Biomedical Engineering Reference
In-Depth Information
proteins, and likely reflects the need for tight regulation
of signaling.
Both naturally occurring and engineered soluble receptors
were initially shown to block cytokine action in vitro,by
effectively acting as decoys. However, in vivo, they failed to
act efficaciously because of relatively low affinity for their
cognate ligand(s) and short in vivo half-lives. These limita-
tions were first shown for the monomeric tumor necrosis
factor receptor (TNFR) ectodomain [13,17,46]. Hence, it
became evident that pharmacologic application of soluble
receptors for inhibition of their cognate ligands in vivo would
either require repeated dosing, or they would need to be
engineered to improve their pharmacokinetic properties. A
time-tested solution to this problem was provided by assem-
bling recombinant hybrid proteins composed of receptor
ectodomains fused to the nonvariable regions of antibodies,
as was initially shown for CD4. In this pioneering work, the
ectodomain of CD4 was fused either to the constant region of
human IgG1 [47] or to the Fc region of mouse IgG2a [48], to
generate “immunoadhesins”—that is, hybrid proteins com-
posed of a receptor ectodomain (or part thereof) and the
constant or Fc region of an immunoglobulin, now commonly
referred to as “receptor-Ig” or “receptor-Fc” fusion proteins.
The resulting hybrids were high-affinity blockers of HIV
infection in T cells. Although these blockers of HIV did
not materialize into therapeutics, they demonstrated several
key properties of receptor-Ig or -Fc fusions:
abounds with such examples. Their use has been more
limited in the clinic, in part reflecting the time and effort
it takes to develop new therapeutics.
9.2 ETANERCEPT AS A PROTOTYPICAL
RECEPTOR-Fc-BASED CYTOKINE BLOCKER
The initial work on CD4-Ig fusions was rapidly followed by
the generation of TNFR-Fc [46,52], a high-affinity blocker
for TNF-
a
and TNF-
b
/lymphotoxin (Figure 9.1a). The idea
of developing soluble TNFR as a therapeutic arose from a
series of preclinical observations that demonstrated a path-
ologic role of TNF-
a
in several inflammatory diseases [54].
The initial clinical proof-of-concept that a TNF-
a
blocker
can halt the progression of pathologic inflammatory pro-
cesses and ameliorate disease was procured using the anti-
TNF-
a
antibody infliximab, and concurrently followed by
TNFR-Fc (etanercept). A detailed perspective of the devel-
opment of etanercept and other TNF blockers can be found
in several excellent reviews on the subject, some of which
also offer a historical perspective [54-58]. The success
of etanercept—it remains one of the most widely used
protein-based therapeutics [59]—has been followed by other
receptor-Fc fusion proteins as well as Traps with the aim of
engineering blockers for their cognate ligands. Some of
those that are entering or have completed clinical develop-
ment or have been approved are listed in Table 9.1.
1. Receptor ectodomains can be fused to Ig domains
without adversely affecting the function of the result-
ing hybrid proteins, while improving their expression
[47-49].
2. Unlike their soluble receptor counterparts, receptor-Ig
or -Fc fusions display long in vivo half-lives.
3. Receptor-Ig or -Fc fusions have antibody-like proper-
ties that retain at least some of the functions of anti-
bodies, such as binding to Fc receptors including
FcRn.
4. Desired effector functions can be potentially engi-
neered into these fusions by choosing the identity of
the Ig subclass constant region (along the lines of what
is done with recombinant therapeutic antibodies [50]).
9.3 HETEROMERIC TRAPS FOR LIGANDS
UTILIZING MULTICOMPONENT RECEPTOR
SYSTEMS WITH SHARED SUBUNITS
Most of the receptor-Fc-based ligand blockers that have
been designed to date have relied on single extracellular
domains (Table 9.1, Figure 9.1) dimerized via the Fc portion
of immunoglobulins, mainly human and mouse IgG (and the
latter mainly for research purposes). However, many recep-
tor systems are composed of more than one receptor subunit.
Thus, they do not lend themselves to simple dimerization of
a single extracellular domain in order to generate a blocker,
at least not one that displays both high affinity and specific-
ity for its cognate ligand(s). For the purposes of this review,
the different multicomponent receptor systems can be parsed
into two broad categories based on the properties of the
subunits comprising the signaling complex:
Historically, receptor-Ig, receptor-Fc, and other variations of
Ig-tagged fusions have been utilized [48,51], but the design
that has prevailed is the receptor-Fc version. Although other
domains and chemical moieties have been used as “carriers”
to engineer proteins with improved pharmacokinetic pro-
files, it is mainly peptide-Fc fusion proteins that have
enjoyed widespread use in research applications. In fact,
tagging proteins with Fc is one of the most commonly
utilized methods employed for studying growth factor
and cytokine function in vitro and in vivo, and the literature
1. Heteromeric Receptor Systems that Utilize Two
Signaling Receptor Subunits. Examples include the
type I and type II receptors for TGF-
b
and
bone morphogenetic proteins (BMPs) [31,116] or
IL-1-RI (IL-1R1) and IL-1-RAcP (IL-1RAP) for
IL-1 (Figure 9.2); a similar concept applies to a
Search WWH ::
Custom Search