Chromosomal translocations that combine distinct functional domains of unrelated proteins are an experiment in nature. They demonstrate how endogenous regulatory checkpoints can be overridden by altered cell biochemistry, informing a means to engineering an aberrant signal that the cell is incapable of counterregulating. Thus, our laboratory and others have synthesized fusions of GM-CSF with peptides, ILs, and chemokines, which we have termed fusokines, with the aim of inducing an enhanced immune response against cancer, aiming to overcome the maladapted biological processes causing disease. In doing so, we found that these fusokines did not behave as merely the sum of their natural unfused counterparts, but as entirely novel ligands co-opting their cognate receptor to communicate a unique message to responsive cellular targets. In this review, we discuss how fusion proteins combining different bioactive ligands can alter immune responses and briefly discuss the regulatory pathways that they circumvent.

Portions of genes can be combined to augment a protein’s normal activities without conferring novel functions unto them. For example, several oncogenic translocations include the IgH gene; when this is combined with Bcl-2 in the (14;18) translocation, the cells will no longer undergo apoptosis, an essential step down the road toward cancer (1). The (9;22) translocation is different from the (14;18) translocation. When the (9;22) translocation was first linked to chronic myelogenous leukemia in 1973 (2) the mechanism was not clear; however, further studies found that the mutation had functional implications in initiating and sustaining chronic myelogenous leukemia (2). Indeed, the translocation aligned the serine–threonine kinase of BCR with the C-terminal tyrosine kinase of Abl, generating a constitutively active tyrosine kinase lacking normal regulatory domains (3). This resulted in the activation of the mitogenic and prosurvival Ras and Erk pathways, leading to cdk2 expression and the initiation of the G1-S transition (4). Such naturally occurring fusion proteins and their dramatic effects are powerful examples that highlight the key weakness to how cells function: although the ways in which different cells are regulated are complex, they are also rigid. The introduction of a different kind of signal that a cell did not evolve to regulate is equivalent to throwing a wrench into its gears, because the cell is inflexible and incapable of adapting.

All oncogenic fusion proteins found in nature are membrane-bound or cytoplasmic tonic-signaling molecules. There are no known naturally occurring incidental oncogenic fusions of extracellular ligands or cytokines. Notwithstanding, the notion of fusing extracellular ligand proteins, such as leukins, with potent signaling properties is particularly appealing for the development of pharmaceuticals. Indeed, the field of “fusion proteins” as drugs rests upon the development of compounds that can be administered as an exogenous substance acting upon defined target tissues. Different groups have synthesized artificial fusion proteins as immunomodulators in an attempt to prevent the development and progression of autoimmune diseases and cancer. In this review, we discuss a few of the strategies that have driven the creation of fusion proteins borne of the marriage of two bioactive ligands, which alter the immune response and function, and briefly discuss the regulatory pathways that they could circumvent.

Cytokines have been artificially combined with Abs to increase their half-life, without neutralizing their ability to signal through their receptors. The objective is to reduce systemic toxicity to patients by increasing the local concentrations of these cytokines while minimizing how much needs to be given to patients to achieve therapeutic efficacy. In these situations, although the cytokine functions normally, how the cells are exposed to the cytokine changes. For example, IL-3 has a half-life in the range of 5–15 min, the duration of which was extended to several hours by combining it with an anti–IL-3 Ab (5). Similarly, IL-2, IL-12, and GM-CSF have been fused with Abs specific for different cell surface markers, such as Her2, to force the cytokines to localize to the site of the tumor (6). In murine studies, doing so significantly increased the effectiveness of the Abs at reducing the rate of tumor growth in different tumor models.

Alternatively, a cytokine can be combined with a ligand for a receptor that is overexpressed by a tumor to force its colocalization. Although this protein was not based on GM-CSF, the idea behind the epidermal growth factor (EGF)/IL-18 fusion was that the EGF moiety would act as an anchor, allowing for the colocalization of IL-18 to EGFR+ tumors (7). Mice treated with EGF/IL-18 and IL-18 intratumorally had reduced tumor growth; once removed from the animals, significantly more tumor cells from the EGF/IL-18 and the IL-18 groups were shown to be in cell-cycle arrest (8).

Although the concept behind EGF/IL-18 was interesting, because the proteins were injected intratumorally the investigators did not assess whether the EGF moiety conferred the IL-18 moiety any tumor tropism in vivo. It would be interesting to know whether the EGF receptor signaling was disturbed by the fusion protein, which would be a tremendous find, or whether the EGF moiety could outcompete the EGF by the local tumor stroma. Answering questions about EGF’s functionality is important, because the EGF/IL-18 fusion is unusual in that it is a protein that attempts to ultimately block proliferation, in part by using a growth factor.

GM-CSF/prostate acid phophatase to stimulate an anticancer immune response.

Dendreon (NASDAQ: DNDN) patented a fusion protein of GM-CSF with prostate acid phophatase (PAP) to break tolerance against PAP-expressing prostate cancer (9). Provenge (sipuleucel-T) is an autologous cellular vaccine generated from dendritic cell (DC) precursors isolated from the patient’s peripheral blood by leukapheresis, meaning that the cells are activated ex vivo in the presence of the GM-CSF/PAP fusion and administered back to the patient. The idea is to provide the immune system with a target Ag, in addition to instructing the immune system about how to respond to this Ag. In a phase III trial published in 2010, the treatment increased average patient survival by 4.1 mo over placebo, had a 22% decreased risk for death, and was well tolerated, with <1% of patients withdrawing from treatment because of infusion-related adverse events (10).

IL-16/myelin basic protein and GM-CSF/myelin basic protein to block an autoimmune response.

