spotlight

Published online April 21, 2006 • 10.1021/cb600147p CCC $33.50

© 2006 by American Chemical Society

116 ACS CHEMICAL BIOLOGY • VOL.1 NO.3 www.acschemicalbiology.org

The Sun Takes Its Toll Mammalian immune cells ward off infection by sensing environmental cues and switching on the proper genes for fighting off the infection. Toll-like receptors (TLRs) function early in the pathway by sensing microbial antigens. Many of the genes downstream of the TLRs have remained elusive in humans. A recent report by Liu et al. (Science 2006, 311, 1770–1773) demon-strates a new cog in the immune machine and an inter-esting link between sunlight, hormones, and immunity.

Using DNA microarrays, human genes were screened to find mRNAs that are upregulated when TLR recep-tors are activated with a ligand. Among the short list of candidates were the vitamin D receptor and Cyp27B1, an enzyme that converts the inactive vitamin D pro-hormone into the active form. This result suggested that vitamin D metabolism may bridge a gap between TLRs and the cell’s ability to kill pathogens. Indeed, when

cells infected with Mycobacterium tuberculosis were treated with active vitamin D, an antimicrobial peptide, cathelicidin, was switched on and bacterial counts decreased. Since Cyp27B1 con-verts pro-hormone to vitamin D, the authors reasoned that media levels of the pro-hormone may play a role in modulating the

immune response. They found that levels of the pro-hormone were vastly different between bovine and human sera. As a consequence, the TLR-coupled boost in mRNAs regulated by vitamin D was nearly absent in human cells cultured in bovine serum. The assay also revealed that the level of the pro-hormone in human serum from different donors is wide-ranging enough to show

dramatically different immune response. For example, serum from African-Americans, on average, had significantly less pro-hormone. Since the synthetic pathway for pro-hormone depends upon sunlight, darker skin indi-rectly casts a shadow on the TLR-dependent immune response. This may explain the higher rates of tuberculosis infection and progression among this ethnic group.

This study highlights one likely pathway among TLRs, vitamin D metabolism, and mRNAs that are expressed during an immune response. The striking pro-hormone data suggest that one’s ability to produce vitamin D may contribute to susceptibility to microbial infection. JU

SpotlightSilencing the MonkeyRNA interference (RNAi) is a gene silencing technol-ogy that has become an invaluable tool for exploring biological processes and has tantalizing therapeutic poten-tial for a variety of disease targets, especially those that have been challenging to pursue using conventional drug discovery methods. Apolipoprotein B (ApoB) is an essential protein in the regulation of cholesterol, but because it is a large, lipid-associated protein, it has been difficult to target for drug discovery. Zimmerman et al. (Nature, advance online publication, March 26, 2006, doi: 10.1038/nature04688) have demonstrated that systemic delivery of small interfering RNA (siRNA) silences ApoB in non-human primates and lowers cholesterol levels in clinically significant amounts.

While local delivery of siRNAs is well established, there are few reports of systemic delivery of siRNAs and none in non-rodent species. Systemic delivery of the APOB-specific siRNA, denoted siApoB-2, was accomplished by encapsulating siApoB-2 in stable nucleic acid lipid particles (SNALPs) and administer-ing the formulation by intravenous injection to cynomolgus monkeys. To evaluate the effectiveness of this approach, the pharmacokinetics, efficacy, and safety of a single dose of SNALP-formulated siApoB-2 was assessed. A dose-dependent reduction in liver APOB mRNA levels of up to 90% was observed 48 h after treatment, and blood levels of ApoB-100 protein, total choles-terol, and LDL were reduced by up to 78%, 62%, and 82%, respectively. Remarkably, these effects lasted through 11 days post-treatment. No effects on complement activation, delayed coagulation, pro-inflammatory cytokine production, or changes in hematology parameters were observed, indicating that the treatment was well tolerated and non-toxic.

This ground-breaking study is the first to dem-onstrate the effectiveness of systemic delivery of RNAi-mediated silencing in non-human primates, substantiating a potential new strategy for treating cholesterol-related and perhaps numerous other diseases. EG

Reprinted with permission from AAAS

Image courtesy of Getty Images

Closing the CNS Gap

Spotlight

117www.acschemicalbiology.org VOL.1 NO.3 • ACS CHEMICAL BIOLOGY

Botulinum neurotoxin A (BoNT/A) is a key ingredi-ent in the anti-wrinkle material Botox. BoNT/A is also prescribed for relief of severe pain and is a potential biological warfare agent. This toxin selectively binds to active neurons, and blocks exocytosis and neurotransmitter release. It has been proposed that BoNT/A receptors are composed of proteins and gangliosides, a group of glycosphingolipids found in the neuronal cell membrane. Now Dong et al. (ScienceExpress, published online March 16, 2006, doi:10.1126/science.1123654) show that BoNT/A enters nerve cells via the synaptic vesicle protein 2 (SV2).

