spotlight

Published online March 17, 2006 • 10.1021/cb600107t CCC $33.50 © 2006 by American Chemical Society

Spotlight

58 ACS CHEMICAL BIOLOGY • VOL.1 NO.2 www.acschemicalbiology.org

Lithium, a common treatment for bipolar disorder, carries a number of side effects. Sleep disturbance is among the most common. Sleep patterns are controlled by circadian rhythm, the 24-hour internal clock in organisms. These rhythms, which are also displayed by cells, are maintained at the mole-cular level through interconnected transcriptional feedback loops of clock genes. Post-translational modification also plays a role in regulation of the cellular clocks. Now, Yin et al. (Science 2006, 311, 1002–1005) suggest a molecular link between lithium, a kinase, an orphan nuclear receptor, and the clock genes.

The orphan nuclear receptor Rev-erbα is responsible for modulating the transcription and rhythmic expression of Bmal1, one of many clock proteins. Previous studies showed that Rev-erbα turns down its own transcription and that of Bmal1 during circadian night. Rev-erbα contains several potential phosphorylation sites for glycogen synthase kinase 3β (GSK3β). Interestingly, this kinase is selectively inhibited by lithium ions and a mutation found in the fly homologue of GSK3β affects circadian timing. To test if this

kinase modulates the activity of Rev-erbα, the authors depleted GSK3β from tissue culture cells using RNA interference. To their surprise, depletion of the kinase resulted in

the disappearance of Rev-erbα as well. Since a number of timing events in the circadian circuit

are regulated by positive and negative feedback on transcription, the mRNA levels of both Rev-erbα and master clock regulator Bmal1 were tested. When GSK3β

is depleted, the mRNA levels of both regulators increased. These data suggested that GSK3β activity

was modulating the amount of Rev-erbα through post-translational modification. To test this hypothesis, the authors inhibited GSK3β kinase activity with lithium at concentrations that mimic patient serum levels. They found very little Rev-erbα protein suggesting that

it was being degraded. The authors show that ubiquitin-mediated proteasomal degradation of Rev-erbα is modulated by two key serine residues which are GSK3β phosphorylation sites. These and other data indicate that the kinase activity of GSK3β is critical for stabilizing Rev-erbα and explain how lithium inhibition of a kinase alters the circadian clock. JU

Trigger PointsMany age-related degenerative diseases, including Alzheimer’s, and diseases connected with medical therapy, such as dialysis-related amyloidosis (DRA), are associated with the oligomerization of protein into amyloid fibrils. The constituents of these amyloids vary widely, but the fibrils share common structural properties. In vitro studies indicate that fibril formation begins with a nucleation event that is followed by a rapid and cooperative

association. As the amyloid product is a heterogeneous aggregate, the molecular basis for these changes are, to date, poorly understood. Now, Eakin et al. (Nat. Struct. Mol. Biol. 2006; 13, 202–208) determine the chemical basis and resultant structural changes required for oligomerization of β-2-microglobulin (β2m) in DRA.

β2m is a subunit necessary for the cell surface expression of class I-like complexes such as the major histocompatibility

complex (MHC). As part of MHC turnover, β2m is released and subsequently broken down by the kidney. The concentration of β2m is higher in patients with kidney disease, but this increased level of the protein by itself does not result in aggre-gation. Under physiological conditions, the transient pres-ence of stoichiometric Cu2+ is required for β2m fibrillogenesis.

The authors show that Cu2+ catalyzes the formation of

Clockwork Lithium

(continued on page 59)

Steve Campbell/Getty Images

SpotlightCollagen, which is Greek for “glue producer”, is the most abundant protein in humans. True to its name, collagen literally holds the human body together as the major component of connective tissue. It also contributes to skin strength and elasticity and has a greater tensile strength than steel. Collagen’s many unique properties have stimulated attempts to develop materials for medicinal and nanotechnological applica-tions. However, natural collagen is difficult to modify and can cause pathological side effects when transplanted into humans. Furthermore, generation of synthetic collagen has proved difficult to control. Now Kotch and Raines (Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 3028–3033) have used synthetic chemistry to create unique frag-ments that self-assemble into collagen.