Instead of using cytokines to break tolerance, Mannie and Abbott (11) combined myelin basic protein (MBP) with IL-1Rα, IL-2, IL-10, IL-13, or IL-16 to create a fusion that could induce Ag-specific tolerance to MBP in the experimental autoimmune encephalomyelitis (EAE) model of multiple sclerosis in rats. The strategy was similar to the one used by Dendreon, in which the immune system was provided with an Ag and a mechanism to influence how the immune system should respond to it. They found that IL-16 was the most effective cytokine to combine with MBP and that the fusion of the two was necessary to induce tolerance by significantly reducing the disease scores of rats with EAE. Although the study did not provide an in-depth understanding of how the proteins functioned or how the immune response was modulated, it is likely that each fusion had a distinct mechanism of action. With regard to the IL-16/MBP fusion, IL-16 was previously found to inhibit IL-2 production by CD4+ T cells by interfering with CD4’s interaction with MHC class II (12, 13). It is unclear how this could induce Ag-specific tolerance, but the investigators speculated that APCs may be involved, because the IL-16/MBP fusion was effective at having APCs present MBP through MHC class II to Rsl.11 T cells (11). Blanchfield and Mannie (14) further explored how cytokines could be fused to more effectively deliver MBP to APCs by combining it with GM-CSF, finding that this fusion was also capable of significantly reducing the severity of EAE in rats. The investigators believed that the GM-CSF moiety acts as a delivery vehicle for MBP, because the two need to be linked for the fusion to suppress EAE.

It is interesting that a single cytokine fused with two different Ags could induce such contradictory results. With Provenge, the GM-CSF/PAP fusion is applied to a purified population of autologous DCs ex vivo to induce an inflammatory response against prostate cancer. It is unfortunate that the MBP studies did not delve further into the cellular biochemistry of the fusions, because it is unusual that the GM-CSF/MBP fusion would induce the differentiation of functional, Ag-presenting DCs from bone marrow monocytes and yet serve a tolerogenic role in vivo. Although the GM-CSF moiety is functional, it is possible that the GM-CSF/MBP fusion interacts with a different GM-CSFR+ population of cells than the DC precursors isolated for Provenge, although this population has yet to be identified. This discrepancy highlights an important nuance in immunotherapy, which is that the therapeutic agent is not the cytokine or the fusion, but the cells with which they interact, where even related cell populations can interpret a signal to mean very different things in different contexts. As seen with Provenge, manipulating the cells ex vivo prior to administering them back to the patient allows for the treatment to be standardized and limits the potential for external factors to influence the differentiation process and the ultimate function of the cells.

The cellular signals launched by the GM-CSF common β-chain and the IL common γ-chain receptors and their regulation.

Jak are recruited upon receptor oligomerization of several receptor families, including, but not limited to, the common β- and γ-chain receptors. They phosphorylate the cytoplasmic tails of their receptors to form SH2-docking sites for STAT proteins, which when phosphorylated, dimerize to regulate gene expression in the nucleus (15). Jak1 and Jak2 are widely expressed and associated with a diversity of receptors. Jak2 is responsible for signaling downstream of the common β-chain and is indispensable for maintaining hematopoiesis. Jak3 is associated with the common γ-chain and the phosphorylation of STAT5, and it regulates cell survival by inducing the expression of Bcl-2 and suppressing Bax in T cells (15, 16).

The suppressor of cytokine signaling (SOCS) proteins have short-term and long-term effects on mediating inflammation by acting as negative-feedback regulators for different signals, with the ultimate consequence of strongly influencing DC and T cell differentiation. The SOCS proteins are essential in regulating the Jak–STAT pathway, because they can compete with STATs for SH2-docking sites on receptors, bind to Jaks to prevent further receptor phosphorylation, and act as ubiquitin ligases that target Jaks for degradation (17, 18).

The GM-CSFR common β-chain is the signaling component for the GM-CSFR, IL-3R, and IL-5R dimers. Because different cell types express the common β-chain receptors, which population responds to these cytokines depends on the α subunit. The α-chains are high-affinity binding proteins that confer specificity to the common β-chain (19). To signal, the receptor dimerizes with another αβ complex, and the β-chains interact to induce proliferation (20) through Jak2/STAT5 and the MAPK pathway. In an inflammatory environment, GM-CSF signaling counterbalances IFN-γ’s antiproliferative effects on T cells by antagonizing STAT1 (21, 22), in addition to promoting the mobilization of monocytes, the maturation of DCs, and the survival and activation of macrophages and granulocytes (23, 24). SOCS3’s regulation of GM-CSFR signaling was made evident through how SOCS3−/− animals developed erythrocytosis and in how the transgenic expression of SOCS3 inhibited erythropoiesis. The SOCS proteins are further involved in mediating GM-CSF’s immunological functions, because SOCS1 and SOCS3 play a central role in mediating GM-CSF’s influence on DC maturation (18, 25).

pIXY 321: the fusion of GM-CSF with IL-3 to stimulate hematopoiesis.

pIXY 321 (aka pixykine), produced by Immunex Corporation (NASDAQ: IMNX), is a fusion protein of GM-CSF with IL-3. It was made because the combination of the two cytokines had complementary effects on different myeloid lineages, although GM-CSF and IL-3 signal through a shared common β-chain (also used by IL-5). It was demonstrated that this pixykine promoted superior immune, erythroid, and megakaryocytic reconstitution than the two cytokines did separately (26, 27). In vitro studies initially found that pIXY 321 was distinct from its unfused counterparts, because it had enhanced binding to GM-CSFR and IL-3R in competitive assays performed on cell lines. These altered interactions resulted in a 10–20-fold increase in the potency of the cytokines, inducing superior colony formation in assays specific to each moiety (28). When pIXY 321 was offered to patients in the initial clinical trials, the treatment was well tolerated with no observed dose-limiting toxicity, whereas it ameliorated thrombocytopenia and neutropenia in patients receiving chemotherapy. In these early studies, success depended on the dosage and the size of the patient’s available marrow reserve at the time of treatment (29, 30). Despite its promising initial results, pIXY 321 trials were discontinued because it did not prove to be superior to GM-CSF when the two were compared in a phase III trial in which patients were treated with 5-Fluorouracil, Leucovorin, Adriamycin, and Cytoxan for advanced breast cancer (31).

