Folding the ferritin alpha helix

A polypeptide emerges from the ribosome as an alpha helix held in such configuration by hydrogen bonds (hbonds) between the backbone nitrogen of each amino acid and the backbone oxygen of the amino acid four positions away (cward, toward the C terminus of the polypeptide). Folding as a molecular process is likely orders of magnitude faster than polypeptide synthesis and occurs as soon as the opportunity arises as the alpha helix emerges from the ribosome.

In order to fold, that is to rotate on the bonds between the CA atom and the N (the phi bond) or C atom (the psi bond), hbonds must be broken. Specifically four hbonds in a row must be broken, allowing rotation on the psi and phi bonds as the helix unwinds until collisions or new hbond formation stops the rotation. Since the specificity of the side-chains determine the folding pattern, the side chains must be involved. From my observations, the following amino acids have side-chains that appear to disrupt alpha helix hbonds:

Glycine: the side chain is just a hydrogen and allows water to compete for hbond formation with the polypeptide. The side-chains of all other amino acids sterically block water's access.

Aspartic acid or asparagine: the side-chain oxygen(s) close to the backbone appears from observations to disrupt the hbond on their oxygen and the hbonds on their nitrogen and the cward amino acids's nitrogen. Interestingly glutamic acid or glutamine, identical but with one more carbon in the side-chain, seem not to disrupt.

Histidine: this amino acid appears to disrupt the hbonds on the oxygens of the amino acids 3 and 4 positions nward (towards the N terminus of the polypeptide).

Proline: the side chain loops around and covalently bonds its backbone nitrogen and so prevents any hbonds on that nitrogen. It also forces a kink in the helix.

Cystine, serine or threonine: if the next amino-acid nward is not aspartic acid and the next amino-acid cward is not methionine or histidine, the hbonds on its oxygen and the hbonds on its nitrogen and the nitrogen of the next cward amino acid are disrupted. Serine and less so threonine can sometimes form side-chain hbonds in the middle of a folding event effectively freezing the rotations.

Lysine: its own back-bone nitrogen hbond appears to be disrupted though I cannot see why.

Arginine: though it appears to not disrupt hbonds, it may be able to prevent aspartic acid four positions away from doing so by double-hbonding its side-chain with its own side-chain. It appears this may also occur in other positions in the middle of a folding event effectively freezing the rotations.

No doubt there are other interactions and combination actions to be discovered and these observations may be simplistic or even wrong. Nevertheless applying just these rules, I have a Jmol script that can make the principal folds in the ferritin polypeptide.

The folding script allows rotation on the phi and psi bonds when four hbonds in a row are disrupted as the polypeptide emerges from the ribosome. The script mostly rotates the freed amino acid's psi bond from a dihedral angle of -26 to about +120 as the helix relaxes and unwinds allowing water access to the backbone. The phi bond is mostly constrained by steric affects and only rotates negatively slightly as the amino acid moves from the alpha helix position in its Ramachandran plot to the beta sheet region.

Water, hbonds and resonance

Some high-school chemistry:
An atom consists of a number of positively charged protons encased by shells containing the same number of negatively charged electrons. The shells form by the attraction of opposite charges and in spite of the repulsion of like charges. The innermost shell has but one "position" where bonds can form whereas others have four positions to accommodate the repulsion of like charges. The four positions in outer shells are tetrahedrally distributed around the nucleus containing the protons where each position contains 0, 1 or 2 elections (though no position in a shell contains 2 when other positions contain 0). For positions with single electrons, the electron may be shared with a single electron from another atom to form a covalent bond, tying the two atoms into a molecule.

Two atoms form bonds with each other by sharing pairs of electrons from their outermost shells. The hydrogen atom has but a single proton and thus but a single electron and so can bind only one other atom. Consequently a hydrogen atom can participate in a chain of atoms only as a terminator. An oxygen atom has 8 protons and thus 8 electrons, 2 in the inner shell and 6 in the outer shell with two positions with single unpaired electrons. Consequently under biological conditions, an oxygen atom can bind two other atoms or even form a double bond with one other atom. Similarly carbon with 6 protons can bind up to four other atoms and nitrogen with 7 protons can bind up to three.

