The Puglisi group investigates the role of RNA in cellular processes and disease. Our goal is to understand RNA function in terms of molecular structure and dynamics using a variety of biophysical and biological tools. We use nuclear magnetic resonance (NMR) spectroscopy to determine structures of biological molecules, and integrate our structural understanding into further mechanistic and functional studies. A long-term goal is to target processes involving RNA with novel therapeutic strategies. Our current research interests include:

NMR spectroscopy of large RNAs and protein-RNA complexes

Dynamics of the ribosome during translation

Molecular basis for antibiotic action against the ribosome

Mechanism of eukaryotic translation initiation and elongation

Structure and function of the Hepatitis C virus (HCV) internal ribosome entry site

RNA-protein interactions in HIV virus

• NMR spectroscopy of large RNAs and protein-RNA complexes.

Figure 1 The challenges and approaches to large RNA NMR spectroscopy. (top) Imino proton NMR spectrum at 800 MHz of a 10 kDa RNA, and a 30 kDa RNA that contains the smaller domain. Both spectral linewidth and overlap increase with increasing molecular weight. (bottom) TROSY at 18.1T leads to significant narrowing in a 1H-13C correlation experiment for adenine H2-C2 for a 25 kDa RNA.

Figure 2 Residual dipolar couplings (RDCs) allow refinement of RNA global shape. Agreement between observed and calculated RDCs for a 25 kDa RNA before and after refinement using the RDCs.

Figure 3 Refinement of a 25 kDa RNA (HCV IRES domain II) using residual dipolar couplings to improve overall RMSD. In the absence of RDC data, the calculated structures have RMSDs of more than 6.

Figure 4 Segmental Labeling of the 100 kDa HCV Internal Ribosome Entry Site (IRES) RNA. A portion of domain II (yellow) of the IRES was labeled with 15N and ligated with the rest of the IRES unlabeled. The overlap of a 1H-15N correlation experiment with that of the isolated domain II oligonucleotide show that domain II forms an independently-folded structure in the full IRES.

NMR spectroscopy has proven a powerful tool for RNA structure determination. However, intrinsic difficulties make RNA structure determination by NMR more difficult than that of proteins. RNAs contain 1/3rd fewer protons per Dalton than proteins, providing much fewer restraints for structure calculation. RNAs often form elongated structures with no long-range restraints to define the overall fold of the molecule. Finally, most biological RNAs are larger than 25 kDa, which is the reasonable size limit for protein structure determination by NMR. We are focusing on extending the use of NMR to larger RNA systems. Our approach has been to:

1. Develop robust procedures for transcription of RNAs
2. Purification methods that avoid polyacrylamide gel electrophoresis
3. Automated experimental setup using the RNAPack Varian suite of programs (written by Dr. Peter Lukavsky and George Gray). RNAPack contains the array of pulse sequences normally used for RNA structure determination.
4. TROSY methods applied to RNAs
5. residual dipolar couplings and RNA structural refinement

• Dynamics of the ribosome during translation.
Figure 5 Three-dimensional structures of the ribosomal particles must be complemented by dynamic information. Our research has initially focused on large ligand (tRNA) movements through the ribosome.

Figure 6 Example of a single-molecule FRET experiment that monitors tRNA movement on the ribosome. FRET is identified as anticorrelated changes in donor (green) and acceptor (red) single fluorophore intensities. The data show millisecond dynamics of relative tRNA positions on the ribosome.

We are using single-molecule fluorescence spectroscopy (in collaboration with the laboratory of Prof. Steve Chu, Applied Physics, Stanford) to investigate the dynamics of translation. The ribosome is a molecular machine that must perform translation with high fidelity and processivity. The structure of the ribosomal particles has revealed the molecular features of ribosome architecture. However, the dynamics of the machine remain to be determined. Single molecule fluorescence analysis avoids the need for synchronization of large numbers of ribosome, such that multiple steps of the translation cycle can be monitored, and rare events revealed. Our goals are to:

1. Understand tRNA movement and dynamics through the ribosomal A, P and E sites.
2. Reveal the dynamics of translocation.
3. Characterize the nature and rates of ribosomal domain movements.
4. Determine the mechanism of translation initiation.
5. Study the overall translation rates of different mRNAs
6. Characterize unusual translational events, such as frameshifting

• The mechanism of antibiotic action against the ribosome.

