[ PDB file ] [ PubMed link ]

In order to get an idea of the conformation of SS3 at a sub-zero temperature, we determined the structure of SS3 at -5 degrees Celcius. Getting the solution down to that temperature without freezing was straighforward – slow cooling together with the salt present in the buffer (~50 mM phosphate) does the trick. No dramatic changes were observed – the structure comprised a large and a small helical region, connected by a short, slightly less well defined sequence.

[ PDB file ] [ PubMed link ]

Many animals that live in extremely cold climates produce proteins that act to prevent the formation of ice crystals in the bodily fluids of the organism. Antarctic fish produce a range of such proteins, as the water temperatures can drop below 0 degrees Celsius. We have determined the structure of the type I antifreeze protein (SS3) from the blood of the shorthorn sculpin at both 5 degrees and -5 degrees. The structure largely comprises a long alpha-helix that displays conserved Thr residues (red) along a single surface. These residues are important for preventing the growth of ice crystals, although the mechanism of action is not currently understood. An additional short helix exists at the N-terminus of the protein, although the conformational relationship between the two helices is not well defined. This is the first solution structure of a wild-type type I antifreeze protein.

[ PDB file ]

This structure shows a two-residue insertion mutant of BKLF F3 that possesses native function but not native structure. This zinc finger retains the ability to bind to GATA-1, even though it contains two additional bulky residues in a structurally important region.

[ PDB file ]

This protein is tryptophan-containing mutant of BKLF F3 that retains native structure and function. The addition of tryptophan enables the use of fluorescence spectroscopy for conducting folding studies.

[ PDB file ] [ PubMed link ]

Red blood cells contain an excess of the alpha-chain of hemoglobin over the beta-chain. It is known that excess alpha-globin is unstable, and its precipitation can lead to membrane damage and apoptosis. This phenomenon is the root of the beta-thalassemia phenotype, in which sufferers cannot make sufficient beta-globin. How is the normal excess of alpha-globin kept ‘safe’? The answer seems to lie in AHSP – a protein that appears to act as a specific chaperone of alpha-globin, binding to the globin and preventing its precipitation. We have determined the structure of AHSP and used NMR titration methods and mutagenesis to reveal which residues are involved in mediating this specific interaction. In addition, as part of a collaboration with Yigong Shi at Princeton, the structure of the AHSP:a-globin complex was determined by X-ray crystallography. This structure confirmed many of our predictions based on biophysical analysis of the interaction. Collaborative work is also ongoing with Peter Lay in the School of Chemistry at USyd to examine the properties of the metal-binding site in the complex.

[ PDB file ] [ PubMed link ]

FOG-1 is vital for the development of red blood cells and platelets. Originally identified as a Friend of GATA, GATA-independent functions are now emerging. Recently, we identified TACC3 as a protein partner of FOG-1. FOG-1 uses its third (of nine) ZnF to bind the coiled-coil domain of TACC3. The structure of FOG-1 finger 3 (left) shows that it is a typical classical CCHH finger (BMRB 6216). NMR and mutagenesis data identified the residues important for the interaction between FOG-1 (yellow) and TACC3 (blue) and this allowed us to use HADDOCK to calculate the structure of the FOG-TACC3 complex (right).

[ PDB file ] [ PubMed link ]

The transcriptional regulator LMO4 consists entirely of two LIM domains, and must bind to the partner protein ldb1 in order to be transported into the nucleus and carry out its role in the regulation of lymphogenesis. We have determined the structure of a complex formed betwen full-length LMO4 and the LMO4-binding region of ldb1 using X-ray crystallography. The interaction interface is completely extended, and the ldb1 peptide (yellow) forms additional beta-strands on beta-hairpins present in both of the LMO4 LIM domains. This tandem beta-zipper arrangement has been observed in one other recent structure – a bacterial pathogen that binds cellular fibronectin repeats (Potts et al., Nature 2003, 423, 177). This structure can now serve as a scaffold from which inhibitors of the LMO4:ldb1 interaction may be designed.

[ PDB file ] [ PubMed link ]

BKLF is a transcriptional repressor involved in erythropoiesis. It has been shown to interact with both DNA (CACCC boxes) and the C-terminal zinc finger of GATA-1. We have determined its solution structure as a first step towards characterizing this interaction. In addition, we have shown that variants of this domain in which the final zinc-ligating histidine is mutated to residues such as asparagine and alanine are still functional in binding both GATA-1 and, in the context of a 3-zinc-finger construct, DNA.

[ PDB file ] [ PubMed link ]

Shown is the structure of a mutant of Mi2-P2, with the second flexible loop (called L3) replaced by the corresponding region from another PHD domain from the protein WSTF. This structure shows that the PHD fold is indeed preserved in the mutant. The backbone of L3-WSTF overlays well with Mi2b-P2 except for some minor differences observed at residues 20-22 and 43-46, which correspond to the two flexible regions (shown in orange and purple).