Because GM-CSF and IL-3 signal through STAT5 downstream of the GM-CSF common β-chain, pIXY 321’s simplicity may have been its downfall. With pIXY 321, it was not the message that the cell received that was altered, but its amplitude; pIXY 321 was intended to be a super GM-CSF/IL-3, and it successfully fulfilled that role, inducing a hyperproliferation of its responder cell lines. Ultimately, because the message itself was not different, it did not confer any additional properties to distinguish itself from its precursors; ultimately, it was not superior to GM-CSF in patients. Simply having more of the same signal did not change the cell’s ability to interpret and regulate it because the cellular machinery remained undisturbed.

The GM-CSF/IL fusion transgenes: GIFTs that keep giving.

The IL common γ-chain family consists of IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21 (32). They all share CD132, the IL-2R γ-chain, as part of their receptor complexes. IL-2 and IL-15 mostly differ by the time at which their receptors are expressed, because they are both involved in NK and CD8+ T cell development and cytolytic functions (33, 34). IL-15’s role is further downstream of IL-2, because it promotes lymphocyte survival by maintaining the expression of Bcl-2, whereas IL-2 promotes activation induced cell death (34, 35). However, IL-15 differs significantly from IL-21 in how it preferentially induces STAT5 over STAT3 and STAT1 and recruits Shc and ERK to the β-chain to promote cell survival (36). IL-21 modulates proinflammatory NK cell and CD4+ and CD8+ T cell functions and, unlike IL-2, it is an inhibitor of regulatory T cell function (3740).

GIFT-2: a hyperagonist ligand for IL-2R.

We originally engineered a fusokine linking as a single polypeptide GM-CSF to IL-2 (GIFT-2) (41). This was the first of a series of GM-CSF IL fusion transgenes (GIFTs), accomplished by linking GM-CSF to common γ-chain ILs. By combining GM-CSF with IL-2, the aim was to influence complementary elements of the myeloid and lymphoid immune systems to stimulate an anticancer immune response. To evaluate whether the distinct GM-CSF and IL-2 domains of GIFT-2 were functional, GIFT-2 was applied to the JAWSII and CTLL-2 cell lines to evaluate whether the fusokine could replicate GM-CSF’s and IL-2’s respective abilities to induce their proliferation. IL-2 was as effective as its unfused counterparts at maintaining these cells lines in culture; however, contrary to pIXY 321, GIFT-2 did not significantly enhance their proliferation (41). In vitro investigations into GIFT-2’s ability to mediate tumor rejection in mice found that GIFT-2–stimulated NK cells adopted a novel proinflammatory phenotype (42); these NK cells produced significantly more IFN-γ than did cells treated with GM-CSF, IL-2, or the combination of the two, and they were highly cytolytic and resistant to GM-CSF’s regulatory effects (43).

Considering the desirable pharmaceutical features of GIFT-2, we went about exploring the pairing of GM-CSF with IL-15 (GIFT-15) and IL-21 (GIFT-21) in the expectation of developing an entirely new class of immune-modulatory fusokines. Significant surprises were in store, which are best understood by defining the molecular response downstream of IL receptor binding to its synthetic ligand: GIFT.

GIFT-15: a partial hyperagonist for IL-15R with counterintuitive consequences.

As a second-generation GIFT fusokine, GM-CSF was linked with IL-15. The finding that GIFT-15 functioned as an immune suppressant was unanticipated, because the only differences between the IL-2R and IL-15R trimers are the α-chains, which only confer ligand specificity to the signaling complex formed by the β- and γ-chains that they share (44). Importantly, the biochemical properties of GIFT-15 were ascribable to IL-15R–driven hyperactivation of STAT3 coupled to hypoactivation of STAT5 on lymphomyeloid cells. Astonishingly, naive B cells responsive to GIFT-15 adopted an IL-10+, B regulatory phenotype and are key mediators of GIFT-15’s immune-suppressive capabilities. Indeed, GIFT-15–treated autologous B cells were used therapeutically to cure mice afflicted with EAE (45).

GIFT-21: a hyperagonist for IL-21R–expressing APCs.

As a third-generation GIFT fusokine, GM-CSF was linked with IL-21. We found that GIFT-21 had unanticipated proinflammatory effects on the monocyte lineage (46). The properties of these cells were investigated in a follow-up study in which bone marrow-derived monocytes were differentiated into DCs in vitro using GIFT-21 and acquired a hybrid conventional/plasmacytoid DC phenotype. The GIFT-21 DCs were injected into mice without any Ag priming and significantly inhibited tumor growth following their migration to tumors and their presentation of tumor Ags to CD8+ T cells in vivo (47).

