T,Q157H 1696 5.66 3.3400 5038 0.93 0.1800 18.600 72.60 

Y290C 988 5.21 5.2700 3792 1.02 0.2700 19.600 49.00 

Y290C,R152C 910 3.15 3.4600 3844 0.90 0.2300 14.800 43.00 

Y290D 10213 3.13 0.3100 9040 0.82 0.0900 3.400 344.00 

Y290F* 1658 5.59 3.3700 3110 0.65 0.2100 16.000 59.70 

Y290F,P97S,D314E 1097 4.64 4.2300 4158 0.71 0.1700 24.800 57.40 

Y290N,N52T 2381 3.44 1.4400 10660 0.26 0.0200 59.500 90.00 

Replacements with beneficial 5-leader mutations 

RKYVYR 183 6.20 33.8800 4108 0.83 0.2000 169.400 9.90 

–3(M)†,E66D 208 2.97 14.2800 5002 0.30 0.0600 238.000 11.90 

–3(M)†,E66D,E331D 267 7.99 29.9000 3403 0.34 0.1000 296.900 14.00 

E82D 221 5.44 24.6200 4735 0.35 0.0700 351.700 14.10 

K100R,A229T,L248M 137 5.58 40.3700 4079 0.45 0.1100 370.300 11.50 

G156R,K289R,Y311H 292 5.40 18.4900 4464 0.54 0.1200 144.800 17.40 

E173D 219 6.39 29.1800 5304 0.89 0.1700 171.600 6.60 

Y337N,G131 171 5.71 33.3900 3367 0.24 0.0700 468.500 8.00 

Y337H,P85S,E181K 318 5.21 16.3800 2409 0.22 0.0900 183.600 12.80 

*Switch from NADP to NAD requires two base substitutions in one codon. This is a possible transitional amino acid produced by 

a single base substitution (11). †The –3(M) designates a G to A substitution at base –3 that creates a new start codon. 

460 

pyranoside (IPTG) to induce a low level terion compatible with reliably identifying 

20 OCTOBER 2006 VOL 314 SCIENCE www.sciencemag.or 

of IMDH expression, we found that cells beneficial mutations in leuB[RKYVYR]. 

harboring leuB[WT] formed large colo- Longer periods of growth (or higher connies 

at 24 hours, whereas cells harboring centrations of IPTG) allowed unmutated 

leuB[RKYVYR] barely formed pinprick leuB[RKYVYR] to form colonies, whereas 

colonies at 48 hours. We chose colony for- shorter periods (or lower concentrations of 

mation at 48 hours as the least stringent cri- IPTG) produced no colonies. 

27 

are benef 

formance 

if there ar 

upper lim 

for NAD 

enzyme p 

with redu 

phenotyp 

remains f 

performan 

of NADP 

cial as co 

NADPH f 

As predi 

mutants h 

and NAD 

increases 

eficial (18 

that NAD 

vivo beca 

NADPH. 

and increa 

the predic 

map cons 

correlation 

and a trad 

Break 

NADP-sp 

increases 

for NADP 

trade-off 

clusions a 

is too stri 

distinct m 

limit, 17 

leader seq 

LeuBERK 

ments in 

zyme pe 

inhibition 

cial muta 

teria dem 

not overly 

Nor ar 

sampling 

adaptive 

sequential 

lution de 

mutants 

coli are n 

force mu 

vantageo

Of the 100,000 mutated plasmids 

screened, 134 (representing 107 distinct 

isolates) formed colonies within 48 hours 

(table S1). Each had either a substitution 

in the 5′ leader sequence upstream of 

leuB[RKYVYR], an amino acid replacement 

in the coenzyme binding pocket, or 

both. Upstream substitutions occurred at 

three nucleotide positions: –3, –9, and –14 

relative to the leuB AUG start codon (fig. 

S1). Ten amino acid replacements were 

found at three codons in the coenzyme 

binding pocket. 