Natural collagen is comprised of three polypeptide strands that fold into a triple

helix, with each strand containing a repeating tripeptide sequence. The researchers generated synthetic peptides containing these tripeptide sequences along with strategically placed cysteines to enable covalent linkage of the strands through disulfide bonds. The pep-tides were designed so that the locations of the cysteines resulted in “sticky ends” that enabled self-assembly of collagen-like triple helices. The authors used a variety of techniques to characterize the syn-thetic collagen triple helices. Circular dichro-

ism spectra confirmed a triple helical structure, and thermal stability experi-ments demonstrated

cooperative denaturation,

characteristic of triple helices. Dynamic light scattering,

atomic force microscopy, and transmission electron micros-

copy provided information about the hydrodynamic radius and length of the

fibrils, which had a width near 1 nm and a length approaching 1 µm,

depending on amino acid composition, temperature,

and solvent. The length of natural collagen is approximately

300 nm; thus this “sticky-ends” method for producing synthetic collagen

is capable of producing strands longer than those of natural collagen. This unique approach presents intriguing opportunities for creating collagen-based materials for diverse applications, ranging from tissue engineering to providing nanowires for electronic devices. EG

59www.acschemicalbiology.org VOL.1 NO.2 • ACS CHEMICAL BIOLOGY

Designer Collagen

M*, an alternative native-like conformation of the protein. This intermediate assembles into dimeric building blocks that are stabilized by Cu2+. The authors hypothesized that divalent cation–mediated backbone isomerization of one proline residue was responsible for M* formation. To test this hypothesis, Eakin et al. mutated the proline to alanine trapping this residue

in a trans conform-ation. This mutant bound Cu2+ 10 000

fold more tightly than WT β2m. The

crystal structure showed that, like WT β2m,

the mutant adopts a β-sandwich structure but its hydrophobic core is repacked. These and other structural

rearrangements result in the formation of an amphipathic dimer. The dimeric structure reveals a basis for its role as

an intermediate building block for the formation of higher oligomeric states.

Metal ions are implicated in other amyloid diseases. The proteins in these systems, like β2m, show cation-triggered conformational transitions possibly involving backbone rearrangements. Understanding the relationships between these metal-triggered transitions and fibril formation will be valuable in understanding amyloid diseases. EJ

Trigger Points, continued

© 2006 N

ational A

cademy of

Science

s, U.S

.A.

Repr

inte

d w

ith

perm

issi

on fr

om N

SMB

Some anticancer agents work by inducing cancer cells to undergo apop-tosis or commit suicide. Cancer cells unfortunately are notorious for acquir-ing resistance to apoptosis-inducing drugs, including the DNA topoisomer-ase II-based DNA-damaging agent doxorubicin. Smukste et al. (Cancer Cell 2006, 9, 133–146) have developed an assay to identify small molecules that boost the toxicity of doxorubicin in resistant cancer cells.

Human papillomavirus (HPV), the major causative factor in cervical cancer, produces an oncogenic protein called E6. This protein increases the resistance of cancerous cells to antican-cer drugs through various mechanisms. E6 forms a complex with E6 Associated Protein, an E3 ligase, and induces p53 ubiquitination and degradation. This allows tumor cells to become resis-tant to apoptosis. E6 also causes the degradation of a proapoptotic protein BAK and activates telomerase, which extends the life of cancer cells. To identify small molecules that overcome

E6-mediated drug resistance, the researchers used an E6-expressing colon carcinoma cell line called RKO-E6 that is two to four times more resistant to doxo-rubicin-mediated apop-tosis than RKO cells alone. Approximately 27 000 small molecules were screened for their ability to increase the toxicity of doxorubicin, and 88 com-pounds were identified that selectively increase doxorubicin’s lethality by at least 20%. The compounds were grouped into five structural classes, including quaternary ammonium com-pounds, protein synthesis inhibitors, 11-deoxyprostaglandin E1 analogues, 1,3-bis(4-morpholinylmethyl)-2-imidaz-olidinethione analogues, and a collec-tion that the authors named indoxins.