Figure 7 Structure of the antibiotic gentamicin C1a bound to its RNA target. The structure was determined using heteronuclear NMR spectroscopy. Gentamicin, an aminoglycoside antibiotic, binds within the major groove of the RNA, which is located in the decoding site of the bacterial ribosome. Aminoglycoside antibiotics cause misreading of the genetic code. Binding of the drug causes a conformational change in ribosomal RNA that dissrupts high-fidelity reading of the genetic code.

Figure 8 Structure of aminoglycoside-ribosomal RNA complexes explain antibiotic resistance mechanisms. 3D structure of the aminoglycoside paromomycin bound to the bacterial rRNA decoding site. The sites of chemical modifications to either the antibiotic or RNA that lead to resistance are highlighted with purple spheres. These are clustered at the drug-RNA interface, such that steric or electrostatic penalties to drug-RNA affinity are introduced. The N7 methylation at G1405 only causes resistance to aminoglycosides like gentamicin that contact this position directly.

The ribosome is the target of many clinically important antibiotics, such as aminoglycosides, tetracyclines, macrolides, streptogramins and oxazolidinones. We have previously focused on the aminoglycosides as a paradigm for ribosome- and RNA-directed antibiotics. Aminoglycosides bind to a conserved domain of 16S rRNA near the site of codon-anticodon interaction, and cause misreading of the genetic code. Our structural and mechanistic work revealed the structural basis for drug-RNA recogntion, the origins of antibiotic resistance and the molecular basis for toxicity. We are currently using the single-molecule fluorescence approach to understand the effect of drug binding on ribosome dynamics. Our goals are to:

1. Determine the effects of drugs on ligand and domain movements
2. Determine the mechanism of action for drugs whose mode of action is currently unclear.

• The mechanism of eukaryotic translation initiation and elongation.

Eukaryotic translation occurs by the same basic mechanism as in prokaryotic organisms. However, the processes of initiation and elongation are more complex and regulated. We are using a number of biophysical and structural approaches to characterized eukaryotic translation. These include:

1. Single-molecule fluorescence spectroscopy (with Prof. Steve Chu, Applied Physics, Stanford) to characterized the mechanism of translation initiation in yeast.
2. Single-molecule fluorescence analysis of the mechanism of translation elongation in yeast.
3. NMR structural studies of translation initiation factors, and proteins involved in cell signaling with translation as an endpoint (eg. PKR system).

• The structure and function of the Hepatitis C virus (HCV) internal ribosome entry site.
Figure 9 HCV polyprotein is translated using an internal ribosome entry site. Translation of most eukaryotic mRNA occurs in a 5cap-dependent fashion. A number of initiation factors are involved in eventual recruitment of the 40S subunit, and subsequent ATP-dependent scanning to the first start codon. In contrast, translation of HCV RNA occurs in a cap-independent manner; the 40S subunit binds directly to the IRES, and only a subset of the initiation factors are required.

Figure 10 Proteins mediate the IRES-40S subunit interaction. Our lab has identified a number of human ribosomal proteins that mediate IRES interaction with the 40S subunit. The positions of these proteins that have bacterial homologs is shown on the 30S ribosome crystal structure (left). These proteins are located on the protein-rich backside of the 40S subunit, and agree with the low resolution cryoEM study of the IRES-40S subunit solved bythe Frank and Doudna groups (shown on the right).

Figure 11 Structure of a domain of the HCV IRES RNA solved by NMR. The secondary structure of the 100 kDa IRES is shown, with domain IIb highlighted. The RNA adopts a loop E motif, with an S-turn in the backbone, as shown in the superposition of final NMR structures.