The authors found that receptors for BoNT/A are localized to secretory vesicles and are exposed during exocytosis. The neurotoxin molecules interact with SV2, a conserved integral membrane glycoprotein with 12 putative transmembrane. BoNT/A bound to the largest luminal loop of three isoforms of SV2 but not to those of other synaptic vesicle proteins indicating that this neurotoxin-vesicle protein interaction was specific. Notably, glycophospholipids bound the toxin

with low affinity and promoted the formation of stable BoNT/A-SV2 complexes. These data indicate that both proteins and lipids are important for the selective entry of BoNT/A into neuronal cells.

Dong et al. also generated a loss of function cell model to determine if the cellular uptake of BoNT/A was dependent on SV2 proteins. The neurotoxin did not enter these SV2 knockout hippocampal neurons. However, entry could be restored by expressing three isoforms of the vesicle protein. Furthermore, mice lacking SV2B, one isoform of SV2, were less sus-ceptible to BoNT/A. Collectively, these data show that SV2 is the receptor for BoNT/A.

Active neurons expose more SV2 receptors during exocytosis and neurotransmitter release. The authors propose that BoNT/A, by binding SV2 and shutting down exocytosis prevents the expo-

sure of more receptors in the poisoned cell. This leaves toxin molecules free to enter other active neurons. This mechanism of action specifically and efficiently shuts down neuronal communication in the poisoned organism. EJ

Imag

e re

prin

ted

wit

h pe

rmis

sion

from

AA

AS

The Sun Takes Its Toll Mammalian immune cells ward off infection by sensing environmental cues and switching on the proper genes for fighting off the infection. Toll-like receptors (TLRs) function early in the pathway by sensing microbial antigens. Many of the genes downstream of the TLRs have remained elusive in humans. A recent report by Liu et al. (Science 2006, 311, 1770–1773) demon-strates a new cog in the immune machine and an inter-esting link between sunlight, hormones, and immunity.

Using DNA microarrays, human genes were screened to find mRNAs that are upregulated when TLR recep-tors are activated with a ligand. Among the short list of candidates were the vitamin D receptor and Cyp27B1, an enzyme that converts the inactive vitamin D pro-hormone into the active form. This result suggested that vitamin D metabolism may bridge a gap between TLRs and the cell’s ability to kill pathogens. Indeed, when

cells infected with Mycobacterium tuberculosis were treated with active vitamin D, an antimicrobial peptide, cathelicidin, was switched on and bacterial counts decreased. Since Cyp27B1 con-verts pro-hormone to vitamin D, the authors reasoned that media levels of the pro-hormone may play a role in modulating the

immune response. They found that levels of the pro-hormone were vastly different between bovine and human sera. As a consequence, the TLR-coupled boost in mRNAs regulated by vitamin D was nearly absent in human cells cultured in bovine serum. The assay also revealed that the level of the pro-hormone in human serum from different donors is wide-ranging enough to show

dramatically different immune response. For example, serum from African-Americans, on average, had significantly less pro-hormone. Since the synthetic pathway for pro-hormone depends upon sunlight, darker skin indi-rectly casts a shadow on the TLR-dependent immune response. This may explain the higher rates of tuberculosis infection and progression among this ethnic group.

This study highlights one likely pathway among TLRs, vitamin D metabolism, and mRNAs that are expressed during an immune response. The striking pro-hormone data suggest that one’s ability to produce vitamin D may contribute to susceptibility to microbial infection. JU

Central nervous system (CNS) injuries often result in tissue damage that does not heal properly. Several factors contrib-ute to inadequate repair of CNS tissue, including scar tissue formation, gaps in nervous tissue from phagocytosis of dying cells, and failure of neurons to initiate axonal extension. Ellis-Behnke et al. (PNAS 2006, 103, 5054–5059) have now developed novel nanobiomedical materials, called self-assembling peptide nanofiber scaffolds (SAPNS), that pro-mote regeneration of axons when injected at sites of CNS trauma.