At first glance, the positive response to 

selection might suggest that coenzyme use 

by IMDH is unconstrained. Upstream substitutions 

in the Shine-Dalgarno sequence 

(positions –9 and –14), as well as a new 

AUG start codon (position –3) that replaces 

the less efficient GUG start codon, 

presumably derive their benefits through 

increases in expression because their kinetics 

are unchanged. Unexpectedly, however, 

beneficial amino acid replacements 

in the coenzyme binding pocket eliminate 

H bonds to the 2′-phosphate of NADP to 

cause striking reductions in NADP performance 

(Table 1). Although some mutants 

have improved NAD performance, others 

remain unchanged and several show reduced 

NAD performance. Isolated amino 

acid replacements outside the coenzyme 

binding pocket, which might have been 

expected to increase kNADP or to break the 

cat 

correlation in affinities between NADP 

and NADPH (K NADP 

m and K NADPH 

i ), have 

no detectable functional effects (table S2, 

those associated with beneficial 5′-leader 

mutations in Table 1). No doubt they, 

along with 126 silent substitutions (table 

S1), hitchhiked through the genetic screen 

with the beneficial mutations. 

Our results suggest that increases in expression 

are beneficial, whereas increases 

in NADP performance are not. This seeming 

paradox is resolved if there are no 

mutations capable of breaking the upper 

limit to kNADP the correlation in affinities 

cat 

for NADP and NADPH, or the trade-off in 

28 

coenzyme performance. With these constraints, 

and with reduced expression in 

the genetic screen, the phenotype-fitness 

map near LeuB[RKYVYR] remains flat 

with respect to increases in NADP performance 

(Fig. 2A). By contrast, severe losses 

of NADP performance are predicted to 

be beneficial as correlated reductions in 

the affinities for NADPH free up IMDH 

for use with abundant NAD. As predicted, 

all beneficial LeuB[RKYVYR] mutants 

have reduced affinities for both NADP and 

NADPH (Table 1). Unaffected by constraints, 

increases in expression are unconditionally 

beneficial (18). These results 

support the hypothesis that NADP-specific 

IMDHs function poorly in vivo because of 

strong inhibition by abundant NADPH. 

That reductions in NADP performance 

and increases in expression are both beneficial 

are the predicted consequences of a 

phenotype-fitness map constrained by an 

upper limit to k NADP 

cat , a correlation in affinities 

for NADP and NADPH, and a tradeoff 

in coenzyme performance. 

Breaking any one constraint would allow 

NADP-specific IMDHs to evolve. 

Yet no mutant increases k NADP 

cat , no mutant 

uncouples the affinities for NADP and 

NADPH, and no mutant breaks the tradeoff 

in coenzyme performance. These conclusions 

are not the result of a selective 

screen that is too stringent. Of 35 colonies, 

representing 26 distinct mutants, that appeared 

after the 48-hour limit, 17 had no 

nucleotide substitutions in the 5′ leader 

sequence or amino acid replacements in 

LeuB[RKYVYR] (table S3). Amino acid 

replacements in the other mutants neither 

improved enzyme performance nor 

decreased NADPH inhibition (table S4). 

That no additional beneficial mutations 

were recovered with relaxed criteria demonstrates 

that the selective screen was not 

overly stringent. 

Nor are the results a consequence of 

inadequate sampling of protein sequence 

space. Natural adaptive evolution fixes 

advantageous mutations sequentially (22–

mpling of 

olutionary 

04 advane 

missing 

(18). We 

eaking mplingthe of 

re olutionary unlikely 

04 e they advan- are 

e), missing because 

oth. (18). We 

an aking NADP- the 

er unlikely NADPH 

ese they expres- are 

elieves , because the 

sources oth. of 

the n NADP- rest of 

rnefit NADPH to be 

se (Fig. expres- 1C) 

lieves such that the 

bove sources wild- of 

the rcome restthe of 

nefit resources to be 

d(Fig. compen- 1C) 

such a protein that 

the oveevolu wildcome 

the 

otypes, resources by 

ring, compen- is not 

constraints a protein 

he constraints evolu- 

e identified 

phenotype, types, by 

ing, landscape. is not 

constraints sed to test 

constraints we have 

lationships ion phenotype 

24). Indeed, experimental evolution demonstrates 

that advantageous double mutants 

in the evolved ß-galactosidase of E. 

coli are not evolutionarily accessible and 

perforce must be accumulated as sequential 

advantageous mutations (25). Hence, 

screening all single substitutions is a 

sufficient sampling of protein sequence 

space for robust evolutionary conclusions. 

We estimate that only 0.04 advantageous 

amino acid replacements are missing from 

the mutant leuB[RKYVYR] library (18). 

We conclude that mutations capable of 

breaking the limit, the correlation, or the 

trade-off are unlikely to ever be fixed in 

populations because they are exceedingly 

rare (they may not exist), because they are 

minimally advantageous, or both. 

