Nucleation sites, proteins

A continuous lipidic cubic phase is obtained by mixing a long-chain lipid such as monoolein with a small amount of water. The result is a highly viscous state where the lipids are packed in curved continuous bilayers extending in three dimensions and which are interpenetrated by communicating aqueous channels. Crystallization of incorporated proteins starts inside the lipid phase and growth is achieved by lateral diffusion of the protein molecules to the nucleation sites. This system has recently been used to obtain three-dimensional crystals 20 x 20 x 8 pm in size of the membrane protein bacteriorhodopsin, which diffracted to 2 A resolution using a microfocus beam at the European Synchrotron Radiation Facility. 

Recent interest has focused on acidic phosphoproteins, such as bone sialoprotein, acting as sites of nucleation. These proteins contain motifs (eg, poly-Asp and poly-Glu stretches) that bind calcium and may provide an initial scaffold for mineralization. Some macromolecules, such as certain proteoglycans and glycoproteins, can also act as inhibitors of nucleation. 

The initial stages of iron incorporation requires the ferroxidase sites of the protein. Thereafter the inner surface of the protein shell provides a surface which supplies ligands that can partially coordinate iron but which leave some coordination spheres available for mineral phase anions, thereby enabling the biomineralization process to proceed, with formation of one or more polynuclear ferrihydrite crystallites. Iron is transferred from the ferroxidase sites to the core nucleation sites by the net reaction (Yang et ah, 1998) ... 

It is unfortunately the case that when we incubate apoferritin with a certain number of iron atoms (for example as ferrous ammonium sulfate), the product, after elimination of non-protein-bound iron, does not have a homogeneous distribution of iron molecules which were able to (i) take up iron rapidly through the three fold channels, (ii) quickly transfer it and form a diiron centre on a ferroxidase site, and (iii) to transfer the iron inward to a nucleation site, where (iv) it will begin to catalyse iron oxidation on the surface of the growing crystallite, will accumulate iron much more rapidly, and in much greater amounts than molecules in which steps (i), (ii) and (iii) are slower, for whatever reasons (perhaps most importantly subunit composition, and the disposition of subunits of the two types H and L, one with regard to the other). This polydispersity makes the analysis of the process of iron uptake extremely difficult. 

Hence, the overall reaction for iron oxidation and hydrolysis at the ferroxidase centre, followed by further hydrolysis and migration to the core nucleation sites consists of two reactions, the protein-catalysed ferroxidase reaction itself and the Fe(II) plus H202 detoxification reaction (Equations (19.7) and (19.8), respectively) ... 

Why mammalian ferritin cores contain ferrihydrite-like structures rather than some other mineral phase is less easy to understand, and presumably reflects the way in which the biomineral is built up within the interior of the protein shell together with the geometry of the presumed nucleation sites. The phosphate content in the intracellular milieu can readily be invoked to explain the amorphous nature of the iron core of bacterioferritins and plants. Indeed, when the iron cores of bacterioferritins are reconstituted in the absence of phosphate, they are found to be more highly ordered than their native counterparts, and give electron diffraction lines typical of the ferrihydrite structure. Recently it has been reported that the 12 subunit ferritin-like Dps protein (Figure 19.6), discussed in Chapter 8, forms a ferrihydrite-like mineral core, which would seem to imply that deposition of ferric oxyhydroxides within a hollow protein cavity (albeit smaller) leads to the production of this particular mineral form (Su et al., 2005 Kauko et al., 2006). 

Nuth M,YoonT, Cowan JA. 2002. Iron-sulfur cluster biosynthesis Characterization of iron nucleation sites for the assembly of the [2Fe-2S] cluster core in IscU proteins. J Am Chem Soc 124 8774-5. 

Helix induction can be augmented by mimicking natural mechanisms using a variety of synthetic expedients. In general, these approaches seek to either stabilize or mimic helix nucleation sites or to naturally or artificially secure side-chain linkages. In most cases, these transformations make use of additional synthetic procedures beyond standard peptide synthesis methodologies to juxtapose amino acids or essential elements thereof in a manner that mimics a protein helix. 

It is not unreasonable to believe, for example, that the collision of two a-helices could become a nucleation site for protein folding. Thus, the alignment of two a-helical chains could determine the overall fold and, as a consequence, the function of the protein. 

In almost all instances of biological mineralization fibrous proteins represent the bulk of the organic matrix. In the past, this phenomenon has been interpreted to mean that proteins such as collagen, keratin or elastin are the key elements in mineralization by providing nucleation sites and at the same time offering structure and space for oriented crystal growth. However, with the advance in the field of biomineralization this model came under severe attack. At present, there is no universal concept which could explain all the intriguing facets of phosphate deposition in cellular systems. 

In the chiton tooth, the organic framework components are synthesized and secreted by the cells into the extracellular space, and there they self assemble. By the time mineralization is about to occur the cells are tens of micrometers away from many of the mineralization sites. They must therefore operate by remote control. The mineralization sites themselves are within a complex chitin framework, the dimensions of which are in the nanometer range. The sea urchin larval spicule represents the exact opposite situation. Mineralization occurs in a vacuole defined by a membrane, and the entire apparatus is within a consortium of fused cells (the syncytium). The membrane of the syncytium tightly surrounds the growing spicule [74], Therefore, it has been proposed that the cells directly control spicule formation. The mineralization vacuole is subdivided by framework proteins. Nothing is known about the structure of the one nucleation site per spicule in the larvae, but in the adult a well-defined location, enclosed within a framework, has been identified as the nucleation site [83]. Dentin formation is intermediate between the two. It is an extracellular process, and the distances between cells or cell processes and mineralization sites are in the range of tens of micrometers or several micrometers respectively. Nucleation occurs within the fibril or at its surface and is associated with a site on the fibril surface some 7 or 8 nm wide [54]. The space available for crystal growth within the fibril is even smaller in one of the dimensions, namely 2 or 3 nm wide. 

The microenvironment in which crystals form is important for understanding the control over mineral formation in biology. In each of the three systems described above, the microenvironments are very different in terms of shape and scale, confining components and the manner in which the space is further sub-divided. Clearly however a key element of the microenvironment is the nucleation site itself. There are numerous indirect observations in vivo and in vitro demonstrating that much of the control over mineral formation is exercised through the design and structure of the nucleation site. We know very little, however, about actual nucleation sites in vivo. In fact to date no protein or other specific components of a nucleation site have been unequivocally identified, let alone characterized. This is certainly one of the major immediate goals of this field. 

It would seem likely that contiguity of a number of such self-determined nucleation sites might well result in an accelerating cascade of interactions that could help account for the remarkable speed of protein folding. Understanding of the additional control and facilitation of the overall process by catalysts of the sort described in this volume might soon lead to a true understanding of this fundamental problem. 

---