Despite the overlapping functions of their source material, the GIFTs differed remarkably from one another in terms of their signaling and their net effects. GIFT-2 retained an IL-2–like phenotype, but with the hyperinduction of STAT1, STAT3, and STAT5 (41). GIFT-15 completely altered IL-15R signaling because it induced a hyperphosphorylation of STAT3 and a hypophosphorylation of STAT5 (44), and GIFT-21 induced a hyperphosphorylation of STAT3, relative to IL-21 (46). It is not clear why similar receptors as those for IL-2 and IL-15 induced such different responses when their respective ligands were fused with GM-CSF or why similar signals as those of GIFT-21 and GIFT-15 induced such different responses downstream of their effector cell populations. We speculate that GIFT-15’s higher affinity for the IL-15Rα–chain and the GM-CSF moiety have altered receptor-binding properties in terms of how GIFT-15 fits into the IL-15–binding cleft (44), but it is not obvious how this alters the phosphorylation of Jaks and STATs further downstream. GIFT-21’s differences from GIFT-15 may be similar to the difference between IL-6 and IL-10; both molecules signal through STAT3 but are differentially regulated by SOCS3 protein to different effects (48). The simplest explanation for the novel ligand properties of GIFTs is that the GM-CSF domain tethered at the N terminus of the polypeptide subtly alters the receptor-binding properties of the C terminus IL. An alternate biological effect may be related to half-life and biodistribution, considering that GM-CSF is a plasma protein with a half-life measured in hours and that ILs often have short half-lives (IL-15’s is <1 min) and distribution compartments. GM-CSF could be the big brother that takes the ILs for a long ride in new quarters in vivo.

In contrast to pIXY 321, the signaling pathways downstream of the GIFTs’ respective receptors are complex. The altered receptor–ligand interactions influenced multiple pathways. Therefore, the cells received entirely different messages through their receptors, inducing novel effector cell phenotypes with useful properties for treating autoimmunity and cancer. An important lesson that we learned from GIFTs is that it is difficult to predict what message a cell will receive as a result of these aberrant interactions or how different populations of cells will interpret it.

GMME more: GM-CSF meets chemokines.

Modulating the biology of lymphoid cells expressing CCRs may be of use for the treatment of autoimmune and infectious diseases. Chemokines guide the migration of cells by binding to glycosaminoglycans, forming a gradient that acts as a path for cells to follow (49). Cells respond to chemokines through G protein-coupled receptors (GPCRs), a unique set of cell-surface receptors that loop across the plasma membrane seven times (50). These receptors associate with different sets of G proteins through their intracellular domains: the Gα proteins are GTPases, which, in the GTP-bound state, dissociate from Gβγ proteins. Gα and Gβγ are involved in different signaling pathways, including STAT phosphorylation (51). Once the Gα protein hydrolyzes its GTP to GDP, it becomes inactive; this process can be accelerated by the regulators of G protein signaling, which are activated as part of negative feedback loops to GPCR signaling (52). Once activated, GPCRs recruit scaffolding proteins called arrestins; they associate with and, thus, modulate the function of other proteins, like ERK1/2, block G proteins from further interacting with the GPCRs, and initiate the internalization of the receptor as a means of desensitizing a cell to its signal (53). It is also interesting to consider how chemokines like CCL2 are regulated because their signaling is fundamentally different and because the way in which they are handled outside of the cell is also different. For example, IL-2 is metabolized and inactivated primarily by cathepsin D in the kidneys (54); this contrasts with CCL2, whose N-terminal amino acids are locally chewed away by matrix metalloproteinases in areas of inflammation, turning CCL2 into an antagonist for CCR2 instead of simply neutralizing it (55, 56).

Initial studies investigating the applicability of GM-CSF–chemokine fusions focused on MCP-1 (CCL2), because it plays a crucial role in the migration of immune cells to areas of inflammation (57, 58), as reflected in why CCR2−/− mice fail to develop EAE (59), and in cancer, where CCL2-recruited monocytes are locally converted to myeloid-derived suppressor cells (60). CCL2 was an ideal candidate to investigate because of its immunology and because it only binds to CCR2 (61).

When fusing GM-CSF with CCL2 (GMME1), Rafei et al. (62) found that, in contrast to CCL2, GMME1’s interaction with CCR2 induced robust intracellular calcium mobilization and the upregulation of Bax, but it did not induce STAT3 phosphorylation. Bioluminescence resonance energy transfer (BRET) studies of GMME1’s interaction with CCR2 found that GMME1 reduced the BRET signal below background resting levels, akin to an inverse agonist and while failing to recruit β-arrestin 2 to CCR2. This unusual signaling pattern and the likely failure of the cell to negatively regulate CCR2 through β-arrestin 2 had the unanticipated consequence of rapidly inducing the apoptosis of CCR2+ cells. To deliver GMME1, mesenchymal stromal cells (MSC) were generated from CCL2−/− mice, retrovirally transduced to express GMME1, and implanted in mice with rheumatoid arthritis and EAE. In each setting, the mice treated with GMME1-expressing cells had significantly reduced disease scores because GMME1 had selectively depleted CCR2+ inflammatory cells, whereas the unmodified CCL2−/− MSCs did not. In vitro recall-response experiments were performed, and the MSC GMME1-treated mice had significantly less myelin oligodendrocyte glycoprotein and type II collagen-reactive T cells. Further experiments also found significantly less cellular infiltrates in the CNS and joints of mice treated for EAE and rheumatoid arthritis, respectively (62, 63). Further research remains to be done with GMME1 against cancer to evaluate whether it can be a receptor-specific chemotherapeutic drug to interfere with the migration of monocytes to cancer and their differentiation into tumor-associated macrophages (60, 64).

Our laboratory is also investigating the fusion of GM-CSF and MCP-3 (GMME3) because it was shown that CCL2 and CCL7 (MCP-3) shared similar functions in regulating Ig production (55). The rationale for GMME3 was that if GMME1 could deplete CCR2+ cells, then GMME3 could potentially deplete more cell types, because CCL7 binds to CCR1, CCR2, and CCR5 (6567). Preliminary experiments disproved this hypothesis, and further investigation found that GMME3 instead could induce the differentiation of IL-10–producing B cells, an unanticipated twist to CCL7’s suppressive properties on B cells (J. Hsieh, P. Williams, M. Rafei, E. Birman, Y.K. Young, and J. Galipeau, manuscript in preparation).

