SIGNAL CONVERTER PROTEINS - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 448

TRAIL-sensitive cells. Furthermore, the discovery that
thrombospondin-1 acts via CD47 to inhibit nitric oxide
signaling throughout the vascular system [142] might
even connect SIRP a -containing SCP to vascular physiology,
and in particular, tumor vasculature. Going even further,
complementary fusions could be contemplated that tap into
the opposing “eat me” signaling system—surface calreticu-
lin binding to CD91 (LDL receptor-related protein)—that
sends a prophagocytic signal to macrophages and serves to
activate them [143].

SCP capable of downmodulating regulatory T cells is one
of them. Since regulatory T cells put the brakes on effector
T cells, the latter can be unleashed using SCP capable of
downmodulating the generation and suppressive activity of
regulatory T cells. Such regulatory T-cell-inhibiting SCP
would be especially useful in the context of cancer immu-
notherapy, given that tumor-infiltrating regulatory T cells
interfere with antitumor effector T-cell immunity within
tumor beds [147-150].
In designing regulatory T-cell-directed SCP, two TNF
superfamily ligand:receptor pairs stand out: GITRL:GITR
and OX40L:OX40. Intriguingly, each of these signaling axes
has been ascribed dual, complementary functional capacities,
effector T-cell activation, and regulatory T-cell inhibition.
That is, molecular agonists for GITR (glucocorticoid-induced
TNF receptor; TNFRSF18) can coordinately costimulate
and coinhibit effector and regulatory T cells, respectively
[151-153]. The same holds true for OX40 receptor agonists
[154,155]. Our group posited that an anti-GITR Ab

30.11 IMMUNE ACTIVATING SCP

Although the focus has been on inhibitory signaling, SCP can
be readily developed for cellular activation. To transition from
T-cell inhibition to activation, the most straightforward
approach is to reformulate T-cell-directed TSCP from a
CoSR

CoS)
configuration. To generate these reconfigured fusion proteins,
one can draw upon the same menu of CoS:CoSR and CoI:
CoIR molecular pairs. For instance, the inhibitory CoSR

CoI to a coinhibitor receptor costimulator (CoIR OX40L

fusion protein, coupling an agonistic anti-GITR antibody
derivative to OX40 ligand, might be especially effective in
eliciting coordinate and synergistic effector T-cell-activat-
ing, regulatory T-cell-inhibiting effects. In addition to
generating a bidirectional, cis loop-back auto-signaling
loop between the GITR and OX40 receptors, dual-signal-
ing anti-GITR Ab

CoI

SCP CTLA-4 FasL can be refashioned into an activating

CoIR CoS SCP by again tapping into the FasL:Fas trans

signaling pathway, this time incorporating Fas as a CoIR
component.
In flipping from CoSR

OX40L also has the potential to function
as a trans signal converter. That is, its binding to GITR on
T cells would interfere with GITR's outward looking
ability to trigger reverse signaling through GITR ligand
on plasmacytoid dendritic cells, which otherwise elicits an
IDO-dependent immune-inhibitory effect [156]. We have
substantiated that anti-GITR Ab

CoS, however, full
mirror imaging is not always feasible. There may be binding
asymmetries peculiar to given costimulator and coinhibitor
pathways that preclude simplistic CoSR-to-CoS and CoI-to-
CoIR reversals. Such asymmetries can arise, for example,
when a single surface ligand has affinity for both a CoSR and
a CoIR, a case in point being the two B7 proteins (CD80 and
CD86), each of which has affinity for both the CoSR CD28
and the CoIR CTLA-4. Yet, even in such cases, the pathway
can still often be invoked for SCP design by considering
granular details pertaining to differential affinities, expres-
sion levels, dynamics of expression, and cellular distribu-
tions. The CD80/CD86 dyad represents just such a situation,
where differences across these categories influence how
interactions with the CD28 versus CTLA-4 receptors play
out [144-146]. Even more complicated situations can gen-
erate asymmetries, such as when a single surface protein
(such as HVEM, TNFRSF14) can function as both a CoI (as
a DC surface protein binding to the inhibitory BTLA
(CD272) and CD160 receptors on T cells) and a CoSR
(as a T-cell surface protein receiving DC-derived LIGHT
costimulator signals) [112,113]. These kinds of complexities
demand significant experimental empiricism for validating
net functional outputs.
While morphing CoSR

CoI to CoIR

OX40L does indeed have
the ability to inhibit the generation and suppressive activity
of CD4 þ CD25 þ FoxP3 þ regulatory T cells, while simulta-
neously costimulating effector T cells.
Another path to T-cell activation invokes fusion proteins
with the capacity to promote the maturation and T-cell
stimulatory capacity of DC. Here too, there are a variety
of fusion protein options, exemplified by anti-CD40 Ab

OX40L, which couples an agonistic anti-CD40 antibody
derivative to OX40 ligand. This chimeric protein features a
number of potential mechanisms of action. Functioning as a
trans bidirectional signaling agent, bridging DC and na ıve
T cells, it has the potential to promote DC maturation (via its
agonistic anti-CD40 Ab end) and simultaneously costimu-
late effector T cells and inhibit regulatory T cells (via its
OX40 ligand end). This chimeric protein also features trans
signal conversion potential, in that it can competitively
block CD40-driven, reverse signaling through CD40L-
bearing monocytes, which otherwise induces a form of
monocytic activation-induced cell death. CoS is the most

direct approach for devising signal converters geared to
T-cell activation, there are other immune-activating
SCP options, some of which would function indirectly.

CoI into CoIR Interestingly,

anti-CD40 Ab OX40L, like various other SCP that we

originally configured with immune modulation in mind,
has anticancer therapeutic possibilities as well. This follows