Introduction
Protein engineering experiments involving fusion of different
proteins/domains that do not naturally adjoin each other forces
biotechnologists to have to choose suitable linker peptides, which can
vary in length, complexity, and conformational characteristics. The
success or failure of the fusion experiment is determined by the choice
of linker peptide, for two primary reasons: (i) the linker can affect
the function, conformation, and/or conformational stability of flanking
proteins/domains; (ii) the linker can itself be naturally susceptible to
undergoing proteolytic degradation by proteases present in the
environment, or become susceptible to proteolysis (as a result of the
influence of its flanking proteins/domains). In the work presented here,
our objective was to focus upon differential proteolytic
susceptibilities of different types of linkers, without distractions
involving either (a) the effects of linkers upon conformations of
flanking domains, or (b) effects of flanking domains upon conformations
of the linkers, to the extent possible. Therefore, the approach we
adopted was to use thermostable flanking proteins/domains that typically
display autonomy of folding, as well as autonomy of conformational
stabilization, in addition to high (natural) resistance to proteolysis
[which is a natural corollary, because high thermal stability
translates into a better-folded state, leading to lower scope for
proteases to gain access to peptide bonds (Mukherjee and Guptasarma,
2005)]. In experiments presented below, through the use of
thermostable flanking domains, we reduced the likelihood of proteolytic
degradation in flanking domains. This allowed us to examine only the
linker peptide for proteolytic degradation.
Two thermostable domains, (i) Coh2, a Type I cohesin fromClostridium thermocellum CipA, and (ii) BSX, a Bacillus
sp. Xylanase, were joined. The structure of BSX consists of a single
(β/α)8-fold barrel (Manikandan et al ., 2006). The structure of
Coh2 consists of a nine-stranded barrel with jelly-roll fold topology,
consisting of two flattened β-sheets (Carvalho et al ., 2003).
Coh2 and BSX were fused with the use of five different linkers: a rigid
linker, a flexible linker, and three linkers derived from the sequence
of a 42 residues-long (native) linker joining the C-terminal end of
Coh2, in Clostridium thermocellum CipA, to the N-terminal end of
an adjoining carbohydrate binding module (CBM). Below, we show that the
linker that is the most structurally-flexible and unstructured is also
the linker that maximally escapes proteolysis. A systematic elimination
of possibilities indicates that this counter-intuitive result owes to
the linker’s facilitation of motions in its flanking domains. The
motions appear to negatively affect the ability of proteases to approach
the linker’s (otherwise scissile) peptide bonds.