SIGNAL CONVERTER PROTEINS - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 449

from its inherent ability to bridge in trans CD40 þ tumor cells
and tumor-infiltrating T cells, leading to tumor cell apopto-
sis, enhanced killing by tumor-directed cytotoxic T lympho-
cytes, and inhibition of local regulatory T cells. These
functional properties are anticipated on the basis of findings
that CD40 engagement on carcinomas and some B-cell
malignancies results in direct antiproliferative and apoptotic
effects, due to upregulation of Fas and TRAIL receptors
[157]. As an aside, this also means that anti-CD40 Ab
OX40L might also be expected to sensitize tumor cells to
Fas- or TRAIL-containing pro-apoptotic SCP.
The constellation of signals amenable to conversion goes
well beyond those that serve as activators, proliferators,
differentiators, inhibitors, or inducers of cell death. Clearly,
there is room for much creativity along these lines. Addi-
tional categories of signals that are targetable by SCP are, by
example, addressins, adhesins, and cell surface-associated
chemokines and their receptors.

prolonged SCP expression in the central nervous system by
this approach. There may be substantial advantages in using
intrathecal administration for screening SCP candidates rele-
vant to autoimmune disorders of the central nervous system.

30.13 SCP FRONTIERS: MINING THE SURFACE
PROTEIN INTERACTOME, REWIRING
CELLULAR NETWORKS

The SCPs described in this chapter were developed in an
eclectic manner and represent but a small sampling of those
that might be generated. A more deliberate and informed
approach toward expanding SCP libraries would be empow-
ered by a systematic cataloging of surface proteins of
therapeutic relevance. Indeed, the most comprehensive
approach would draw upon a cell surface protein interac-
tome (CSPI), which maps protein interaction networks. The
Human Interactome Database constitutes one platform being
developed to visualize and integrate information pertaining
to protein interaction networks and disease associations for
each protein in the human proteome [158,159]. A fully
annotated CSPI, narrowing in on the subset of proteins
relevant to cell contact-dependent interactions and providing
cell distribution and protein function data for them, would
substantially advance SCP design. In filtering the large
matrix of potential SCP protein pairs that would emerge
from such an exercise, one would conceptually navigate
across several dimensions: interactive cell pairs or clusters
linked to specific pathogenic processes; cell surface molec-
ular pairs of special significance for the functional interplay
among these pathogenic cell groupings; choice of SCP
functionality most appropriate for targeting the cell surface
molecules and cellular interactions of interest.
In thinking about SCP function in vivo, one per force
enters a realm of higher order complexity, as one marries
multifunctional SCP with multicellular networks. As dis-
cussed throughout this chapter, individual SCP commonly
reveal multiple functionalities and mechanisms of action,
even in the simplest of contexts where there are but one or
two cell types in play. As one adds more cell types to a given
interactive mix, the functional consequences ripple outward
and are amplified. Furthermore, additional systemic com-
plexity emerges when one considers the wide cellular dis-
tribution of some of the cognate counter-receptors that are
targeted by SCP. This in effect can bring additional cellular
networks into the fold. To decipher changes in the net output
from SCP-perturbed cellular networks, more sophisticated
readouts will be called for, invoking ever more multiplexed
analytic tools, for example, cytometric fingerprinting [160]
and high-dimensional microfluidic dynamic arrays [161].
Yet, despite the potential complexity of SCP:network
overlays, they create new therapeutic opportunity and can
even streamline interventions. By targeting critical nodes

30.12 EXPERIMENTAL TOOLS FOR SCREENING
SCP CANDIDATES

Given the breadth of SCP possibilities, robust experimental
approaches for vetting fusion protein candidates become
essential. Protein production poses an experimental bottle-
neck. Since producing panels of candidate SCP proteins for
functional screening is simply too cumbersome, gene trans-
fer represents a preferable option. One experimental path,
which we have deployed with success, starts with in vivo
gene transfer as a gateway for achieving functional valida-
tion and garnering first-order efficacy information that ena-
bles early prioritization of SCP candidates.
By example, we have recently leveraged a robust trans-
poson-based mammalian expression system for in vivo SCP
evaluation [80]. This expression system enables systemic
expression of the fusion protein candidates, with sustained
serum levels over months.
As an alternative to transposon-based systemic expression
of SCP, one can turn to local gene delivery for SCP functional
screening. In this case, functional readouts that assess local
SCP effects in a straightforward manner are essential. This is
exemplified by our coupling of intradermal naked SCP-
encoding DNA transfer with delayed type hypersensitivity
suppression as a functional readout for local T-cell immuno-
modulation. This system has allowed us to screen the immune
inhibitory potential of several new CosR CoI TSCP candi-

dates. This gene transfer-based, rapid functional screening
approach pointed to OX40 TRAIL as a particularly interest-

ing SCP candidate for memory T-cell inhibition.
Intradermal injection is but one of multiple local gene
delivery options. Another is intrathecal gene delivery. We
have developed the capability of administering SCP expres-
sion constructs intrathecally in mice, demonstrating