90 5 Functional Requirements in the Program and the Cell Chassis for Next-Generation Synthetic Biology
solvent, essential for the development of life as we know it [50]. Proteinogenic
amino acids are hydrophilic, amphiphilic, or hydrophobic. Surprisingly, proline
is not an amino acid but an imino acid, and this has considerable consequences
for translation, with requirement of a specific elongation factor [51]. As building
blocks of proteins, these molecules need to be activated as aminoacyl adenylates
and loaded onto the 3’OH extremity of the ribose of an acceptor tRNA molecule.
This process introduces considerable constraints in the selection of amino acids
relevant to translation: for example, ornithine, homoserine, and homocysteine
will cyclize during the process and create toxic dead‐end compounds [35, 52].
Norleucine and selenomethionine can substitute for methionine [53], and this
changes activity only in a restricted number of proteins but is deleterious when
methionine is limiting as the methionine side chain is specifically used in some
metalloproteins, for example [54]. 2‐Aminobutyrate is an analog of cysteine, and
its concentration must be stringently controlled as it mimics cysteine metabolism.
Furthermore, the presence of non‐proteinogenic amino acids in cells implies that
they are prone to affect negatively translation accuracy via their wrong incorpo
ration into proteins. Hence the cell must cope with this hurdle either by main
taining a very low level of non‐proteinogenic amino acids or by modifying them
(generally by N‐acylation and sometimes N‐methylation) so that they do not
enter the wrong pathway [55].
Interestingly, this ubiquitous protection pathway (in the sense given by engi
neers in organic chemistry) is likely to have been recruited for other helper
functions such as further protection or regulation. For example, ribosomal pro
teins are generally acylated, but the exact function of the modifications remains
unknown, except for an obvious coupling with metabolism and a protective
role for reactive amines [56, 57]. In the case of nucleotides, it may well be that
formation of the triphosphates, besides providing a way to drive forward bio
syntheses via pyrophosphate hydrolysis, plays also the role of a recognition
group, resulting in the selection of a subset of nucleotides for insertion in
polynucleotides.
Another mode of metabolism organization derives from a distinctive match
between matter and space constraints. Carbon chemistry allows formation of a
considerable number of specific stereoisomers (remember Pasteur’s exclama
tion) that are recognized by enzyme cavities in a highly space‐constrained way.
Proteinogenic amino acids are of the l‐type. As a consequence most proteases
and peptidases are active on chains made of these stereoisomers. This opened
up the possibility of a selection pressure, leading to protease‐insensitive protec
tive structures that evolved toward containing the d‐isomers (e.g., antibiotics
[58]). Another most important selection pressure is on compounds that are sim
ilar to amino acids, with a hydroxyl group in the place of the alpha‐amino group,
making them good mimics of amino acids. For example, glycerate is quite simi
lar to serine and could take its place in many enzymes, an unwanted stereo
chemical toxic property. d‐Glycerate is therefore the preferred stereoisomer
[59]. This has consequences on the make‐up of nucleic acids. Because of the link
between the latter metabolite and those involved in glycolysis/gluconeogenesis,
this stereochemical constraint explains why most biologically relevant carbo
hydrates, ribose, and deoxyribose, in particular, are of the d‐isomer type [60].