Birgit Claasen

The strength of the template effect attracting nucleotides to naked DNA

The transmission of genetic information relies on Watson–Crick base pairing between nucleoside phosphates and template bases in template–primer complexes. Enzyme-free primer extension is the purest form of the transmission process, without any chaperon-like effect of polymerases. This simple form of copying of sequences is intimately linked to the origin of life and provides new opportunities for reading genetic information. Here, we report the dissociation constants for complexes between (deoxy)nucleotides and template–primer complexes, as determined by nuclear magnetic resonance and the inhibitory effect of unactivated nucleotides on enzyme-free primer extension. Depending on the sequence context, Kd′s range from 280 mM for thymidine monophosphate binding to a terminal adenine of a hairpin to 2 mM for a deoxyguanosine monophosphate binding in the interior of a sequence with a neighboring strand. Combined with rate constants for the chemical step of extension and hydrolytic inactivation, our quantitative theory explains why some enzyme-free copying reactions are incomplete while others are not. For example, for GMP binding to ribonucleic acid, inhibition is a significant factor in low-yielding reactions, whereas for amino-terminal DNA hydrolysis of monomers is critical. Our results thus provide a quantitative basis for enzyme-free copying.

15N relaxation NMR studies of prolyl oligopeptidase, an 80 kDa enzyme, reveal a pre-existing equilibrium between different conformational states.

Open and closed: The characterization of protein mobility is crucial for the understanding of biological functions. We have applied NMR spectroscopy to study the conformational dynamics of the 80 kDa enzyme prolyl oligopeptidase (POP). Our results revealed that POP is highly dynamic and that inhibition of catalytic activity shifts this conformational equilibrium towards a less dynamic state.

A cost-effective labeling strategy for the NMR study of large proteins: selective 15N-labeling of the tryptophan side chains of prolyl oligopeptidase.

NMR spectroscopy is a useful tool for the study of protein structure, protein dynamics and molecular recognition processes, including both protein–protein and protein–ligand interactions. 4] However, the application of NMR experiments to large proteins remains a challenge. Transverse relaxation processes are accelerated as the size of the macromolecule grows and perdeuteration of proteins with low tumbling rates may be required. Thus, cells must be grown in D2O, [5] which, in general, reduces protein expression levels and significantly raises the cost of the NMR sample. Moreover, the assignment of spectra is limited by signal overlap, thus simplification of spectra by an appropriate selective labeling scheme is often required. 6] Selective labeling of specific amino acids can be achieved by using auxotrophic cell strains and adding the amino acid with a suitable isotope label to the medium. However, given that the biosynthetic pathways of amino acids are complex, cell growth may be limited when one of these pathways is disrupted, and this leads to lower expression levels. A less intrusive approach consists of the exploitation of the cell’s metabolic machinery to produce selectively labeled proteins and the particular choice of precursor determines whether a subset or a specific amino acid ends up labeled. Various authors have described such labeling approaches. As an example, if C-labeled a-ketobutyrate and a-ketoisovalerate are added to the growth medium, the cell then incorporates the C-label into valine, leucine and isoleucine sidechains. Similarly, addition of [2-C]or [4-C]-labeled indole to the medium allows the labeling of the tryptophan residues. Tryptophan, tyrosine and arginine are the most common amino acids in the hot spots of a given protein and are therefore involved in most of the binding energies in protein–protein and protein–ligand interactions. An inexpensive and reliable labeling strategy for the above-mentioned residues that is applicable to large proteins would be extremely useful. Here we report a cost-effective labeling strategy for the NMR study of large proteins. In this approach the N-label is selectively incorporated into the tryptophan side chains of the protein and the spectrum can be acquired without the need for deuteration. We applied this labeling to prolyl oligopeptidase (POP; EC 3.4.21.26), a serine protease of 80 kDa. In recent years, POP has gained relevance as a target for the treatment of cognitive disturbances. An array of strategies are currently being used to identify POP inhibitors, as these compounds show neuroprotective and cognition-enhancing effects in experimental animals. The X-ray structure of POP from porcine muscle revealed a distinctive two-domain structure: a catalytic domain with an a/b hydrolase fold and an unusual b-propeller domain. The costructure of POP in the presence of Z-prolyl-prolinal (ZPP), a canonical POP covalent inhibitor, shows that the specificity of the binding between the enzyme and the proline-containing inhibitor is provided by the hydrophobic interaction between POP tryptophan 595 (Trp595) and the ZPP proline ring. 16] The acquisition of a spectrum of a perdeuterated and uniformly N-labeled POP sample (U-[H,N]-POP; Figure 1 A) showed that even with perdeuteration and transverse relaxation optimized spectroscopy (TROSY), signal overlap was still a considerable handicap in cases with a large protein of 80 kDa. Therefore, a POP sample that was selectively labeled at tryptophan side chains (Trp[N-indole]-POP) was produced by supplementing the minimal growth medium with N-indole. The incorporation of the label was checked by mass spectrometry (Figure S1 in the Supporting Information). The resulting [H,N]-TROSY HSQC spectrum recorded with the Trp [NIndole]-POP sample showed a satisfactory signal-to-noise ratio (S/N) and eleven of the twelve expected tryptophan signals were present and well-dispersed (Figures 1 B, 2). It is noteworthy that despite of the high molecular weight of POP this result was obtained with a non-perdeuterated protein sample. The application of a previously described Trp[2-C-indole labeling was also evaluated. However, only few signals with poor S/N were detected in the [H,C]-aromatic TROSY spectrum acquired with a perdeuterated Trp[2-C-indole]-labeled POP sample (see SI). Thus, comparison of the results obtained for the selectively Trp[2-C-indole]and Trp[N-indole]-labeled POP shows that the Trp[N-indole] labeling is more convenient in cases with large proteins. This can be rationalized by considering the main differences between the TROSY effect of a H[a] Dr. T. Tarrag , Dr. B. Claasen, N. Kichik, Prof. E. Giralt Institute for Research in Biomedicine, Barcelona Science Park Baldiri Reixac, 10, 08028 Barcelona (Spain) Fax: (+ 34) 93-4037126 E-mail : ernest.giralt@irbbarcelona.org [b] Prof. E. Giralt Department of Organic Chemistry, University of Barcelona Mart Franqu s, 1, 08028 Barcelona (Spain) [c] Dr. R. A. Rodriguez-Mias Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood Avenue, Boston, MA 02115 (USA) [d] Dr. M. Gair NMR Facility, Scientific-Technical Services, University of Barcelona Barcelona Science Park, Baldiri Reixac, 10, 08028 Barcelona (Spain) Supporting information for this article is available on the WWW under http ://dx.doi.org/10.1002/cbic.200900575.