Methane is a carbon atom bound to four hydrogens:

As you can see the molecule is symmetrical but, if one, two or three hydrogens are removed it is no longer so. Likewise a water molecule, an atom of oxygen bonded to two hydrogens, is necessarily asymmetric:

The consequence of asymmetry is that the internal charges of the molecule are not evenly distributed so that one end has a net partial positive charge and the other a net partial negative charge. Multiple identical molecules will then tend to align so that the opposite charges face each other forming a weak so-called ionic bond. This bond when one atom is hydrogen is termed a hydrogen bond, which, because of peculiarities of hydrogen, is stronger than other ionic bonds and is thought to even have some covalent properties.

Each water molecule with hydrogens on two of the four binding positions and electron pairs on the other two positions can make hydrogens bonds with two other water molecules forming the crystal structure we know as ice when temperatures are below the freezing point. At biological temperatures, the thermal vibration of the molecules disrupts enough of the hydrogen bonds such that water is a liquid but no so many that it becomes a gas.



Other molecules such as polypeptides can also form hydrogen bonds with water and this is in fact how water acts as so effective a solvent. Further, thanks to a phenomenon called resonance, ultra-pure water can even attack glass and metals. Resonance explains how helium can be made a liquid even though its atoms are perfectly charge-symmetrical. The theory is that, for two adjacent helium atoms at a given instance, their electrons could be positioned such that the atoms have some degree of charge separation that causes the two atoms to attract each other. Of course as likely, the electrons will position such that the two atoms repel each other and so no net attraction would be expected. The attraction event, however, is self-reinforcing whereas the repulsion event is self-diminishing. In the attraction event, the molecules in a sense move closer increasing the strength of the attraction whereas they move apart during the repulsion event diminishing the strength. Though this affect is overwhelmed by other forces compelling the movement of electrons, it can set up a resonance where the electrons move in paths to maximize attraction.

In pure water, such a resonance can occur through the entire mesh of hbonds effectively energizing water molecules in contact with the vessel walls sufficient to cause corrosion. Even the minor amount of impurities in tap water is sufficient to break-up the mesh such that tap water is much less corrosive than distilled water.

A chain of hbonds exists in the helical form of polypeptides and likewise resonance can energize atoms at the ends of the chain. Furthermore, resonance can cause attraction between any two atoms and such so-called dispersion bonds also play a role in protein folding. Though weak, dispersion bonds are numerous in large molecules and their contributions to 3D structure are likely significant.

The basic premise of protein folding

A molecule is a set of atoms connected to each other by covalent bonds. A covalent bond involves the sharing of an electron pair across two nuclei.  When two atoms are so bonded, their movements relative to each other become very restricted in all degrees of freedom, translational and rotational, except one - rotation on the axis of the bond. Within the domain of temperatures viable to life on Earth, single-bonded atom pairs freely rotate on the bond (though certain angles will be energetically favored over others and may even be forbidden by steric constraints in particular instances).

Polypeptides emerge from the ribosome in an alpha-helix form, effectively a straight rod. The polypeptide then spontaneously folds into its mature form almost exclusively by rotation on the two single bonds internal to each amino acid. All amino acids involved in protein synthesis have a central carbon (CA) with a single bond to a nitrogen (N) and another single bond to another carbon (C) that is itself double bonded to an oxygen (O). When two amino acids are joined by the ribosome, a bond is formed between the C of one and the N of the next. No rotation is seen on this new bond, however, because of the nature of oxygen and nitrogen. Nitrogen has a weaker hold on its electrons than does carbon whereas oxygen has a stronger hold. The result is that an other electron pair is drawn to some degree across the bond between the amino acids preventing rotation on that peptide bond.

The central carbon (CA) of all amino acid species except glycine also single bonds to a third carbon (CB) that further bonds to other atoms depending on the particular amino acid species involved. This so called side chain ( as opposed to the N-CA-C "backbone") must interact with the alpha-helix to direct the folding process since polypeptides differing only in the particular sequence of amino acid types are seen to fold in various but deterministic ways.
 

More on Ferritin

About 2.4 billion years ago the Great Oxygenation Event began and, for the next 600 million years, oxygen from photosynthetic blue-green algae entered the oceans and the atmosphere. Life on Earth then had to adapt to the then poisonous effects of free oxygen.  Some organisms persisted even to this day in compartments where oxygen does not reach. Others had to evolve to withstand and eventually to exploit the new conditions.

