Structure-based design of a bispecific receptor mimic that inhibits T cell responses to a superantigen.

Key surface proteins of pathogens and their toxins bind to the host cell receptors in a manner that is quite different from the way the natural ligands bind to the same receptors and direct normal cellular responses. Here we describe a novel strategy for "non-antibody-based" pathogen countermeasure by targeting the very same "alternative mode of host receptor binding" that the pathogen proteins exploit to cause infection and disease. We have chosen the Staphylococcus enterotoxin B (SEB) superantigen as a model pathogen protein to illustrate the principle and application of our strategy. SEB bypasses the normal route of antigen processing by binding as an intact protein to the complex formed by the MHC class II receptor on the antigen-presenting cell and the T cell receptor. This alternative mode of binding causes massive IL-2 release and T cell proliferation. A normally processed antigen requires all the domains of the receptor complex for its binding, whereas SEB requires only the alpha1 subunit (DRalpha) of the MHC class II receptor and the variable beta subunit (TCRVbeta) of the T cell receptor. This prompted us to design a bispecific chimera, DRalpha-linker-TCRVbeta, that acts as a receptor mimic and prevents the interaction of SEB with its host cell receptors. We have adopted (GSTAPPA)(2) as the linker sequence because it supports synergistic binding of DRalpha and TCRVbeta to SEB and thereby makes DRalpha-(GSTAPPA)(2)-TCRVbeta as effective an SEB binder as the native MHC class II-T cell receptor complex. Finally, we show that DRalpha-(GSTAPPA)(2)-TCRVbeta inhibits SEB-induced IL-2 release and T cell proliferation at nanomolar concentrations.

The high-resolution structure of the triplex formed by the GAA/TTC triplet repeat associated with Friedreich's ataxia.

Expansions of the triplet repeat, GAA/TTC, inside the first intron of the frataxin gene causes Friedreich's ataxia (FRDA). It was of interest to us to examine whether the FRDA repeat forms an unusual DNA structure, since formation of such structure during replication may cause its expansion. Here, we show that the FRDA repeat forms a triplex in which the TTC strand folds on either side of the same GAA strand. We have determined the high-resolution NMR structures of two intramolecularly folded FRDA triplexes, (GAA)2T4(TTC)2T4(CTT)2 and (GAA)2T4(TTC)2T2CT2(CTT)2 with T.A.T and C+.G.C triads. T4 represents a synthetic loop sequence, whereas T2CT2 is the natural loop-folding sequence of the TTC strand. We have also made use of site-specific 15N-labeling of the cytosine residues to investigate their protonation status and their interaction with other protons. We show that the cytosine residues of the Hoogsteen C+.G pairs in this triplex are protonated close to physiological pH. Therefore, it appears that the triplex formation offers a plausible explanation for the expansion of the GAA/TTC repeats in FRDA.

DNA repeats in the human genome.

