DNA codes

Stereochemistry refers to the spatial arrangements of atoms in molecules and complexes. The linear model can account for no stereochemistry of a polypeptide beyond the sequential arrangement of amino acids.
Primary sequence – the sequence of amino acids in a polypeptide.
The linear model holds that nucleotide information is translated into primary sequences of amino acids. However, there is almost surely stereochemical information in some form as well. So I reject the widely accepted notion – the definition actually – that primary sequence is synonymous with primary structure. I advocate separate definitions for these obviously different terms.
Primary structure – the structure of the polypeptide backbone before folding.
This is the insanity of Rafiki. Of all the ideas presented in these web pages, the notion that genetic information can be fundamentally stereochemical is the most unconventional. Nucleotides are translated into primary structure instead of just primary sequence - the heretical conclusions of a child unsupervised.
A brief sketch of the cellular genetic mechanisms will serve as a base for more specifics of this concept.
Genetic information is stored in the cell (in the nucleus of eukaryotes) by DNA. This information is transcribed into messenger RNA (mRNA) and transported out of the nucleus into the cytoplasm. The information is translated into polypeptides by a mechanism that involves mRNA, transfer RNA (tRNA), ribosomes (rRNA) and amino acids.
Determining genetic information at the level of the polypeptide is a simple matter of counting. Consider the case of a hypothetical sequence of 100 amino acids. Given all the possible ‘synonymous’ nucleotide sequences that can produce the same sequence of 100 amino acids, how many discrete polypeptide backbones can emerge from translation before the protein begins to fold?
The linear model says that at this point we are certain of primary sequence, and that all primary sequences are created equal. It says that the genetic code has no mechanism to impart more information into the molecule beyond the identity of its sub-units, so the count of polypeptide backbones is unimportant. The Rafiki model says that primary structure is the real name of the game before the translated polypeptide begins to fold. The code has the ability to translate many primary structures that all share the same primary sequence, and that each discrete structure has a potential to fold quite differently and therefore become a functionally different protein. Rafiki sees a much greater amount of genetic information getting through this step in translation and being stored in the conformations of each peptide bond.
This idea is clearly nuts, but is it true? Common sense says it should be, and the weight of existing evidence seems to support it. Why is it so hard to accept? Why not test it?
A codon only gives part of the information in the genetic code. Triplets are seemingly capable of translating only enough information to define the sequence of amino acids. However, of equal importance to the ultimate shape and function of the protein is the critical information regarding the nature of each peptide bond. For proper translation of stereochemical information, the genetic code looks beyond the individual codon. All of the genetic information taken together during translation defines the entire peptide bond, information which includes:
1. Amino acid identities.2. Major bond configuration.3. Bond rotations.
For this reason, the salient unit of information in the genetic code goes beyond the codon to specify the complete peptide bond. I will call this unit the pepton. Mapping this information is discussed elsewhere. The following illustrations are designed to demonstrate the three components of the peptide bond.
The peptide bond combines two amino acids, or tetrahedrons, and it is formed by a group of atoms from each side of the bond called the peptide group, illustrated here by a rectangle. The first amino acid contributes to the peptide group its cabonyl carbon (C1), a-carbon (alpha 1) and the carbonyl oxygen (Oxygen). The second amino acid contributes its amide nitrogen (N), the a-carbon (alpha 2), and the amide hydrogen (Hydrogen). These molecule schematics are for relative position and not scale.
Although there is some wiggle to the group as a whole, for the most part the bond is planar, and the group can take one of two major configurations: cis or trans. The above illustration demonstrates the trans configuration. Once formed, the bond will not spontaneously switch configurations; however, there are enzymes capable of catalyzing the switch, which involves a 1800 rotation around the axis of the peptide bond.
As we can see here, the cis configuration of the peptide group has a much greater steric hindrance between the two amino acids. This means that the atoms get in the way of each other, and they do not like this. For this reason nearly all of the peptide bonds in naturally occurring proteins are in the trans configuration, at the very least 95% of them are trans-peptide bonds. However, the cis configuration can and does naturally occur. Apart from thermodynamic folding, is there genetic information, or a translation mechanism to dictate when a cis bond should occur?
When a cis bond does occur, it almost always involves an amino acid called proline. Proline is like a latch that locks the backbone in place. Cis bonds and proline are prominent in protein structures called turns, where the peptide backbone makes a dramatic change of direction. Proline and cis bonds therefore have a significant impact on the final shape and function of a protein. It logically follows that the genetic code should somehow specify the creation of bond configuration during translation. There are a variety of mechanisms that might do this, but whether or not it can is a key to information content in the protein.