One problematic condition facing life was the disappearance of soluble iron as the soluble ferrous form was driven to the insoluble ferric form as oxygen entered the oceans. Iron then precipitated out and formed the banded iron formations that supply much of modern iron ores. Iron, which before was plentiful, suddenly became scarce and, even to this day, algae cannot thrive in the central oceans for lack of it. Near shore where iron is available from land runoff and from cycling of shallow sediments photosynthetic life persisted, principally by evolving the protein ferritin so that iron could be stored in a compact mineral form in the cell. 24 ferritin molecules interact to form a hollow sphere into which ferritin precipitates iron atoms by reversibly removing an electron to form the ferric form from the soluble ferrous form. A single sphere can then store up to 4500 iron atoms. So successful was this approach that ferritin has remained structurally identical in nearly every organism on Earth for 2 billion years.

These filled ferritin spheres are ferromagnetic and could then have served as the basis for sensors of the Earth's magnetic field. Also because of their density they could serve as the basis for sensors of gravity and motion as well. Perhaps, thanks to ferritin, cells could then evolve to sense their macro environment.  Motile cells could then move with deliberation and animal life could become possible.

Ferritin

Iron is the most abundant element on Earth by mass. Super-massive stars are able to fuse elements to higher and higher atomic weights as they evolve. They then die in supernova explosions and spew out their fusion products.  It happens that fusion up to the atomic weight of iron is heat generating but then heat absorbing after that, so it is not surprising that so much iron has accumulated in the universe. Iron is a transition element meaning that it has unfilled lower orbitals in its neutral state.  It can thus accept or surrender electrons with little change in its free energy which makes it ideal for mediating chemical reactions. It also makes it dangerous as it can readily react to whatever it bumps into.  Life, therefore from the very beginning has had to deal with it and the simple protein ferritin has been a universal answer, at least for the storage aspect. All life forms, prokaryotes, eukaryotes, and archaea have variants of ferritin.  Mitochondria and plastids even have their own variants.  The structure of all variants is highly conserved. 24 ferritin molecules interact to form a hollow ball. Iron ions in the +2 ferrous form are then deposited by ferritin inside the shell in the +3 ferric form (up to 4500 atoms per shell!), available for withdrawal as the need arises.

I have chosen ferritin as the ideal target of my folding software for several reasons.  It is relatively simple consisting of just four alpha helixes with a strap between them, and the many variants all have basically the same structure.  Hundreds of x-ray crystallization studies have been done on its many variants to high resolution including waters of hydration. Finally it interacts with  itself to form predictable macro structures. Mutant forms have even been generated and crystallized to provide insights on its folding and aggregation dynamics. It does, however, lack beta sheets so I will have to find another target eventually to fool around with.

New Title!

For the last few years, apart from minor revisions to the Protein Cycling Diet book, I have been playing around with the question of predicting how a polypeptide would fold if you knew only its amino acid sequence. This question is the holy grail of protein chemistry and has proven to be an exceedingly difficult problem and I hardly expect to solve it. I am now approaching it with a tool that I have not found others to have yet employed namely Jmol, an interactive viewer for three-dimensional chemical structures and a brilliant piece of work currently maintained by Bob Hanson of St. Olafs university.

I used to do my folding work in Java and then use Jmol to display the results. After Bob added the ability to edit polypeptides and in particular to rotate bonds, I was able to do all my work with the Jmol scripting language.  My initial output was a library of Jmol scripts: a set of tools for generating and manipulating polypeptides (and polynucleotides) found here. Now I am attempting to write scripts for predictive folding and have decided to discuss my project in this blog. Accordingly I have changed the title to the more general topic, Protein folding. After all, the earlier diet topic was about how to help dispose of misfolded proteins, so I have only broadened the scope of the blog and will continue to speculate here on relevant developments in autophagy science as well.

New revision!

A new revision of the book is available in the links on the left.  I added chapters on anti-oxidants, cancer, cardiovascular diseases and supplements in view of recent studies, especially on the potential for protein cycling to benefit cardiovascular outcomes. The WordPress editor is a real bear and I've yet to figure out how to make a link to the middle of a page.  Instead I directed the footnotes directly to their sources and have updated the links to PDFs where available.