The two remaining ways to evolve an 

underlying NADP-specific a conserved IMDH phenotype are to can reduce be identified intra- 

from cellular the relationships NADPH pool amongand, genotype, as our phenotype, results 

and show, fitness to increase that define expression. an adaptive Reducing landscape. in- 

Experimental tracellular NADPH evolution relieves can then the be used inhibition to test 

their 

but, 

existence. 

as experiments 

Using this 

deleting 

approach, 

sources 

we have 

underlying of 

shown howa certain conserved structure-function phenotype can be relationships identified 

from NADPH show (13), the disruption to the 

in IMDH the relationships have constrained among itsgenotype, coenzyme phenotype, phenotype 

and since rest fitness of themetabolism lastthat common defineancestor. costs an adaptive far more landscape. than the 

Experimental evolution can then be used to test 

their existence. Using this approach, we have 

shown References how certain andstructure-function Notes relationships 

1. S. J. Gould, R. C. Lewontin, Proc. R. Soc. London Ser. B 

in IMDH have constrained its coenzyme phenotype 

205, 581 (1979). 

since 2. J. the Antonovics, last common P. H. vanancestor. Tienderen, Trends Ecol. Evol. 6, 

166 (1991). 

3. M. R. Rose, G. V. Lauder, Adaptation (Academic Press, 

References San Diego, CA, and 1996). Notes 

1. 4. S. M. J. Pigliucci, Gould, R. J. C. Kaplan, Lewontin, Trends Proc. Ecol. R. Evol. Soc. 15, London 66 (2000). Ser. B 

5. 205, N. Eldredge 581 (1979). et al., Paleobiology 31, 133 (2005). 

2. 6. J. M. Antonovics, Travisano, J. P. A. H. Mongold, van Tienderen, A. F. Bennett, Trends Ecol. R. E. Evol. Lenski, 6, 

166 Science (1991). 267, 87 (1995). 

3. 7. M. H. R. Teotónio, Rose, G. M. V. R. Lauder, Rose, Nature Adaptation 408, (Academic 463 (2000). Press, 

8. San P. Beldade, Diego, CA, K. Koops, 1996). P. M. Brakefield, Nature 416, 844 

4. M. (2002). Pigliucci, J. Kaplan, Trends Ecol. Evol. 15, 66 (2000). 

5. 9. N. W. Eldredge A. Frankino, et al., B. J. Paleobiology Zwaan, D. L. 31, Stern, 133 P. (2005). M. Brakefield, 

6. M. Science Travisano, 307, J. 718 A. Mongold, (2005). A. F. Bennett, R. E. Lenski, 

10. Science R. Chen, 267, A. Greer, 87 (1995). A. M. Dean, Proc. Natl. Acad. Sci. 

7. H. U.S.A. Teotónio, 93, 12171 M. R. Rose, (1996). Nature 408, 463 (2000). 

11. 8. P. M. Beldade, Lunzer, S. K. P. Koops, Miller, P. R. M. Felsheim, Brakefield, A. Nature M. Dean, 416, Science 844 

(2002). 310, 499 (2005). 

12. 9. W. A. A. M. Frankino, Dean, G. B. J. Golding, Zwaan, Proc. D. L. Stern, Natl. Acad. P. M. Sci. Brakefield, U.S.A. 

Science 94, 3104 307, (1997). 718 (2005). 

10. 13. R. G. Chen, Zhu, G. A. B. Greer, Golding, A. M. A. Dean, M. Dean, Proc. Science Natl. Acad. 307, Sci. 1279 

U.S.A. (2005); 93, published 12171 (1996). online 13 January 2005 (10.1126/ 

11. M. science.1106974). 

Lunzer, S. P. Miller, R. Felsheim, A. M. Dean, Science 

14. 310, G. Wallon 499 (2005). et al., J. Mol. Biol. 266, 1016 (1997). 

12. 15. A. J. H. M. Hurley, Dean, G. A. B. M. Golding, Dean, Structure Proc. Natl. 2, 1007 Acad. (1994). Sci. U.S.A. 

16. 94, J. H. 3104 Hurley, (1997). A. M. Dean, D. E. Koshland Jr., R. M. Stroud, 

13. G. Biochemistry Zhu, G. B. 30, Golding, 8671 A. (1991). M. Dean, Science 307, 1279 

ole for 17. (2005); T. Penfound, LEDGF/p75 

published J. W. Foster, onlinein13 Escherichia January 2005 coli and (10.1126/ Salmonella: 

science.1106974). 