Repetitive DNA sequences, interspersed throughout the human genome, are capable of forming a wide variety of unusual DNA structures with simple and complex loopfolding patterns. The hairpin formed by the fragile X repeat, (CCG)n, and the bipartite triplex formed by the Friedreich's ataxia repeat, (GAA)n/(TTC)n, show simple loopfolding. On the other hand, the doubly folded hairpin formed by the human centromeric repeat, (AATGG)n, the hairpin G-quartet formed by (TTAGGG)n at the 3' telomere overhang, and the hairpin G-quartet, and hairpin C+.C paired i-motif formed by the insulin minisatellite, [formula: see text] show multiple and complex loopfolding. We have performed high resolution nuclear magnetic resonance (NMR) spectroscopy and in vitro replication to show that unique base-pairing and loopfolding render stability to these unusual structures under physiological conditions. The formation of such stable structures offers a mechanism of unwinding which is advantageous during transcription. For example, the formation of the hairpin G-quartet, and hairpin C+.C paired i-motif upstream of the insulin gene may facilitate transcription. These unusual DNA structures also provide unique 'protein recognition motifs' quite different from a Watson-Crick double helix. For example, the hairpin G-quartet formed by (TTAGGG)n at the 3' telomere overhang is specifically recognized and stabilized by the human repair protein, Ku70/Ku80 hetero-dimer, which may be important in the stability of the telomere. However, the formation of the same unusual DNA structures during replication is likely to cause instability in the lengths of the DNA repeats. If the altered (generally expanded) length enhances the probability of the unusual structure during the next cycle of replication, it further increases the instability of the repeat causing a 'dynamic mutation'. In fact, NMR and in vitro replication studies show that the longer the repeat length the higher is the probability of hairpin formation by the fragile X repeat, (CCG)n. In addition, the hairpin of the fragile X repeat, upstream of the FMR-1 gene, is more susceptible to CpG methylation than its duplex thereby leading to methyl-directed suppression of transcription. Thus, the selective advantage of the unusual structures formed by the DNA repeats in the regulation of gene expression may be offset by the genomic instability caused by the same structures during replication. The repeat number is a critical parameter that helps maintain a balance between the advantage gained from an unusual structure during gene expression and the disadvantage posed by the same structure during replication.

Cystosine-rich strands of the insulin minisatellite adopt hairpins with intercalated cytosine+.cytosine pairs.

Previously, we reported the high resolution NMR structure of the hairpin G-quartet structure formed by the G-rich strand of the insulin minisatellite of repeat sequence, (ACAG4TGTG4/TGTC4ACAC4) located upstream of the human insulin gene. Here, we report structural studies on the C-rich strand of this insulin minisatellite. First, we show by high resolution NMR that (C4TGTC4) forms a hairpin dimer with intercalated C+.C pairs (referred to as the hairpin i-motif); 340 NOE distance constraints uniquely define the nature of hairpin folding and the pattern of C+.C intercalation. Second, we show by one-dimensional NMR spectroscopy and molecular modeling studies that (C4TGTC4ACA4TGTC4) forms an intramolecularly folded hairpin with intercalated C+.C pairs. Third, we demonstrate by in vitro replication studies that several such hairpin i-motifs are present in long (C4TGTC4ACA)n (n>/=6) sequences, even in the presence of their complementary strands. Finally, we discuss structural and biological significance of the hairpin i-motifs formed by the C-rich strands of the insulin minisatellite.

Structure-function correlations of the insulin-linked polymorphic region.

The insulin minisatellite of the insulin-linked polymorphic region (ILPR), a 14 base-pairs long tandem repeat of: 5'-ACAGGGGTGTGGGG-3' 3'-TGTCCCCACACCCC-5', is located 363 base-pairs upstream of the human insulin gene. A locus for insulin-dependent diabetes mellitus (IDDM) has been mapped to the ILPR. It has been shown that the ILPR is polymorphic in length and this length polymorphism is also related to the transcriptional activity of the insulin gene and the susceptibility to IDDM. Here, we attempt to decipher the role of the ILPR structure in length polymorphism and transcriptional regulation. We show by gel electrophoresis, circular dichroism (CD) and one and two-dimensional nuclear magnetic resonance spectroscopy (1D/2D NMR) that the G-rich strand of the ILPR adopts an intramolecularly folded hairpin G-quartet structure. A detailed analysis of 1D/2D NMR data of d(G4TGTG4) and d(G4TGTG4ACAG4TGTG4) enables us to define the nature of chainfolding, the stacking interaction of the G-tetrads in the stem, and the interactions of the bases in the loops. d(G4TGTG4ACAG4TGTG4) happens to be the smallest unit of the G-rich strand that can form the intramolecular hairpin G-quartet structure. For long ILPR sequences, several such hairpin G-quartet structures can be linked in space. Indeed, by an in vitro replication assay, we show the presence of such multiple hairpin G-quartet structures for the G-rich strand of the ILPR of repeat length 6. This observation suggests that the formation of multiple hairpin G-quartets may explain slippage during replication and the observed length polymorphism. From our high resolution structure, we are able to identify a set of interactions that are critical for the structure and stability of the hairpin G-quartet. Single or double mutations in the ILPR that destabilize these interactions also lower the transcriptional activity of the insulin gene. Therefore, the hairpin G-quartet structure of the ILPR has a direct correlation with the transcriptional activity of the human insulin gene.