Despite a decided predilection for trans configuration (perhaps all bonds are created trans) there is another opportunity for the genetic code to specify the shape of the polypeptide backbone. Each peptide bond has a measure of rotational freedom around the bond between the a-carbon and its peptide group member on either side. The rotation between the a-carbon and the amide nitrogen has been named phi. The rotation between the a-carbon and the carbonyl carbon has been named psi.
Together the two rotations determine the bond angle between adjacent peptide bonds. The planes of the peptide groups are not typically parallel in an actual protein, and the location of the R-groups can vary relative to the polypeptide backbone. This is a critical factor in determining the ultimate folding of the protein, and therefore the function of the protein. It is not unrealistic to expect a system that determines the ultimate function of the overall organism to somehow determine this essential element of organism building - before folding.
The overall peptide bond angles are determined by the combination of these two rotational angles, phi and psi. These combinations can be “good” or “bad” from the standpoint of steric hindrance and therefore energy stability. A plot of these combinations was done by Ramachandran and co-workers, and was later confirmed by empiric data. A simplified view is shown below: (Original URL for this picture: http://www.cryst.bbk.ac.uk/PPS95/course/3_geometry/rama.html)
The range of actual bond angles will in a large part be dictated by the actual amino acids participating in a particular bond. The spread of a Ramachandran plot involving glycine, for instance, will be much larger than one involving proline, as will be described shortly.
Information involves quantizing and specifying, and it seems then that the peptide bond, like everything in the universe, can be neatly quantized. We can describe the bond with a few parameters. We can simply quantize each parameter, digitize them so to speak, and then combine them to create many possible bonds. A language capable of executing this process would go a long way toward folding a protein.
In honor of Ramachadron’s work, I have named the quantity of information contained in the genetic code regarding the peptide bond rotation the R-number. The information regarding the major bond configuration is the C-number. The C-number tells us the bend and the R-number tells us the twist in each peptide bond.
There is one other property of the bond that merits attention; it is the laxity of the bond. Every bond has some wiggle or play. The degree to which this is true depends on the specific functional groups on the amino acids involved in the bond. For instance, proline has no play, but Glycine is a virtual swivel. A Glycine-Glycine bond is sloppy, but a Proline-Proline bond is tight. A Glycine-Proline bond will combine features of each. In this way the key contribution to the backbone made by each amino acid is its impact on the flexibility of the peptide bond it forms with a given partner at a certain angle and major conformation. Residue identities play other roles with respect to folding a protein, of course, but Rafiki sees this as the key role of each functional group in determining the overall shape of a protein. The amino acid identities can determine just how ‘locked-in’ at a certain angle that a peptide bond might be at the time it is formed.
Before we can begin to quantify actual bond information in the genetic code, we must first attempt to identify the mechanisms and components that can carry and translate this information. In so doing, we begin to see that a major reformation of our view of the genetic code is required. The first element pertains to what we define as the actual code, or more appropriately, where the code actually resides. The central dogma holds that information flows DNA » RNA » Protein.
Transcription is the process of moving information: DNA » RNA.Translation is the process of moving information: RNA » Protein.
The genetic code is about translation. Information is moved from nucleic acids to amino acids. DNA serves a storage function only and conventionally plays no direct role in translation.Messenger RNA is like a modified transcript of information stored in DNA – important, but not everything. It is far from the only nucleic acid molecule that participates in translation. Also integral to translation are transfer RNA (tRNA) and ribosomal RNA (rRNA) both of which must be considered active participants in translation. They are not merely mute cogs, they are active components of the genetic code, providing an additional voice in the translation process. Genetic information is directly communicated by these components during translation. This concept is easily lost when one focuses entirely on primary sequence and neglects primary structure.
To get a feel for what’s really happening on the shop floor, the actual action at the point of translation, one must imagine being an amino acid; we must visit the shop floor. If you were an amino acid in a living cell, what would you see at the moment of translation? Let’s use a colored chessboard to make the point visually. The board is a graphical representation of nucleotides, codons, amino acids and their relative water affinities. (Red = hydrophobic, blue = hydrophilic)
This graphic is equivalent to the standard Watson-Crick codon table. All the information in the table is in this graphic, but the table relies on letters and words whereas this graphic uses shapes and colors. There is something radically wrong with both – they are each misleading. Why?