Cellular and Molecular Biology, F. C. Neidhardt, Ed. (ASM 

Press, Washington, DC, ed. 2, 1996), pp. 721–730. 

18. See supporting material on Science Online. 

19. O. Kuchner, F. H. Arnold, Trends Biotechnol. 58, 1 (1997). 

20. F. H. Arnold, P. L. Wintrobe, K. Miyazaki, A. Gershenson, 

Meehan, Phonphimon Wongthida, Mary Peretz, 

Trends Biochem. Sci. 26, 100 (2001). 

ole for LEDGF/p75 

REPORTS 

benefit to be gained. The phenotype-fitness 

map (Fig. 1C) imposes a law of diminishing 

returns such that LeuB[RKYVYR] 

must be expressed above wild-type levels 

by a factor of 100 to overcome the inhibition 

by NADPH (18). Diverting resources 

away from other metabolic needs toward 

compensatory protein synthesis would impose 

a protein burden (26–29) sufficient 

to prevent the evolution of NADP-specific 

IMDHs. 

The production of unnatural phenotypes, 

by artificial selection or molecular 

engineering, is not sufficient to conclude 

that evolutionary constraints are absent 

entirely. Rather, potential constraints underlying 

a conserved phenotype can be 

identified from the relationships among 

genotype, phenotype, and fitness that define 

14. G. an Wallon adaptive et al., J. Mol. landscape. Biol. 266, 1016 Experimental 

(1997). 

evolution 15. J. H. Hurley, can A. M. then Dean, be Structure used 2, to 1007 test (1994). their 

16. J. H. Hurley, A. M. Dean, D. E. Koshland Jr., R. M. Stroud, 

existence. Biochemistry Using 30, 8671 this (1991). approach, we have 

shown 17. T. Penfound, how J. certain W. Foster, in structure-function Escherichia coli and Salmonella: re- 

Cellular and Molecular Biology, F. C. Neidhardt, Ed. (ASM 

lationships 14. G. Press, Wallon Washington, etin al., IMDH J. DC, Mol. ed. Biol. have 2, 1996), 266, constrained 1016 pp. 721–730. (1997). its 

coenzyme 15. 18. J. See H. supporting Hurley, phenotype A. M. material Dean, on Structure since Science2, the Online. 1007 last (1994). com- 

16. 19. J. O. H. Kuchner, Hurley, F. A. H. M. Arnold, Dean, D. Trends E. Koshland Biotechnol. Jr., 58, R. M. 1 Stroud, (1997). 

mon 

20. Biochemistry ancestor. 

F. H. Arnold, 30, P. L. 8671 Wintrobe, (1991). K. Miyazaki, A. Gershenson, 

17. T. Trends Penfound, Biochem. J. W. Foster, Sci. 26, in100 Escherichia (2001). coli and Salmonella: 

21. Cellular L. G. Otten, and Molecular W. J. Quax, Biology, Biomol. F. C. Eng. Neidhardt, 22, 1 (2005). Ed. (ASM 

22. Press, S. Wright, Washington, in Proceedings DC, ed. 2, of 1996), the Sixth pp. International 721–730. 

18. See Genetics supporting Congress, material D. F. on Jones, Science Ed. (Brooklyn Online. Botanic 

19. O. Garden, Kuchner, Menasha, F. H. Arnold, WI, 1932), Trendspp. Biotechnol. 356–366. 58, 1 (1997). 

20. 23. F. J. H. Maynard Arnold, Smith, P. L. Wintrobe, Nature 225, K. Miyazaki, 563 (1970). A. Gershenson, 

24. Trends D. M. Weinreich, Biochem. Sci. N. F. 26, Delaney, 100 (2001). M. A. Depristo, D. L. Hartl, 

21. L. Science G. Otten, 312, W. 111 J. Quax, (2005). Biomol. Eng. 22, 1 (2005). 

22. 25. S. B. Wright, G. Hall, inAntimicrob. Proceedings Agents of theChemother. Sixth International 46, 3035 

Genetics (2002). Congress, D. F. Jones, Ed. (Brooklyn Botanic 

26. Garden, S. Zamenhof, Menasha, H. H. WI, Eichhorn, 1932), pp. Nature 356–366. 216, 456 (1967). 

23. 27. J. K. Maynard J. Andrews, Smith, G. D. Nature Hegeman, 225, J. 563 Mol. (1970). Evol. 8, 317 (1976). 