Examination of the mechanisms by which the compounds exerted their effects revealed that doxorubicin resistance is tied to regulation of its

target, topoisomer-ase IIα. Many of the compounds were found either to increase cellular topoisomerase IIα levels or to induce S phase arrest. The

latter could affect topoisomerase IIα transcription. The indoxins were especially interesting because they upregulated topoisomerase IIα and caused S

phase arrest, increasing the sensitivity of resistant cells to these compounds. Photoaffinity pull-down experiments revealed that the molecular target of the indoxins is a nuclear actin-related protein complex involving myosin 1c and ARP2, proteins that are known to be involved in transcriptional regulation. The mechanisms of action of the compounds identified in this screen shed light on pathways of doxorubicin resistance in cancer, and this knowledge could lead to improved treatment strategies for drug-resistant cancers. EG

Spotlight


How do cells know where to end tran-scription of a gene, and what do they do when the transcription process stalls? Transcription termination is the answer, and E. coli have evolved a number of different mechanisms to accomplish this task. Termination can be accomplished by an intrinsic mechanism relying on the formation of a hairpin structure within the RNA molecule. Alternatively termina-tion can rely on an enzymatic process involving either the Rho terminator factor or the Mfd protein, which is important for removing stalled polymerases. Parks and Roberts (Proc. Natl. Acad. Sci U.S.A.; doi:10.1073/pnas.0600145103) have explored the similarities between these termination mechanisms and have found

there are common elements in each of them, all involving the transcription bubble.

The transcription bubble is the region of unwound DNA in which RNA polymerase elongates the new RNA molecule. Termination and release of the RNA involve the rewinding of the DNA duplex within this region and the unwinding of the DNA/RNA hybrid. Intrinsic RNA termination requires the formation of a hairpin structure that disrupts the RNA/DNA hybrid. This disruption allows the DNA transcription bubble to be partially reformed, favoring dissociation of the RNA/DNA hybrid. Parks and Roberts examined whether Rho-mediated or Mfd-mediated

termination also involved the rewinding of the transcription bubble. They show that transcription bubbles prevented from reannealing by incorporation of mismatches within the bubble region are less efficient at releasing transcripts in both Rho-dependent and Mfd-dependent termination. They further show that both Rho and Mfd translocate the transcription bubble downstream, effectively moving the transcription bubble and further favoring dissociation of the RNA from the DNA template. These results indicate that all known mechanisms of termination involves the remodeling of the transcription bubble, and yield a clearer picture of the mechanism of prokaryotic transcription termination. ST

Assisted Suicide

Reforming Bubbles

Imag

e co

urte

sy o

f B. S

tock

wel

l

Capturing the CapsidMany viruses, including hepatitis B virus (HBV), West Nile

virus, and HIV, have macromolecular protein cores, or capsids,

that contain the viral genome. Assembly of the core is an

integral part of the viral life cycle, has no counterpart in

the infected cell, and may be particularly sensitive to small

changes in local conformation, making it an attractive target

for antiviral therapy. However, capsid assembly has not been

extensively pursued as a drug target, in part due to difficulty

in adapting relevant assays for high throughput screening.

Stray et al. (Nat. Biotechnol. 2006, 24, 358–362) has devel-

oped an in vitro assay for HBV capsid assembly, facilitating

the search for small molecule inhibitors. This assay can be

readily adapted to other virus assembly systems (such as

avian influenza).

In HBV, the icosahedral core consists of 240 copies of

capsid protein that enclose the viral DNA and reverse tran-

scriptase. In vitro, assembly of the capsid protein into a core-

like particle is dependent on many factors, such as protein

concentration, pH, and ionic conditions, and is nucleated

by a trimer of protein dimers. This assembly process results

in close proximity of the C termini of the capsid proteins,

suggesting that fluorescence quenching could be employed

to monitor assembly progression. The researchers gener-

ated a fluorescent capsid protein, termed C150BO. This

modified protein was highly fluorescent prior to assembly,

but fluorescence dropped markedly upon capsid assembly

due to fluorescence quenching. Two small molecules known

to disrupt capsid assembly, urea and a heteroaryldihydro-

pyrimidine (HAP-1), were used to determine whether pertur-

bation of capsid assembly could be detected using C150BO.