[ PDB file ] [ PubMed link ]

We have used NMR methods to determine the solution structure of the second PHD finger from the transcriptional regulator Mi2b. This structure indicates that PHD fingers contain two relatively flexible loop regions (shown in orange and red) and a conserved stable core that ligates two zinc atoms (zinc ligation residues are shown in yellow and green). By using a strategy of mutagenic substitution and expansion, we have also shown that a wide variety of mutations in these loops can be tolerated without major disruption of the PHD fold. If you want to see for yourself, the solution structure of one of these mutants is shown in the next figure!

[ PDB file ] [ PubMed link ]

Bacterial superantigens are involved in overstimulating T cells. YPM has a structure that is unique amongst classical superantigens. Typical superantigens consist of 2 b-strand domains and a central α-helix. However YPM consists of 8 antiparallel β-strands that form a jelly roll motif. There are 2 main regions in the YPM structure, one that is quite rigid and another that undergoes intermediate exchange. The unique structure of YPM suggests that it may activate T cells through a mechanism that is distinct from other superantigens.

[ PDB file ] [ PubMed link ]

We have solved the structure of a complex of the transcription regulators LMO4 and ldb1 (PDB: 1M3V). To do this we engineered FLIN4, a fusion of the LID domain of ldb1 and the N-terminal LIM domain of LMO4. The structure of FLIN4 shows that ldb1-LID (yellow) binds to both Zn1 (cyan) and Zn2 (magenta) of LMO4-LIM1 in an extended fashion with ldb1-LID forming a short b-strand that extends a b-hairpin present in LMO4-LIM1.

[ PDB file ] [ PubMed link ]

FLIN2 (PDB: 1J2O) is an engineered intramolecular complex comprising a fusion of the LID domain of ldb1 and the N-terminal LIM domain of LMO2. The solution structure of FLIN2 shows that ldb1-LID (yellow ribbon) binds to the LMO2 LIM1 domain (surface representation) in an extended fashion. Half of the interaction interface involves predominantly backbone-backbone and side-chain hydrophobic interactions, the other involves mainly electrostatic interactions.

[ PDB file ] [ PubMed link ]

ZNF265 is a mammalian splicing factor, which contains an Arg-Ser rich domain and a putative double zinc finger motif that is capable of binding to RNA (putting it in the SR class of proteins). Shown here is the structure of one of the two zinc fingers. The structure comprises two beta-hairpins, crossing each other at an angle of ~80 degrees, and sandwiching a single zinc atom. The fold is unlike any known zinc-binding motifs and we have termed it the crossed finger. Remarkably, the positions of the residues shown in space-fill mimic the RNA-binding residues of the unrelated dsRBM (double-stranded RNA-binding module) domain, indicating the convergent evolution of a binding surface on different scaffolds.

[ PDB file ] [ PubMed link ]

This non-native structure corresponds to a fragment of the CH1 domain of CBP. Although the CH1 domain in its native form binds 3 zinc ions (Dames et al, PNAS 2002, 99, 5271-5276), this shorter domain binds only one zinc ion, through a different combination of zinc ligating residues. This novel zinc-binding fold is quite remarkable, given that a sequence, taken out of the context of a whole domain, is able to fold into a stable, yet different structure.

[ PDB file ] [ PubMed link ]

USF9 has the same zinc ligation topology as USF1 but the two fingers share only 14% sequence identity outside of the zinc ligands (blue). However, USF9 adopts an almost identical fold to USF1, with a short β-hairpin (orange) and an α-helix. Furthermore, USF9 is also able to interact with the N-terminal finger of GATA-1.

[ PDB file ] [ PubMed link ]

U-shaped is a member of the FOG family of transcription factors, which interact with GATA proteins to control gene expression in various tissues. U-shaped has nine zinc fingers. The structure of the first finger of U-shaped (USF1) is very similar to that of a classical zinc finger, consisting of a short β-hairpin (yellow) and an α-helix. The central zinc atom (green) is ligated by 1 His and 3 Cys residues (blue), in contrast to classical fingers, which have a CCHH zinc ligation topology. While many classical fingers bind to DNA, USF1 has been shown to interact with the N-terminal zinc finger of GATA-1.

[ PDB file ] [ PubMed link ]

The structure of GNF consists of two distorted β-hairpins and an α-helix, with a central zinc atom (green) ligated by four cysteine residues (red). GATA-1 interacts with transcription factors from the FOG family using this N-terminal zinc finger. This interaction is crucial for GATA-1’s ability to correctly regulate gene expression in hematopoetic cells.