Of all of the fusions discussed, GMME1 is the simplest to translate because it does not depend on the differentiation of a distinct, novel cell type to mediate its effects, but the depletion of CCR2+ cells. The demonstration that coupling GM-CSF with a GPCR ligand could gain such a novel function is significant, because it dramatically expanded the repertoire of potential fusion proteins. However, chemokine-based fusions are likely to be more demanding to study because CCLs can be promiscuous with regard to which CCR they interact with and because of how CCRs can heterodimerize.

When most fusokines were originally designed, they were expected to function as the sum effect of their components would naturally do. One strategy has been to force a fusion to localize to tumors (68). The use of Ab–cytokine fusions is another approach, but like with vaccine approaches, it is limited by the Ags to target. Furthermore, the effectiveness of a ligand in anchoring a second signal has yet to be clearly shown. Fusions were also made to provide the immune system with a target and a set of instructions to direct responses against this target, as seen with Provenge and the MBP fusion proteins. Cytokines have also been combined to deliver complementary messages to different population of cells, as was the case with pIXY 321, the fusion of GM-CSF and IL-3, to help patients recover from myelosuppression. However, combining cytokines rapidly becomes complicated when factors, such as the size of the linker, affect the molecule’s interactions with its receptors and furthermore, on the molecular mechanisms on which the molecule depends to functionally distinguish itself.

Interestingly, the GIFTs and GMMEs were based on GM-CSF, but the functions of these fusions were consistently not driven by GM-CSFR signaling. This is not to say that GM-CSF is not an important component of the fusion protein or that the GM-CSF moiety cannot modulate the immune system, but that GM-CSFR signaling has yet to be shown to be necessary for the effects of GIFTs or GMMEs. Because the function of GM-CSF in these fusions is mainly to alter its partner ligand’s interactions with its receptor(s), it would be interesting to see what would happen if GM-CSF were combined with molecules involved in signaling pathways outside of the immune system or if other growth factors, such as G-CSF, were used instead of GM-CSF. However, it would be difficult to anticipate how this would alter the partner ligand’s interaction with its receptor, because despite the similarities in nomenclature, G-CSF and GM-CSF are completely different molecules. Although these fusions were based on cytokines found in nature, the novel signaling, as well as the responses that the GIFTs and the GMMEs elicit, distinguishes them from their unfused counterparts. For this reason, we prefer to think of them as entirely new molecules that co-opt a set of receptors that are readily available on the surface of immune effector cell types.

The major limitation for the generation of new fusions is creativity. If we assume that there are >200 bioactive leukines, then there are >40,000 possible binomial combinations. A logical hypothesis-driven approach would be to couple functionally unrelated, yet potentially synergistic, partners and pay particular attention to the impact of altered half-life, biodistribution, and the affinity of each domain to its cognate receptor and subservient signaling. Alternatively, the coupling of inflammatory cytokines to decoy receptors that block suppressor checkpoints, such as TGF-β, or antiangiogenic mechanisms could also be attractive for cancer therapy.

The bounty of unheralded effects involving dramatic gain-of-function properties arising from fusokine interaction with target cells suggests that the synthetic coupling of functionally distinct cytokines is a profitable strategy to alter the immune response for improving clinical outcomes.

Abbreviations used in this article:

BRET

bioluminescence resonance energy transfer

DC

dendritic cell

EAE

experimental autoimmune encephalomyelitis

GIFT

GM-CSF/IL fusion transgene

GMME1

GM-CSF fused with CCL2

GMME3

GM-CSF fused with MCP-3

GPCR

G protein-coupled receptor

MBP

myelin basic protein

MSC

mesenchymal stromal cells

PAP

prostate acid phosphatase

SOCS

suppressor of cytokine signaling.