Structure and polymorphism of HIV-1 third variable loops.

The third variable (V3) loop of HIV-1 surface glycoprotein, gp120, has been the target of neutralizing antibodies. However, sequence variation inside the V3 loop diminishes its effectiveness as a potential vaccine against HIV-1. The elusive nature of the V3 loop structure prompted us to carry out a systematic study on different isolates in an attempt to identify a common structural motif in the V3 loop regardless of the amino acid sequence variability. We have previously determined the structural features of two V3 loops: V3 Thailand and V3 MN. In this paper, we present the structure of two other variants: V3 Haiti and V3 RF. Our results show that similar secondary structures are observed in all the four V3 loops: a GPG(R/K/Q) crest in the center of the neutralizing domain, two extended regions flanking the central crest, and a helical region in the C-terminal domain. For the Haitian V3 loop, we also show how the conserved structural features are masked through a conformational switch encoded in the amino acid sequences on the C-terminal side of the GPGK crest.

Solution structures of the individual single strands of the fragile X DNA triplets (GCC)n.(GGC)n.

Three-dimensional structures of the fragile X triplet repeats (GCC)n and (GGC)n are derived by using one- dimensional/two-dimensional NMR. Under a wide range of solution conditions (10-150 mM NaCl,pH6-7)(GCC)5-7 strands form exclusively slipped hairpins with a 3' overhanging C. The slipped hairpins of (GCC)n strands show the following structural characteristics: (i) maximization of Watson-Crick G.C pairs; (ii) formation of C.C mispairs at the CpG steps in the stem; (iii) C2'-endo, anti conformations for all the nucleotides. The ability of (GCC)n strands to form hairpin structures more readily than complementary (GGC)n strands suggests preferential slippage during replication and subsequent expansion of the (GCC)n strands. In addition, the C.C. mispairs at the CpG site of (GCC)n hairpins account for their exceptional substrate efficiencies for human methyltransferase. Gel electrophoresis data show that (GGC)n strands form both hairpin and mismatched duplex structures in 10-150 mM NaCl (ph 6-7) for n < 10, but for n > or + 11 hairpin structures are exclusively present. However, (GGC)n strands remain predominantly in the duplex state for n=4-11 under NMR solution conditions, which require DNA concentrations 100- to 1000-fold higher than in gel electrophoresis. NMR analyses of [(GGC)n]2 duplexes for n=4-6 show the presence of Watson-Crick G.C and mismatched G anti G syn pairs. The mismatches adjacent to the CpG step introduce local structural flexibility in these duplexes. Similar structural properties are also expected in the stem of the hairpins formed by (GGC)n strands.

Structure of a tumor associated antigen containing a tandemly repeated immunodominant epitope.

Human mucins are T or S glycosylated tandem repeat proteins. In breast cancer, mucins become under or unglycosylated. Two-dimensional nuclear magnetic resonance experiments are performed on chemically synthesized mucin tandem repeat polypeptides, (PDTRPAPGST-APPAHGVTSA)n the unglycosylated form for n=1,3 where (APDTR) constitutes the antigenic sites for the antibodies isolated form the tumors in the breast cancer patients. These studies demonstrate how the tandem repeats assemble in space giving rise to the overall tertiary structure, and the local structure and presentation of the antigenic site(APDTR) at the junction of two neighboring repeats. The NMR data reveal repeating knob-like structures connected by extended spacers. The knobs protrude away from the long-axis of Muc-1 and the predominant antigenic site (APDTR) forms the accessible tip of the knob. Multiple tandem repeats enhance the rigidity and presentation of the knob-like structures.