24. 28. D. M. E. Weinreich, Dykhuizen, N. Evolution F. Delaney, 32, M. 125 A. Depristo, (1978). D. L. Hartl, 

29. Science A. L. Koch, 312, J. 111 Mol. (2005). Evol. 19, 455 (1983). 

25. 30. B. Supported G. Hall, Antimicrob. by NIH grant Agents GM060611 Chemother. (A.M.D.). 46, We 3035 thank 

(2002). three anonymous reviewers whose constructive criticisms 

26. S. weZamenhof, found invaluable. H. H. Eichhorn, Nature 216, 456 (1967). 

27. K. J. Andrews, G. D. Hegeman, J. Mol. Evol. 8, 317 (1976). 

Supporting 28. D. E. Dykhuizen, OnlineEvolution Material 32, 125 (1978). 


29. A. L. Koch, J. Mol. Evol. 19, 455 (1983). 

Materials 30. Supported and Methods by NIH grant GM060611 (A.M.D.). We thank 

SOM three Text anonymous reviewers whose constructive criticisms 

Figs. we S1 found to S3 invaluable. 

Tables S1 to S5 

Supporting References Online Material 


4 August 2006; accepted 20 September 2006 


10.1126/science.1133479 

SOM Text 


Tables S1 to S5 

References the HIV-1 IN protein from rapid degradation in 

4the August 26S2006; proteasome accepted 20(8). September In the2006 bona fide viral 

10.1126/science.1133479 

context, drastic knockdown of p75 changed the 

genomic pattern of HIV-1 integration by reducing 

the viral bias for active genes, which suggests that 29 

the p75HIV-1 influences IN protein integration fromtargeting rapid degradation (9). However, in 

the 26S proteasome (8). In the bona fide viral 

REPORTS 

REPORTS

Tech Notes 

Large Insertions: Two Simple Steps Using 

QuikChange® II Site-Directed Mutagenesis Kits 

Abstract 

This Technical Note describes a modified 

method to introduce large insertions in 

plasmids in two simple steps using the 

QuikChange ® II Site-Directed Mutagenesis 

Kit. Dr. Wenge Wang in Dr. Wafik S. 

El-Deiry’s lab at the University of Pennsylvania 

has used the QuikChange kit and 

a modification of the method described 

by Geiser et al. (2001) 1 to perform a large 

insertion of a PCR product into a target 

plasmid. He had inserted the enhanced 

green fluorescent protein (EGFP) open 

reading frame (760 bp) following the Cterminus 

of death receptor TRAIL receptor 

1/death receptor 4 (TRAIL-R1/DR4) 

right after the signal sequence, using the 

QuikChange II site-directed mutagenesis 

kit with some modifications to the basic 

protocol. Dr. Wang transfected cells with 

the recombined plasmid and it generated 

a fusion protein with the correct molecular 

size (Figure 1). 

Background 

Dr. Wafik S. El-Deiry’s lab is focused on 

studying the mechanism of action of the 

tumor suppressor p53 and the contribution 

of its downstream target genes to 

control cell growth. Their lab has identified 

a number of genes that are directly 

regulated by p53 and which can inhibit 

cell cycle progression (p21WAF1), induce 

apoptosis (KILLER/DR5, Bid, caspase 6, 

Traf4 and others) or activate DNA repair 

(DDB2). The research has provided knowledge 

into the tissue specificity of the DNA 

damage response in vivo and into the 

mechanism by which wild-type p53 sensitizes 

cells to killing by anti-cancer drugs. 

High-Efficiency Mutagenesis Method 

All QuikChange ® kits a offer a one-day 

method to introduce point mutations, amino 

acid substitutions, small insertions, and 

deletions in virtually any double-stranded 

30 

plasmid template at efficiencies greater 

than 80% (Figure 2). The QuikChange II 

and QuikChange II XL kits feature our 

highest fidelity DNA polymerase, our gold 

standard for accuracy, and a linear amplification 

strategy. Together, they reduce the 

mutation frequency due to incorporation 

errors typically caused by using an exponential 

PCR amplification approach. The 

result is high-efficiency mutagenesis without 

unwanted errors. 

Ideal for Large Insertions and Deletions 

We have used the QuikChange kit method 

for small insertions and deletions (~12 bp) 

and observed greater than 80% efficiency 2 . 