1
Iqbal
J.
,
Sanger
W. G.
,
Horsman
D. E.
,
Rosenwald
A.
,
Pickering
D. L.
,
Dave
B.
,
Dave
S.
,
Xiao
L.
,
Cao
K.
,
Zhu
Q.
, et al
.
2004
.
BCL2 translocation defines a unique tumor subset within the germinal center B-cell-like diffuse large B-cell lymphoma.
Am. J. Pathol.
165
:
159
166
.
2
Rowley
J. D.
1973
.
Letter: A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining.
Nature
243
:
290
293
.
3
Kurzrock
R.
,
Kantarjian
H. M.
,
Druker
B. J.
,
Talpaz
M.
.
2003
.
Philadelphia chromosome-positive leukemias: from basic mechanisms to molecular therapeutics.
Ann. Intern. Med.
138
:
819
830
.
4
Cortez
D.
,
Reuther
G.
,
Pendergast
A. M.
.
1997
.
The Bcr-Abl tyrosine kinase activates mitogenic signaling pathways and stimulates G1-to-S phase transition in hematopoietic cells.
Oncogene
15
:
2333
2342
.
5
Jones
A. T.
,
Ziltener
H. J.
.
1993
.
Enhancement of the biologic effects of interleukin-3 in vivo by anti-interleukin-3 antibodies.
Blood
82
:
1133
1141
.
6
Ortiz-Sánchez
E.
,
Helguera
G.
,
Daniels
T. R.
,
Penichet
M. L.
.
2008
.
Antibody-cytokine fusion proteins: applications in cancer therapy.
Expert Opin. Biol. Ther.
8
:
609
632
.
7
Lu
J.
,
Peng
Y.
,
Zheng
Z. J.
,
Pan
J. H.
,
Zhang
Y.
,
Bai
Y.
.
2008
.
EGF-IL-18 fusion protein as a potential anti-tumor reagent by induction of immune response and apoptosis in cancer cells.
Cancer Lett.
260
:
187
197
.
8
Lu
J. X.
,
Peng
Y.
,
Meng
Z. F.
,
Jin
L. Q.
,
Lu
Y. S.
,
Guan
M. X.
.
2005
.
Rational design of an EGF-IL18 fusion protein: implication for developing tumor therapeutics.
Biochem. Biophys. Res. Commun.
334
:
157
161
.
9
Small
E. J.
,
Fratesi
P.
,
Reese
D. M.
,
Strang
G.
,
Laus
R.
,
Peshwa
M. V.
,
Valone
F. H.
.
2000
.
Immunotherapy of hormone-refractory prostate cancer with antigen-loaded dendritic cells.
J. Clin. Oncol.
18
:
3894
3903
.
10
Kantoff
P. W.
,
Higano
C. S.
,
Shore
N. D.
,
Berger
E. R.
,
Small
E. J.
,
Penson
D. F.
,
Redfern
C. H.
,
Ferrari
A. C.
,
Dreicer
R.
,
Sims
R. B.
, et al
.
2010
.
Sipuleucel-T Immunotherapy for castration-resistant prostate cancer.
N. Engl. J. Med.
363
:
411
422
.
11
Mannie
M. D.
,
Abbott
D. J.
.
2007
.
A fusion protein consisting of IL-16 and the encephalitogenic peptide of myelin basic protein constitutes an antigen-specific tolerogenic vaccine that inhibits experimental autoimmune encephalomyelitis.
J. Immunol.
179
:
1458
1465
.
12
Ogasawara
H.
,
Takeda-Hirokawa
N.
,
Sekigawa
I.
,
Hashimoto
H.
,
Kaneko
Y.
,
Hirose
S.
.
1999
.
Inhibitory effect of interleukin-16 on interleukin-2 production by CD4+ T cells.
Immunology
96
:
215
219
.
13
Liu
Y.
,
Cruikshank
W. W.
,
O’Loughlin
T.
,
O’Reilly
P.
,
Center
D. M.
,
Kornfeld
H.
.
1999
.
Identification of a CD4 domain required for interleukin-16 binding and lymphocyte activation.
J. Biol. Chem.
274
:
23387
23395
.
14
Blanchfield
J. L.
,
Mannie
M. D.
.
2010
.
A GMCSF-neuroantigen fusion protein is a potent tolerogen in experimental autoimmune encephalomyelitis (EAE) that is associated with efficient targeting of neuroantigen to APC.
J. Leukoc. Biol.
87
:
509
521
.
15
O’Shea
J. J.
,
Gadina
M.
,
Schreiber
R. D.
.
2002
.
Cytokine signaling in 2002: new surprises in the Jak/Stat pathway.
Cell
109
(
Suppl.
):
S121
S131
.
16
Wen
R.
,
Wang
D.
,
McKay
C.
,
Bunting
K. D.
,
Marine
J. C.
,
Vanin
E. F.
,
Zambetti
G. P.
,
Korsmeyer
S. J.
,
Ihle
J. N.
,
Cleveland
J. L.
.
2001
.
Jak3 selectively regulates Bax and Bcl-2 expression to promote T-cell development.
Mol. Cell. Biol.
21
:
678
689
.
17
Kubo
M.
,
Hanada
T.
,
Yoshimura
A.
.
2003
.
Suppressors of cytokine signaling and immunity.
Nat. Immunol.
4
:
1169
1176
.
18
Yoshimura
A.
,
Naka
T.
,
Kubo
M.
.
2007
.
SOCS proteins, cytokine signalling and immune regulation.
Nat. Rev. Immunol.
7
:
454
465
.
19
Hara
T.
,
Miyajima
A.
.
1996
.
Function and signal transduction mediated by the interleukin 3 receptor system in hematopoiesis.
Stem Cells
14
:
605
618
.
20
Balzac
F.
,
Belkin
A. M.
,
Koteliansky
V. E.
,
Balabanov
Y. V.
,
Altruda
F.
,
Silengo
L.
,
Tarone
G.
.
1993
.
Expression and functional analysis of a cytoplasmic domain variant of the beta 1 integrin subunit.
J. Cell Biol.
121
:
171
178
.
21
Kasper
S.
,
Kindler
T.
,
Sonnenschein
S.
,
Breitenbuecher
F.
,
Böhmer
F. D.
,
Huber
C.
,
Fischer
T.
.
2007
.