SAPNS consist of alternating positive and negative L-amino acids that form highly hydrated scaffolds in physiological solutions. Use of SAPNS as biomaterials provides several advantages over cur-rently available polymer biomaterials. First, the individual nanofibers are approximately 10 nm in diameter, which

is similar in scale to the native extra-cellular matrix and thus may facilitate cell growth, migration, and differentiation. In addition, degradation of SAPNS leaves natural L-amino acids that can poten-tially be used by the surrounding tissue. Finally, because SAPNS are synthetic materials and appear to be immunologically inert, biological contamination and tissue rejection issues prevalent in other biomaterials are avoided.

To test the ability of the SAPNS to promote CNS wound healing, the researchers generated a lesion in the optic tract of the hamster midbrain, resulting

in a tissue gap and loss of vision. When RADA-1, a SAPNS that is known to support neuronal growth and development, was injected at the site of injury, significant axonal growth through the tissue gap was

observed and, remarkably, vision was restored. The authors hypoth-esize that interac-tion of SAPNS with the extracellular matrix promotes cell migration into the lesion area and creates an environ-ment conducive to

axon regeneration. By overcoming some of the obstacles in CNS wound healing previously thought to be impenetrable, this groundbreaking work launches a new method to treat CNS trauma. EG

Imag

e re

prin

ted

wit

h pe

rmis

sion

from

PN

AS.

Botox on the Brain

Fortuitous Promiscuity

Spotlight


Labeling reagents for visualizing biological interactions come in all shapes and sizes. Proteins, peptides, small molecules, and inorganic nanoparticles have all been used to label proteins, but each method must tackle many issues, including labeling specificity, speed, stability, tag size, and toxic-ity. In addition, it has been especially difficult to target specific proteins in the context of living cells. Lin et al. (JACS 2006, 128, 4542–4543) have developed a method for labeling proteins in vitro and on the surface of live cells using the enzyme trans-glutaminase.

Transglutaminases (TGases) cata-lyze amide bond formation between glutamine (Q) and lysine residues, serving to cross-link proteins in diverse cellular processes including

cell migration and apoptosis. The authors exploit this activity using guinea pig TGase (gpTGase), which is specific for its glutamine-containing substrate but can be quite promis-cuous for its amine-containing substrate. This promiscuity allows for a variety of amine-containing labeling reagents, such as biotin or fluorescein derivatives, to be ligated to peptide substrates of gpTGase (called “Q-tags”) genetically fused to the target protein of interest.

To demonstrate the utility of this approach in live cells, two proteins, cyan fluorescent protein and the epidermal growth factor receptor, were Q-tagged, expressed on the surface of HeLa cells, and labeled. The labeling was demonstrated to be site-specific, enzyme-dependent, and non-toxic and did not affect the

activity of the proteins. In addition, this approach was used to attach a photoaffinity probe to the NFkB transcription factor p50 in vitro. This allowed the researchers to examine the effects of reagents other than DNA on the homodimerization of p50, a distinct advantage over the currently favored gel-shift assay. Though there are disadvantages to using gpTGase, such as the inability to label intracellular proteins and the fact that gpTGase specificity for the Q-tag is lower than that of other enzymes (such as biotin ligase) for their peptide substrates, this work opens the door to using transglutaminase-introduced probes to investigate protein trafficking, protein–protein interactions, and protein conformational changes in the cellular context. EG

New Light on the Lightning BugFireflies don’t just fascinate six year olds. Scientists have shown that their bright yellow light allows the insects to communicate, whether to attract members of the opposite sex or to warn predators to stay away. The biochemistry behind the bioluminescence comes from oxidation of the small molecule luciferin by the enzyme luciferase; light is emitted upon relaxation of excited oxyluci-ferin to its ground state. Remarkably, a single amino acid change in luciferase, serine 286 to asparagine (S286N), changes the emission color from yellow-green to red, but the molecular basis for the color change has not been clear. Now, Nakatsu et al.

(Nature 2006, 440, 372–376) have uncovered a structural explanation for this fascinating phenomenon. The researchers use X-ray crystallography to investigate structural differences between wild-type and mutant luciferase

enzymes. Three crystal structures of wild-type luciferase, the first complexed with its natural reactant, the second with a stable analog of the high energy intermediate, 5´-O-[N-(dehydroluciferyl-)-

sulfamoyl]adenosine (DLSA), and the third with its natural products, provided snapshots of the reaction pathway; changes in luciferase structure reflect movement of amino acids as the reac-tion progresses. Although the structures with the reactant and the products were essentially identical, the DLSA structure revealed significant movement of isoleucine 288, close to serine