Many of our customers have used the 

QuikChange method or have modified it 

to introduce deletions of 31 bp 3 , 87 bp 4 , 

and 3 kb 5 or insertions of 31 bp 3 and up 

to ~1 kb 5 . In order to achieve large insertions 

of up to 1 kb, megaprimers must be 

generated from an initial PCR reaction. To 

ensure the highest mutagenesis efficiency, 

we recommend gel purification with our 

StrataPrep ® PCR Purification Kit. To create 

the megaprimers, a minimum of 20 bp 

upstream and downstream of the insertion 

sequence should be complementary to 

your template vector that will be used in 

the QuikChange kit reaction. Therefore, 

each oligo used in the initial PCR reaction 

will require 20 bp overhangs on the 

5’ ends. 

Using the QuikChange kit for large deletions 

is even easier since oligos can be 

synthesized, rendering the initial PCR reaction 

and fragment purification unnecessary. 

Again, 20 to 30 bp may be required 

to bind to your template DNA both upstream 

and downstream of the region that 

you want to delete. However, the largest 

oligo required should not exceed 60 bp. 

This Technical Note describes performing

a large insertion that can be completed 

in two simple steps. First, PCR-amplify 

the megaprimer. Second, add your gel 

purified megaprimer to our QuikChange 

site-directed mutagenesis kit (Figure 2). 

This strategy avoids tedious and time-consuming 

sub-cloning. With our QuikChange 

kits, you will have confidence in generating 

an error free clone while saving valuable 

time. 

Materials & Methods 

Part I. PCR reaction to generate 

the megaprimer 

Reaction and cycling conditions. 

PCR reagents: PfuUltra ® II Fusion HS DNA 

Polymerase (Stratagene) 

Thermal Cycler used: TECHNE: Touchgene 

® Gradient 

• The forward primer: GACTCCGAATCC 

CGGGAGCGCAGCGGTGAGCAAGG 

GCGAGGAG, the first 25 bp overlaps 

the target vector, the remaining primer 

sequence is complementary to EGFP. 

The primers were resuspended in 50 

mM NaCl. 

• The reverse primer: CGCGGCTGCCTC 

TGTCCCACTCTTGTACAGCTCGTCCA 

TGCC, the first 22 bp targets the vector 

and the remaining primer sequence is 

complementary to EGFP. The primers 

were resuspended in 50 mM NaCl. 

• Template for PCR reaction: 200 ng of 

plasmid DNA (pEGFP-N1; Clontech) 

• PfuUltra ® II Fusion HS DNA Polymerase 

(Stratagene) 

• PCR product was purified with 

QIAGEN’S QIAEX II Gel Extraction Kit 

• The quantity of the PCR product was 

estimated on agarose gel with Ethidium 

Bromide 

Note: The size of the PCR product is 760 

bp, which includes the insert (EGFP) of 

714 bp plus overlapping sequence of the 

target vector. 

Fig. 1 Enhanced green fluorescent protein (EGFP) 

was inserted in the open reading frame (760 bp) following 

the C-terminus of death receptor TRAIL receptor 

1/death receptor 4 (TRAIL-R1/DR4) right after the 

signal sequence. Cell staining shows the member 

protein fused with EGFP. 

1. Mutant Strand Synthesis 

Perform thermal cycling to: 

• Denature DNA template 

• Anneal mutagenic primers 

containing desired mutation 

• Extend and incorporate primers 

with high-fidelity DNA polymerase 

2. Dpn I Digestion of Template 

Digest parental methylated and 

hemimethylated DNA with Dpn I 

3. Transformation 

Transform mutated molecule into 

competant cells for nick repair 

Fig. 2. The QuikChange ® II One-Day Site-Directed 

Mutagenesis Method. 1. Mutant strand synthesis. 2. 

Dpn I Digestion of parental DNA template. 3. Transformation 

of the resulting annealed double-stranded 

nicked DNA molecules. After transformation, the XL-1 

Blue E. coli cell repairs nicks in the plasmid. 

31

Tech Notes 

Part II. The QuikChange II 

mutagenesis reaction can 

be set up using the following 

guidelines (Tables 1 

and 2). 

• Primer for the insertion 

reaction 300-500 

ng of PCR product (760 

bp EGFP) to serve as a 

megaprimer. 