Cross-inhibition of interferon-induced signals by GM-CSF through a block in Stat1 activation.
J. Interferon Cytokine Res.
27
:
947
959
.
22
Kotone-Miyahara
Y.
,
Yamashita
K.
,
Lee
K. K.
,
Yonehara
S.
,
Uchiyama
T.
,
Sasada
M.
,
Takahashi
A.
.
2004
.
Short-term delay of Fas-stimulated apoptosis by GM-CSF as a result of temporary suppression of FADD recruitment in neutrophils: evidence implicating phosphatidylinositol 3-kinase and MEK1-ERK1/2 pathways downstream of classical protein kinase C.
J. Leukoc. Biol.
76
:
1047
1056
.
23
Hamilton
J. A.
2008
.
Colony-stimulating factors in inflammation and autoimmunity.
Nat. Rev. Immunol.
8
:
533
544
.
24
Selleri
S.
,
Deola
S.
,
Pos
Z.
,
Jin
P.
,
Worschech
A.
,
Slezak
S. L.
,
Rumio
C.
,
Panelli
M. C.
,
Maric
D.
,
Stroncek
D. F.
, et al
.
2008
.
GM-CSF/IL-3/IL-5 receptor common beta chain (CD131) expression as a biomarker of antigen-stimulated CD8+ T cells.
J. Transl. Med.
6
:
17
.
25
Jackson
S. H.
,
Yu
C. R.
,
Mahdi
R. M.
,
Ebong
S.
,
Egwuagu
C. E.
.
2004
.
Dendritic cell maturation requires STAT1 and is under feedback regulation by suppressors of cytokine signaling.
J. Immunol.
172
:
2307
2315
.
26
Donahue
R. E.
,
Seehra
J.
,
Metzger
M.
,
Lefebvre
D.
,
Rock
B.
,
Carbone
S.
,
Nathan
D. G.
,
Garnick
M.
,
Sehgal
P. K.
,
Laston
D.
, et al
.
1988
.
Human IL-3 and GM-CSF act synergistically in stimulating hematopoiesis in primates.
Science
241
:
1820
1823
.
27
Vadhan-Raj
S.
,
Broxmeyer
H. E.
,
Andreeff
M.
,
Bandres
J. C.
,
Buescher
E. S.
,
Benjamin
R. S.
,
Papadopoulos
N. E.
,
Burgess
A.
,
Patel
S.
,
Plager
C.
, et al
.
1995
.
In vivo biologic effects of PIXY321, a synthetic hybrid protein of recombinant human granulocyte-macrophage colony-stimulating factor and interleukin-3 in cancer patients with normal hematopoiesis: a phase I study.
Blood
86
:
2098
2105
.
28
Curtis
B. M.
,
Williams
D. E.
,
Broxmeyer
H. E.
,
Dunn
J.
,
Farrah
T.
,
Jeffery
E.
,
Clevenger
W.
,
deRoos
P.
,
Martin
U.
,
Friend
D.
, et al
.
1991
.
Enhanced hematopoietic activity of a human granulocyte/macrophage colony-stimulating factor-interleukin 3 fusion protein.
Proc. Natl. Acad. Sci. USA
88
:
5809
5813
.
29
Izquierdo
M. A.
,
Degen
D.
,
Raymond
E.
,
Caron
D.
,
Ortiz
V.
,
Banks
P.
,
Von Hoff
D. D.
.
1996
.
Effects of PIXY321, a granulocyte-macrophage colony-stimulating factor/interleukin 3 fusion protein, on human tumor colony-forming units taken directly from patients.
Clin. Cancer Res.
2
:
1713
1716
.
30
Vadhan-Raj
S.
,
Papadopoulos
N. E.
,
Burgess
M. A.
,
Linke
K. A.
,
Patel
S. R.
,
Hays
C.
,
Arcenas
A.
,
Plager
C.
,
Kudelka
A. P.
,
Hittelman
W. N.
, et al
.
1994
.
Effects of PIXY321, a granulocyte-macrophage colony-stimulating factor/interleukin-3 fusion protein, on chemotherapy-induced multilineage myelosuppression in patients with sarcoma.
J. Clin. Oncol.
12
:
715
724
.
31
O’Shaughnessy
J. A.
,
Tolcher
A.
,
Riseberg
D.
,
Venzon
D.
,
Zujewski
J.
,
Noone
M.
,
Gossard
M.
,
Danforth
D.
,
Jacobson
J.
,
Chang
V.
, et al
.
1996
.
Prospective, randomized trial of 5-fluorouracil, leucovorin, doxorubicin, and cyclophosphamide chemotherapy in combination with the interleukin-3/granulocyte-macrophage colony-stimulating factor (GM-CSF) fusion protein (PIXY321) versus GM-CSF in patients with advanced breast cancer.
Blood
87
:
2205
2211
.
32
Asao
H.
,
Okuyama
C.
,
Kumaki
S.
,
Ishii
N.
,
Tsuchiya
S.
,
Foster
D.
,
Sugamura
K.
.
2001
.
Cutting edge: the common gamma-chain is an indispensable subunit of the IL-21 receptor complex.
J. Immunol.
167
:
1
5
.
33
Fehniger
T. A.
,
Caligiuri
M. A.
.
2001
.
Interleukin 15: biology and relevance to human disease.
Blood
97
:
14
32
.
34
Kovanen
P. E.
,
Rosenwald
A.
,
Fu
J.
,
Hurt
E. M.
,
Lam
L. T.
,
Giltnane
J. M.
,
Wright
G.
,
Staudt
L. M.
,
Leonard
W. J.
.
2003
.
Analysis of gamma c-family cytokine target genes. Identification of dual-specificity phosphatase 5 (DUSP5) as a regulator of mitogen-activated protein kinase activity in interleukin-2 signaling.
J. Biol. Chem.
278
:
5205
5213
.
35
Carson
W. E.
,
Fehniger
T. A.
,
Haldar
S.
,
Eckhert
K.
,
Lindemann
M. J.
,
Lai
C. F.
,
Croce
C. M.
,
Baumann
H.
,
Caligiuri
M. A.
.
1997
.
A potential role for interleukin-15 in the regulation of human natural killer cell survival.
J. Clin. Invest.
99
:
937
943
.
36
de Totero
D.
,
Meazza