286, creating a hydrophobic pocket in the active site. The authors suggest that the hydrophobic environment prevents energy loss from the excited state of oxyluciferin, resulting in yellow-green

bioluminescence.To investigate the basis for the color change, the structure of luciferase (S286N) complexed with DLSA

was compared with that of the wild-type enzyme. The mutant enzyme does not exhibit the conformational changes seen in the wild-type system, and thus no hydrophobic pocket is created. The authors hypothesize that this results in energy loss from the excited state, manifested by lower energy red light emission. EG

Image courtesy of Getty images

Resisting the Resistance

Protein Capture… Without a TraceProtein microarrays are an attractive method for investigating protein inter-actions due to the small amount of protein required and the large number of interactions that can be probed on a single glass microscope slide. However, methods for attaching pro-teins to the microarray surface suffer from technological challenges. Non-covalent attachment methods leave proteins attached at random orienta-tions or may not be resilient enough to last through subsequent assays, and covalent attachment strategies can be inefficient or leave capture reagents attached to the surface. Kwon et al. (Angew. Chem., Int. Ed. 2006, 45, 1726–1729) have developed a traceless capture ligand approach for creating protein microarrays.

The traceless capture method is based on protein trans-splicing, a naturally occurring process wherein a segment of a protein, called an intein, excises itself and rejoins the remaining portions. The authors use the DnaE intein from cyanobacteria,

which contains two fragments, the N-intein (IN ) and the C-intein (IC ), that spontaneously self-assemble to form a functional splicing domain. In the study, the IN fragment was fused at the DNA level to the C-termini of two proteins, maltose binding protein and enhanced green fluorescent protein, and the IC fragment was immobilized onto a glass slide. Spotting of the protein solutions on the glass slide should allow for the intein frag-ments to interact, ligate the protein to the slide, and splice themselves out into solution. Indeed, detection of the proteins was dependent on

the presence of IN in the protein and IC on the slide, suggesting that the proteins covalently attached to the surface by this mechanism. In addi-tion, the authors demonstrated that IN-containing fusion proteins from soluble cellular fractions of protein expressed in E. coli and from crude in vitro transcription/translation reactions could be immobilized using this approach, expanding its utility to proteins in biological mixtures. EG

Spotlight


Reprinted with permission from Angewandte Chemie, International Edition

Antibiotic resistance is a growing health concern across the globe. Many bac-teria acquire resistance through genetic mutations that circumvent the antibiotic’s mechanism of action. Antibiotics that possess multiple modes of action are therefore more likely to resist resistance. Nisin, a common food preservative and member of a group of antimicrobial peptides called lantibiotics, utilizes several mechanisms to kill bacteria, including forming pores in the mem-brane, disrupting cell wall biosynthesis, and inhibiting the outgrowth of bacterial spores. Nisin A contains five cyclic

thioethers, called lanthionines and methyllanthionines, that are created by addition of cysteine to dehydroalanine and dehydro-butyrine. It is believed that the enzyme NisC is responsible for formation of the rings, but the cyclase activity of NisC has not previously been demonstrated. Li et al. (Science 2006, 311, 1464–1467) report the in vitro enzymatic synthesis of nisin A and the crystal structure of NisC. Their findings confirm the cyclase activity of NisC and provide insight into the mechanism of

biosynthesis of this important antibiotic.Sequencing of the nisin gene clus-

ter several years ago revealed that the

(continued on page 120)

Imag

e re

prin

ted

wit

h pe

rmis

sion

from

AA

AS

.

Spotlight


The sequences found at each end of a typical eukaryotic mRNA lack protein-coding potential and instead play important roles in gene regulation. Recently, these untranslated regions (UTRs) have garnered extra attention as more sequence motifs are implicated in

RNA stability and the efficiency of translation. A new study by Baben-dure et al. (RNA, published online March 15, 2006, doi: 10.1261/rna.2309906) takes on an old UTR question: how does RNA secondary struc-

ture at the 5′ end of an RNA affect the translation efficiency of the message?

The authors used a remarkably simple cell-based assay where

translation efficiency is monitored by a ratio of two different colors of fluorescent protein. Red fluor-escent protein was encoded after a standard 5′ UTR, while the UTR of green fluorescent protein (GFP) was flavored with various hairpins of RNA secondary structure. These RNA elements were engineered between the GFP mRNA transcrip-tion start site and the start codon for protein synthesis. Structural elements that reduced transla-tion efficiency in living cells were identified by a change in the red to green signal ratio. Both the stabil-ity of the inserted RNA helix and its position could be easily altered and assayed. Hairpin RNAs at the extreme 5′ end of the RNA had the most profound effect on transla-tion. Also, a surprisingly sharp drop in translation occurred as the stability of the hairpin pushed into the –25 kcal mol-1 range. Some of the rules developed from this

assay stand in contrast to earlier studies with in vitro translation systems, thus highlighting the need to caution and comparison. These synthetic hairpin elements were compared with natural UTR sequences that have strong predicted secondary structures near the 5′ end of the mRNA. These natural 5′ UTRs also imparted lower translational efficiency.