When HAP-1, which is known to accelerate capsid assembly

and nucleate the formation of irregular particles, was added

to the C150BO assembly reactions, a decrease in fluorescence

was observed. This result suggests that HAP-1 increased

the rate and extent of C150BO assembly as it does with

wild-type capsid protein. Unlike HAP-1, urea inhibits capsid

assembly. When it was added to C150BO assembly reactions,

an increase in fluorescence was observed, confirming the

decrease in assembly. These results indicate the C150BO

assay can identify small molecule inhibitors of virus assembly.

Furthermore, this assay is amenable to a high throughput

format, providing a novel method for discovering small

molecule antiviral agents that misdirect capsid formation to

yield aberrant and noninfectious virus particles. EG

Spotlight


Image courtesy of A. Zlotnick

Spotlight


Several neurodegenerative disorders,

such as Huntington’s, Alzheimer’s, Parkin-

son’s, and amyotrophic lateral sclerosis

(Lou Gehrig’s disease), are associated

with the misfolding of proteins. In some

of these disorders, protein misfolding

and aggregation are caused by expan-

sion of polyglutamine (polyQ) residues

in specific proteins, but the link between

aggregation and disease pathogenesis

is not clearly defined. Gidalevitz et al.

(Science, 2006, 311, 1471–1474.) have

uncovered a connection between polyQ

expansion and cellular

malfunction that helps

explain the role of

aberrant protein folding

in neurodegenerative

conditions.

The researchers

employed genetic methods using

Caenorhabditis elegans (C. elegans)

temperature-sensitive (ts) mutations

to explore the effects

of the presence of an

aggregation-prone pro-

tein on cell homeostasis.

Temperature-sensitive

mutant proteins are dependent on the

folding environment, so the extent of

aggregation versus folded protein in the

ts mutant worms serves as an indicator of

the state of the protein-folding quality-

control system. C. elegans strains contain-

ing either nonaggregating or aggrega-

tion-prone polyQ proteins were crossed

with several structurally

and functionally unrelated

ts mutants. The authors

observed that the presence

of an aggregation-prone

polyQ protein, but not the

nonaggregating protein,

was sufficient for the worms to exhibit

the mutant phenotype at the permissive

condition. These observations indicate

that expression of the

aggregation-prone polyQ

protein generally interferes

with the protein-folding

capacity of the cell.

The results are surprising considering

that normally the presence of misfolded

protein activates a stress response

that increases protein refolding. The

researchers hypothesize that the chronic

presence of aggregation-prone proteins,

while insufficient to activate the stress

response, initiates a positive feedback

mechanism that enhances disruption of

protein folding. Over time, the gradual

accumulation of misfolded and damaged

proteins compromises cellular function

and results in a disease state. This work

puts forward a general mechanism for the

catastrophic effects of protein misfolding,

and further discernment of this process

could lead to new therapeutic strategies

to treat neurodegenerative disorders. EG

Misfortunes of Misfolding

Is data coming at you faster than you can catalogue it? You’re not alone but don’t despair. A group of RNA scientists, the RNA Ontology Consortium (ROC) is developing an RNA Ontology (RO) (RNA 2006, published online February 16, 2006, 10.1261/rna.2343206). According to Wikipedia, ontology has one basic question: “What actually exists?” For the RNA scientists what exist are the kinds, structures, and sequences of RNA, and the events and processes in which they are involved. RNA Ontology is an initiative to create a taxonomy of all this information along with

the methodologies used to examine and interpret it. RO also aims to create software to bring computational tools to the bench scientists and to make possible pre-cise searches for all RNA information. With this aggre-gate of taxonomies, tools, and common vocabulary, scientists with different backgrounds, experiences, and interests can discuss their favorite RNA using a mean-ingful, mutually intelligible dialect. The ROC invites those interested in RNA to contribute to the on-going discussion of RNA ontology through their Community Discussion Board (http://roc.bgsu.edu) GM

Integrating Your RNA

Repr

inte

d w

ith

perm

issi

on fr

om A

AA

S.