HCV is an important health problem, and infects more the 2% of the worlds population. HCV is an positive-strand RNA virus, and its RNA genome is translated immediately after uncoating of the virus in the cytoplasm of infected (liver) cells. An internal ribosome entry site (IRES) mediates the initiation of translation of the HCV genome. The 40S ribosomal subunit binds directly to the IRES RNA structure, and thus avoids the need for a 5 cap and scanning; upon IRES binding, the start codon for translation is positioned near the P site. We have been studying the mechanism and structure of the HCV IRES. We have previously delineated the domains of the IRES that are critical for 40S subunit binding, and the proteins on the surface of the 40S subunit that mediate this interaction. Our current interests include:

1. Determination of the structure of the IRES using NMR spectroscopy. The IRES is a 100 kDa RNA, and we are using novel NMR approaches, including residual dipolar couplings.
2. High resolution NMR structures of isolated IRES RNA domains.
3. Determining the conformational changes that the IRES undergoes on 40S subunit binding.
4. Understanding the mechanism of IRES-mediated initiation. What are the order of bidning events? How do initiation factors influence the process? What is the rate of initiation? How does IRES conformation affect initiation rate?

• RNA-protein interactions in HIV virus.
Figure 12 The Tat-TAR protein RNA complex regulates transcription of HIV and BIV RNAs. The TAR element is located at the 5 end of both retroviral RNAs. In HIV, the Tat protein interacts with a trinucleotide bulge, and a host cyclin interacts with a hairpin loop. In BIV, the Tat protein binds with high affinity to the RNA site, and cyclin does not interact with the hairpin looo. (right) Three-dimensional structure of the HIV Tat-TAR interaction solved by NMR. Superposition of final NMR structures of the BIV Tat-peptide-TAR RNA complex solved using 13C, 15N-labeled BIV peptide. The peptide forms a beta turn within the RNA major groove.

Figure 13 A host tRNALys3 primes reverse transcription of HIV genomic RNA. A binary complex of tRNA and genomic RNA is specifically recognized by HIV reverse transcriptase. We are studying the structure of the RNA-RNA complex using NMR.

Figure 14 HIV genomic RNA and tRNA form a high-affinity complex. Imino proton NMR spectrum of free tRNALys3 and a 1:1 complex (50 kDa total) with HIV genomic RNA.

RNA plays a criticial role in the HIV virus replication cycle. We have a longstanding interest in critical protein-RNA interactions that occur in HIV biology. First, the Tat protein is a transcriptional regulator coded by the HIV virus. It binds to a stem-loop structure at the 5-end of HIV mRNAs, and binding to the nascent transcript stimulates transcription. We have studied the HIV and Bovine immunodeficiency virus (BIV) Tat-TAR interaction using peptides that correspond to the RNA-binding domains. We have refined the structure of the BIV Tat-TAR interaction using modern heteronuclear NMR methods. The protein adopts a beta-turn conformation upon RNA binding. We are using this as a template to design small molecule inhibitors of the Tat-TAR interaction. Current interests include:

1. Design of cyclic peptide mimics of the BIV Tat peptide that recognized TAR RNA.
2. Design of cyclic peptides that interact with HIV TAR RNA.

A complex between a human tRNALys and HIV genomic RNA is the priming site for reverse transcription of the HIV RNA genome. HIV reverse transcriptase (RT) specifically recognizes this RNA-RNA complex. We are investigating the structural basis for initiation of RT using a combination of methods. The results of this work may have important ramifications for the mechanism of action of RT inhibitors. Our current work focuses on:

1. NMR structure determination of the 50 kDa tRNALys3-genomic RNA binary complex.
2. x-ray crystallographic analys of the RT initiation complex
3. The use of NMR to study the RT initiation complex.
Joseph Puglisi   Profile
Professor, Department of Structural Biology
Director, Stanford Magnetic Resonance Laboratory
e-m. puglisi\@\

Mailing Address:
Stanford University School of Medicine
D105 Fairchild Science Building
299 Campus Drive West
Stanford, CA 94305-5126
Manolia Margaris   Profile
Associate Director, Research Administration & Organizational Affairs

e-m. manolia\@\