Chemoenzymatic Route to Oxyfunctionalized Cembranoids Facilitated by Substrate and Protein Engineering

Cembranoids constitute a large family of 14-membered oxygenated macrocyclic diterpenoids with potential as therapeutic agents. Selective late-stage oxidations of cembranoid scaffolds remain a challenge for chemical catalysts but can be accomplished by enzymes. Here, a new chemoenzymatic route to oxyfunctionalized 14-membered macrocycles including cembranoids is described. This route combines a metal-catalyzed ring-closing metathesis with a subsequent P450 BM3-catalyzed hydroxylation and delivers cembranoid-like analogues. Systematic substrate probing with a set of synthetic 14-membered macrocycles revealed that the regioselectivity of a P450 BM3-based biocatalyst increased with increasing ring rigidity as well as size and polarity of the exocyclic substituents. Enzyme regioselectivity could further be improved by first-sphere active site mutagenesis. The V78A/F87A variant catalyzed hydroxylation of cembranoid-ol (9S/R)-3 d with 90 % regioselectivity for C5 position. Extensive NMR analysis of Mosher esters and single crystal X-ray structure determination revealed a remarkable diastereoselectivity of this P450 BM3 mutant depending on substrate stereochemistry, which led exclusively to the syn-cembranoid-diols (5S,9S)-4 and (5R,9R)-4.

High‐Fidelity Recognition of RNA: Solution Structure of a DNA:RNA Hybrid Duplex with a Molecular Cap

Binding RNA targets, such as microRNAs, with high fidelity is challenging, particularly when the nucleobases to be bound are located at the terminus of the duplex between probe and target. Recently, a peptidyl chain terminating in a quinolone, called ogOA, was shown to act as a cap that enhances affinity and fidelity for RNAs, stabilizing duplexes with Watson–Crick pairing at their termini. Here we report the three‐dimensional structure of an intramolecular complex between a DNA strand featuring the ogOA cap and an RNA segment, solved by NMR and restrained torsion angle molecular dynamics. The quinolone stacks on the terminal base pair of the hybrid duplex, positioned by the peptidyl chain, whose prolinol residue induces a sharp bend between the 5′ terminus of the DNA chain and the glycine linked to the oxolinic acid residue. The structure explains why canonical base pairing is favored over hard‐to‐suppress mismatched base combinations, such as T:G and A:A, and helps to design improved high‐fidelity probes for RNA.