Learning from Proteins and Drugs: Receptors That Mimic Biomedically Important Binding Motifs - Designing Receptors for the Next Generation of Biosensors - page 34

1 Introduction: Weak Interactions

Weak interactions determine protein size and shape and are therefore an essential
part of normal protein function. Discreet binding pockets and motifs that have
evolved to be highly selective for only a very particular class of substrates come
about as a result of a myriad of these non-covalent interactions. The helical shape of
DNA is so because of an intricate combination of H-bond donor and acceptor pairs,
stacking between the bases and solvation effects. Weak hydrogen bonding, electro-
static, and hydrophobic interactions play roles in all protein-substrate and
protein-protein interactions.
The most commonly observed weak interactions and arguably most important for
normal protein function are hydrogen bonds. Ubiquitous in complex natural systems,
almost all biological processes involve hydrogen bonding in some form or another
and these interactions have been the topic of extensive study for decades [ 1 - 6 ].
Proteins which act on anionic substrates universally contain highly evolved hydrogen
bond networks in their active sites. Figure 1a depicts the pore of a ClC chloride
channel whose crystal structure was solved in 2002 [ 7 ]. The ion is coaxed into the
pore by four key hydrogen bond donating residues. These attractive interactions pull
the chloride in close proximity to an aspartic acid residue which is displaced, thus
opening the ion channel. During drug design, medicinal chemists often seek to
emulate the natural substrate of a biological target and attempt to preserve all
attractive forces in the host-guest complex. Replacement of the phosphate linker in
natural RNA with an acylsulfonamide in a simple dinucleoside mimic (2)preserveda
key H-bond with His119 and resulted in inhibition of RNase A (Fig. 1b, c )[ 8 ].
Other essential, yet not as well understood non-covalent interactions present in
biological systems are those involving aromatic residues [ 9 - 12 ] and the hydropho-
bic effect [ 13 - 16 ]. The former is often considered to be a result of the latter, as
hydrophobic aromatic residues pack close together in the core of the protein while
hydrophilic residues are more often observed near the protein surface [ 16 ]. Although
disfavored on the basis of configurational entropy, as a folded protein loses huge
numbers of degrees of freedom relative to its unfolded state, the favored enthalpic
gains of water molecules able to hydrogen bond to each other causes the overall
energy of the system to be favorable. As a direct result, hydrophobic aromatic
residues are often seen stacked on one another. Doubly mutating two stacked
tyrosine residues in bacterial ribonuclease (barnase) resulted in a 4.6 kcal/mol
decrease in protein stability (Fig. 2a )[ 9 ]. Because many hydrophilic residues reside
near protein surfaces, salt bridges and cation-pi interactions are often seen at
protein-protein interfaces [ 3 ]. The latter is sometimes observed as a quaternary
ammonium residue bound inside an electron-rich pocket heavily populated with
aromatic residues. These binding sites, or “hotspots,” contain highly preorganized
tryptophan, phenylalanine, and tyrosine residues which stabilize the incoming
cation through their electron-rich pi-clouds. An important interaction of this type
is between methylated lysine and the CBX class of proteins, when methylation of
lysine is misregulated, disease often follows [ 17 - 20 ]. Figure 2b depicts the co-