Repr

inte

d w

ith

perm

issi

on fr

om A

AA

S.

Spotlight


Distinguishing Cathepsins Cathepsins are cysteine proteases whose expres-sion levels and activity are increased in human and mouse cancer cells. In some tumors, cathepsins are also mislocalized to the cell surface and can be secreted into the extracellular space. There are 11 cathepsin family members but their individual functions in cancer and normal cells are not clear. In a recent study, Gocheva et al. (Genes Dev. 2006, 20, 543–556) used a genetic approach to define the roles of four cathepsin genes in cancer cells.

The author generated knockouts of four different cathepsin genes in cancer bearing mice and examined the phenotypes of the mutants. Mutations in cathepsins B and S impaired the formation of tumors and angiogenesis. Removing cathepsin B, L or S resulted in reduced tumor growth. Absence of any one of these three cathepsins impaired tumor invasion. In contrast, removing cathepsin C had no significant effect on the tumor parameters examined. Further analysis showed that cathepsin B, L, and S cleaved E-cadherin, a protein involved in cell adhesion, in vitro, but cathepsin C did not. In mice lacking cathepsin B, L, and S, the levels of E-cadherin levels are maintained. The authors propose that the cathepsin B, L, and S-mediated tumor invasion is likely due to loss of cell adhesion resulting from the cleavage of E-cadherin. The molecular mechanism by which each cathepsin mediates other aspects of tumorigensis requires additional studies. However, the data from Gocheva et al. may be useful in designing inhibitors that target specific cathepsins and thereby specific stages of cancer progression. EJ

A Life or Death DecisionCellular response to the proinflam-matory cytokine tumor necrosis factor (TNF) can result in cell survival or cell death, depending on the circumstances. TNF activates the c-Jun NH2-terminal kinase (JNK) signaling pathway, but understanding how JNK determines the cell’s destiny requires deciphering its apparent dual role in mediating both life and death. TNF-induced activation of JNK is characterized by an early, robust but transient phase followed by a late, low-level but sustained period. Ventura et al. (Mol. Cell 2006, 21, 701–710) have used chemical genetics to investigate how the time course of JNK activity affects the fate of the cell.

The redundancy of kinases in the cell has hampered the ability to find specific small molecule inhibitors to probe kinase function, and JNK is no exception. To circumvent this problem, the researchers created mutant JNK (m-JNK) containing an enlarged ATP binding pocket and an inhibitor, 1-naphthylmethyl-4-amino-1-tert-butyl-3-(p-methyl-phenyl)pyrazolo[3,4-d]pyrimidine (1NM-PP1), that binds specifically to this pocket but not to wild-type JNK or other kinases. The researchers also created murine embryonic fibroblasts (MEFs) that express wild-type JNK, no JNK, or m-JNK to facilitate interpretation of the specific function of JNK in the cell.

The temporal role of JNK in apoptosis and survival pathways was investigated using these unique biological tools. The authors found that JNK activity during the late phase of JNK signaling mediates proapoptotic signaling. In contrast, it was observed that the early phase of TNF-induced JNK activation is critical for signaling survival. These data explain the apparent paradoxical roles of JNK in the cell and indicate that a cell’s fate is critically dependent on the temporal regulation of this kinase. The details of how JNK controls each signaling process will require additional mechanistic studies. EG

Channels, Receptors and SynapsesCold Spring Harbor Meeting April 18–21, 2006 Cold Spring Harbor Laboratory, NY

Molecular Chaperones and the Heat Shock Response May 3–7, 2006 Cold Spring Harbor Laboratory, NY

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

RNA 2006Annual Meeting of the RNA SocietyJune 20–25, 2006Seattle, WA

UPCOMING CONFeReNCeS

Spotlights written by Eva Gordon, Evelyn Jabri, Grace Miller, Sarah Tegen, and Jason Underwood.

Image courtesy of R. Davis

spotlight

Documents