Hairpins are formed by the single DNA strands of the fragile X triplet repeats: structure and biological implications.

Inordinate expansion and hypermethylation of the fragile X DNA triplet repeat, (GGC)n.(GCC)n, are correlated with the ability of the individual G- and C-rich single strands to form hairpin structures. Two-dimensional NMR and gel electrophoresis studies show that both the G- and C-rich single strands form hairpins under physiological conditions. This propensity of hairpin formation is more pronounced for the C-rich strand than for the G-rich strand. This observation suggests that the C-rich strand is more likely to form hairpin or "slippage" structure and show asymmetric strand expansion during replication. NMR data also show that the hairpins formed by the C-rich strands fold in such a way that the cytosine at the CpG step of the stem is C.C paired. The presence of a C.C mismatch at the CpG site generates local flexibility, thereby providing analogs of the transition to the methyltransferase. In other words, the hairpins of the C-rich strand act as better substrates for the human methyltransferase than the Watson-Crick duplex or the G-rich strand. Therefore, hairpin formation could account for the specific methylation of the CpG island in the fragile X repeat that occurs during inactivation of the FMR1 gene during the onset of the disease.

Local and global structural properties of the HIV-MN V3 loop.

Studies of the feasibility of a subunit vaccine to protect against human immunodeficiency virus (HIV) infection have principally focused on the third variable (V3) loop. The principal neutralizing determinant (PND) of HIV-1 is located inside the V3 loop of the surface envelope glycoprotein, gp120. However, progress toward a PND-based vaccine has been impeded by the amino acid sequence variability in the V3 loops of different HIV isolates. Theoretical studies revealed that the variability in sequence and structure of the V3 loop is confined to the N- and C-terminal sides of the conserved GPG crest. This leaves three regions of the V3 loop conserved both in sequence and secondary structure. We present the results of NMR studies that test the validity of our theoretical predictions. Structural studies are reported for the HIV-V3 loop (HIV-MN) in the linear and cyclic (S-S-bridged) forms. For the V3 loop sequence of the HIV-MN isolate, the three conserved secondary structural elements are as underlined below: turns turn helix CTRPNYNKRKRIHIGPGRAFYTTKNIIGTIROAHC Finally, the conformational requirement of the PND in the V3 loop-antibody interaction is tested by monitoring the monoclonal antibody binding to the HIV-MN V3 loop in the linear and cyclic forms by enzyme-linked immunosorbent assay. The binding data reveal that the cyclic V3 loop is a better ligand for the monoclonal antibodies than the linear form although the latter has the same sequence. This means that the monoclonal antibodies recognize the PNDs as conformational epitopes.

Unusual structures of the tandem repetitive DNA sequences located at human centromeres.

The presence of the highly conserved repetitive DNA sequence d(AATGG)n.d(CCATT)n in human centromeres argues for a special role for this sequence in recognition, most probably through the formation of an unusual structure during mitosis. Quantitative one- and two-dimensional nuclear magnetic resonance (1D/2D NMR) spectroscopic studies reveal that the Watson-Crick duplex d(AATGG)n.d(CCATT)n adopts the usual B-DNA conformation as illustrated by taking d(AATGG)3.d(CCATT)3 as an example, whereas the d(CCATT)n strand is essentially a random coil. In contrast, the d(AATGG)n strand adopts an unusual stem-loop motif for repeat lengths n = 2, 3, 4, and 6. In addition to normal Watson-Crick A.T pairs, the stem-loop structures are stabilized by mismatched A.G and G.G pairs in the stem and G-G-A stacking in the loop. Stem-loop structures of d(AATGG)n are independently verified by gel electrophoresis and nuclease digestion studies and were also previously shown to be as stable as the corresponding Watson-Crick duplex d(AATGG)n.d(CCATT)n [Grady et al. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 1695-1699]. Therefore, the sequence d(AATGG)n can, indeed, nucleate a stem-loop structure at little free energy cost, and if, during mitosis, it is located on the chromosome surface, it can provide specific recognition sites for kinetochore function.