This paper represents a rigorous next step toward understanding yet another RNA-mediated control mechanism. In addition to second-ary stuctures, ligand-induced RNA tertiary structures such as those found in bacterial riboswitches may also regulate protein synthesis in eukaryotes. Finally, the authors suggest that small engineered RNAs might be used to alter the structural character of natural or designer UTRs, thereby providing a mechanism for cells to tinker with translation. JU

Resisting the Resistance, continued

Untranslated Roadblock

peptide precursor to nisin A is synthe-sized via translation and then post-translationally modified to contain the cyclic thioethers. However, the in vitro reconstitution of nisin A, as well as that of other lantibiotics, has been challeng-ing due to their structural complexity and the requirement for access to NisC and its substrate. To obtain NisC, the authors cloned the enzyme from Lactococcus lactis and expressed it in Escherichia coli. In order to generate the NisC substrate, an L. lactis strain was engineered to contain the genes that code for the

precursor peptide, the dehydratase that creates dehydroalanine and dehydrobu-tyrine, and the transporter that secretes the dehydrated peptide. Treatment of the secreted dehydrated peptide with NisC, followed by removal of the leader peptide with trypsin, generated active nisin A, as shown by a clear zone of inhibition when spotted on a lawn of an indicator strain.

The crystal structure of NisC was deter-mined and found to contain an a-toroidal core structure that is topologically similar to bacterial cellulases, endoglucanase, terpenoid cyclases, and the b subunit of

protein farnesyl transferase. In addi-tion, a zinc ion is strategically located in the center of the toroid bound by two conserved cysteines and a histidine, where it assists activation of the nisin cysteine thiol for intramolecular nucleo-philic attack on the dehydroalanine or dehydrobutyrine. This structure provides insights into the possible roles, such as posttranslational modification of cysteine residues, of structurally similar proteins found in a variety of organisms including humans but whose functions are currently unknown. EG

Imag

e co

urte

sy o

f Get

ty im

ages

Spotlight


RNA interference (RNAi),

where specific RNA

fragments suppress

gene expression and

result in loss of func-

tion phenotypes, has

revolutionized the

exploration of gene

function in mammals. In

an effort to adapt RNAi

technology to genome-

wide exploration, The

RNAi Consortium (TRC) has been

created to generate genome-scale

short hairpin RNA (shRNA) librar-

ies and develop high throughput

screening methods to characterize

these libraries in mammalian cells.

To this end, Moffat et al. (Cell 2006,

124, 1283–1298) have created

the initial portion of a lentiviral

shRNA library, denoted TRC1, and

screened a subset of this library in

human colon cancer cells.

TRC1 contains 100,000 arrayed

shRNA constructs that target

22,000 human and mouse genes.

Several challenges in high through-

put RNAi screening were addressed

in the design and screening of

TRC1. First, introduction of shRNA

into cells was accomplished

using lentiviral vectors, which can

transduce a wide range of cell types

including primary and nondividing

cells. Second, variable effective-

ness of different shRNA constructs

and off-transcripts effects were

limited by including five shRNA

constructs per gene and requiring

that at least two shRNAs induce a

similar phenotype in order for the

effects of the gene to

be considered. Finally,

a semiautomated

protocol was developed

to adapt the library to

high throughput pheno-

typic assays.

As a test of the utility

of TRC1, a subset of the

library was screened

to identify regula-

tors of mitosis. One

hundred genes were identified that

significantly altered the mitotic

index in human colon cancer cells.

Further experiments revealed that

although some of these genes

also affected mitosis in human

fibroblasts, others did not, point-

ing to potential cancer targets.

The TRC1 library is an initial

realization of an effort to anno-

tate the genome using RNAi.

Information generated from this

project will contribute to a global

understanding of the genetics

of biological processes. EG

First Annual Meeting Japanese Society for Chemical BiologyMay 8–9, 2006 Tokyo, Japan

Yale Chemical Biology Symposium May 12, 2006New Haven, CT

71st Symposium: Regulatory RNAs May 31–June 5, 2006 Cold Spring Harbor Laboratory, NY

UPCoMING CoNFERENCES

Spotlights written by Eva Gordon, Evelyn Jabri, and Jason Underwood

Repr

inte

d w

ith

perm

issi

on fr

om C

ell

Genome-Wide Interference

spotlight

Documents