Note: Follow reaction TABLE 2. Cycling Method 

with Dpn I digestion and 

transformation per the 

QuikChange 

Segment 

1 

Cycles 

1 

Temperature 

95°C 

Time 

30 seconds 

2 5 95°C 30 seconds 

52°C 1 minute 

68°C 

1 min/kb of 

plasmid* (7 min) 

3 13 95°C 30 seconds 

55°C 1 minute 

68°C 

1 min/kb of 

plasmid* (7 min) 

* For example, a 7 kb plasmid would have a 7 minute extension time 

® II manual. 

For PfuUltra ® II Fusion HS 

DNA Polymerase information 

please refer to 

the manual. 

Results 

The results of this mutagenesis 

experiment 

demonstrate that large 

PCR fragments can be successfully 

inserted into the target plasmid DNA. The 760 bp PCR fragment had at 

each end only short regions of homology where the insertion occurred. There were 

approximately 100 colonies produced. About 20 colonies were selected for further 

analysis of mutagenesis efficiency, and they were 100% positive by sequencing. 

References 

1. Geiser, M., et al. (2001) BioTechniques 31: 

88-92. 

2. Papworth, C., Braman, J., and Wright, D.A. 

(1996) Strategies 9: 3-4. 

3. Wang, W. and Malcolm, B. (1999) BioTechniques 

26: 680-682. 

4. Burke, T., et al. (1998) Oncogene 16: 1031-1040. 

5. Makarova, O., et al. (2000) BioTechniques 29: 

970-972. 

Legal 

QuikChange ® and StrataPrep ® are registered 

32 

TABLE 1. Reaction Set-up 

Amount 

Component per reaction 

Distilled water (dH 2 O) X µl 

10x QuikChange reaction buffer 5.0 µl 

dNTP mix 1 µl 

DNA template (50 ng) X µl 

Megaprimer (300 - 500 ng) X µl 

High Fidelity DNA polymerase (2.5 U/µl) 1.0 µl (2.5U) 

Total reaction volume 50 µl 

Acknowledgments 

This Technical Note was written based on data provided by Wenge Wang, M.D., 

Ph.D., in Dr. Wafik S. El-Deiry’s laboratory, Division of Hematology and Oncology in 

the Department of Medicine at the University of Pennsylvania. 

trademarks of Stratagene, an Agilent Technologies 

company. 

QIAGEN ® is a registered trademark of QIAGEN. 

Touchgene ® is the registered trademark of Techne 

Incorporated. 

Clontech © logo is the trademark of Clontech 

Laboratories, Inc. 

a. U.S. Patent Nos. 7,176,004; 7,132,265; 

7,045,328; 6,734,293; 6,489,150; 6,444,428; 

6,391,548; 6,183,997; 5,948,663; 5,932,419; 

5,866,395; 5,789,166; 5,545,552, and patents 

pending.

�� 

With GeneMorph s 

® 

II mutagenesis kits, a balanced spectrum of mutations is right at your fingertips 

Our GeneMorph ® 

II Kits include Mutazyme ® 

II 

Polymerase, which delivers a balanced mutational 

spectrum. 

Mutazyme ® 

II 

DNA Polymerase 

Mutazyme ® 

I 


Taq 


AT to N GC to N 

Value 10% 20% 30% 40% 50% 60% 70% 80% 

u.s. and canada 1-800-424-5444 x3 

europe 00800-7000-7000 

www.stratagene.com 

U.S. Patent No. 7,045,328; 6,803,216; 6,734,293; 6,489,150; 6,444,428; 6,183,997; 5,489,523 and patents pending 

GeneMorph ® 

and Mutazyme ® 

are registered trademarks of Stratagene, an Agilent Technologies Company, 

in the United States. 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

� �� 

� �� 

�� 

� �� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

��

�� 

2X Faster 

�� 

�� 

�� 

� �� 

� �� 

�� 

�� 

u.s. and canada 1-800-424-5444 x3 

europe 00800-7000-7000 

www.stratagene.com�� 

�� 

�� 

� � � �� 

�� 

�� 

�� 

QuikChange ® 

is a registered trademark of Stratagene, an Agilent Technologies Company, in the United States. 

*U.S. Patent Nos 6,734,293; 6,489,150; 6,444,428; 6,391,548; 6,183,997; 5,948,663; 

5,932,419; 5,866,395; 5,789,166; 5,545,552; 6,713,285 and patents pending. 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

� �� 

� �� 

�� 

� �� 

�� 

�� 

�� 

��

Mutagenesis Techniques - Science

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?