Thermal stability of protein secondary structure in Langmuir-Blodgett films.

The temperature dependence of the secondary structure of photosynthetic reaction centres from Rhodobacter sphaeroides in solution and in Langmuir-Blodgett film was studied by circular dichroism. It was shown that the secondary structure of the protein was not affected in Langmuir-Blodgett films by heating up to 200 degrees C, while in solution it was completely lost at 55 degrees C. Molecular order rather than decreased hydration degree was held responsible.

DNA internal motions within nucleosomes during the cell cycle and as a function of ionic strength.

We have used 31P NMR spectra to show that DNA internal motions are greatly hindered within oligonucleosomes. The fluctuations seem to be a function of both the cell cycle and the number of nucleosomes interlinked. Namely, the resonance areas, directly related to unbound phosphate, are consistently smaller in M-phase than in S-phase; at the same time, the resonance line width, inversely related to base plane, deoxyribose, and phosphate internal motions, is consistently larger in mononucleosomes than in oligonucleosomes. In all cases, the removal of chromosomal proteins, by a progressive increase of ionic strength up to 2 M NaCl, increases the internal motion, as monitored by a decrease in line width toward that of free DNA. While for both oligo- and mononucleosomes in S-phase the decrease in line width is strictly correlated to a sharp increase in resonance area, in M-phase it is not, with the 31P resonance area rather low even at 2.0 M NaCl extraction. Similarly, while S-phase 31P line widths steadily grow from mono- to oligonucleosomes, in M-phase they do not. Moreover, the increase of the ionic strength to 0.6 M NaCl, as compared to 0.35, 1.2, and 2 M NaCl, displays significant variations on 31P line width and resonance area, independent of the cell cycle phase and the number of nucleosomes interlinked. These observations agree with earlier suggestions on the differential role of the various chromosomal protein subfractions, known to preferentially dissociate at the different ionic strengths in question, in the sealing of mononucleosomes and in the overall stability of polynucleosomes.

Improvement of protein secondary structure prediction by combination of statistical algorithms and circular dichroism.

Three different approaches (propensity curve shifting, hydropathy index evaluation, and iterative attribution/cancellation of secondary structure) to the use of secondary structure percentages derived from circular dichroism measurements to improve the success rate of a protein secondary structure prediction method, without using decision constants, are described and compared. Propensity-curve shifting appears to be the best-performing approach, bearing an increase of 5.3% in the success rate of single-residue structural prediction when exact information on the secondary structure, obtained by X-ray crystallography, is employed; with information of an accuracy comparable to that obtainable by circular dichroism, the improvement stays between 3.5 and 4.9%, for a three-state prediction. Although developed with circular dichroism in mind, the method can use percentages of secondary structure obtained by any other experimental methodology from which they can be inferred, for instance Raman spectroscopy and infrared spectroscopy.

The histone H1 globular region. A possible supersecondary structure from spectroscopic and statistical studies.

The central region of the basic nuclear protein, histone H1, has a highly conserved amino acid sequence and a globular structure which is still not known at atomic resolution. A possible secondary and supersecondary structure was predicted by combining experimental measurements of circular dichroism and NMR spectroscopy with a statistical method based on the amino acid sequence. Our results showed the protein fragment as being highly structured and having a total alpha-helix content of about 40%.

Biophysical approach to the determination of the secondary structure of the histone H1 globular region.

A possible secondary structure of the globular part of the histone H1 was obtained with a statistical approach based on the GOR method. The results of circular dichroism measurements on the protein were taken into account in order to choose between theoretically equivalent structures.