R.
,
Capaia
M.
,
Fabbi
M.
,
Azzarone
B.
,
Balleari
E.
,
Gobbi
M.
,
Cutrona
G.
,
Ferrarini
M.
,
Ferrini
S.
.
2008
.
The opposite effects of IL-15 and IL-21 on CLL B cells correlate with differential activation of the JAK/STAT and ERK1/2 pathways.
Blood
111
:
517
524
.
37
Kumano
M.
,
Hara
I.
,
Furukawa
J.
,
Oniki
S.
,
Nagai
H.
,
Miyake
H.
,
Fujisawa
M.
.
2007
.
Interleukin-21 activates cytotoxic T lymphocytes and natural killer cells to generate antitumor response in mouse renal cell carcinoma.
J. Urol.
178
:
1504
1509
.
38
Roda
J. M.
,
Parihar
R.
,
Lehman
A.
,
Mani
A.
,
Tridandapani
S.
,
Carson
W. E.
 III
.
2006
.
Interleukin-21 enhances NK cell activation in response to antibody-coated targets.
J. Immunol.
177
:
120
129
.
39
Li
Y.
,
Yee
C.
.
2008
.
IL-21 mediated Foxp3 suppression leads to enhanced generation of antigen-specific CD8+ cytotoxic T lymphocytes.
Blood
111
:
229
235
.
40
Zorn
E.
,
Nelson
E. A.
,
Mohseni
M.
,
Porcheray
F.
,
Kim
H.
,
Litsa
D.
,
Bellucci
R.
,
Raderschall
E.
,
Canning
C.
,
Soiffer
R. J.
, et al
.
2006
.
IL-2 regulates FOXP3 expression in human CD4+CD25+ regulatory T cells through a STAT-dependent mechanism and induces the expansion of these cells in vivo.
Blood
108
:
1571
1579
.
41
Stagg
J.
,
Wu
J. H.
,
Bouganim
N.
,
Galipeau
J.
.
2004
.
Granulocyte-macrophage colony-stimulating factor and interleukin-2 fusion cDNA for cancer gene immunotherapy.
Cancer Res.
64
:
8795
8799
.
42
Penafuerte
C.
,
Bautista-Lopez
N.
,
Boulassel
M. R.
,
Routy
J. P.
,
Galipeau
J.
.
2009
.
The human ortholog of granulocyte macrophage colony-stimulating factor and interleukin-2 fusion protein induces potent ex vivo natural killer cell activation and maturation. [Published erratum appears in 2010 Cancer Res. 70: 6106.]
Cancer Res.
69
:
9020
9028
.
43
Faisal
M.
,
Cumberland
W.
,
Champlin
R.
,
Fahey
J. L.
.
1990
.
Effect of recombinant human granulocyte-macrophage colony-stimulating factor administration on the lymphocyte subsets of patients with refractory aplastic anemia.
Blood
76
:
1580
1585
.
44
Rafei
M.
,
Wu
J. H.
,
Annabi
B.
,
Lejeune
L.
,
François
M.
,
Galipeau
J.
.
2007
.
A GMCSF and IL-15 fusokine leads to paradoxical immunosuppression in vivo via asymmetrical JAK/STAT signaling through the IL-15 receptor complex.
Blood
109
:
2234
2242
.
45
Rafei
M.
,
Hsieh
J.
,
Zehntner
S.
,
Li
M.
,
Forner
K.
,
Birman
E.
,
Boivin
M. N.
,
Young
Y. K.
,
Perreault
C.
,
Galipeau
J.
.
2009
.
A granulocyte-macrophage colony-stimulating factor and interleukin-15 fusokine induces a regulatory B cell population with immune suppressive properties.
Nat. Med.
15
:
1038
1045
.
46
Williams
P.
,
Rafei
M.
,
Bouchentouf
M.
,
Raven
J.
,
Yuan
S.
,
Cuerquis
J.
,
Forner
K. A.
,
Birman
E.
,
Galipeau
J.
.
2010
.
A fusion of GMCSF and IL-21 initiates hypersignaling through the IL-21Ralpha chain with immune activating and tumoricidal effects in vivo.
Mol. Ther.
18
:
1293
1301
.
47
Williams
P.
,
Bouchentouf
M.
,
Rafei
M.
,
Romieu-Mourez
R.
,
Hsieh
J.
,
Boivin
M. N.
,
Yuan
S.
,
Forner
K. A.
,
Birman
E.
,
Galipeau
J.
.
2010
.
A dendritic cell population generated by a fusion of GM-CSF and IL-21 induces tumor-antigen-specific immunity.
J. Immunol.
185
:
7358
7366
.
48
Yasukawa
H.
,
Ohishi
M.
,
Mori
H.
,
Murakami
M.
,
Chinen
T.
,
Aki
D.
,
Hanada
T.
,
Takeda
K.
,
Akira
S.
,
Hoshijima
M.
, et al
.
2003
.
IL-6 induces an anti-inflammatory response in the absence of SOCS3 in macrophages.
Nat. Immunol.
4
:
551
556
.
49
Proost
P.
,
De Meester
I.
,
Schols
D.
,
Struyf
S.
,
Lambeir
A. M.
,
Wuyts
A.
,
Opdenakker
G.
,
De Clercq
E.
,
Scharpé
S.
,
Van Damme
J.
.
1998
.
Amino-terminal truncation of chemokines by CD26/dipeptidyl-peptidase IV. Conversion of RANTES into a potent inhibitor of monocyte chemotaxis and HIV-1-infection.
J. Biol. Chem.
273
:
7222
7227
.
50
Rollins
B. J.
1997
.
Chemokines.
Blood
90
:
909
928
.
51
Yuen
J. W.
,
Poon
L. S.
,
Chan
A. S.
,
Yu
F. W.
,
Lo
R. K.
,
Wong
Y. H.
.
2010
.
Activation of STAT3 by specific Galpha subunits and multiple Gbetagamma dimers.
Int. J. Biochem. Cell Biol.
42
:
1052
1059
.
52
De Vries
L.
,
Zheng
B.
,
Fischer
T.
,
Elenko
E.
,
Farquhar
M. G.
.
2000
.
The regulator of G protein signaling family.
Annu. Rev. Pharmacol. Toxicol.
40
:
235
271
.
53
Tohgo
A.
,
Choy
E. W.
,
Gesty-Palmer
D.
,
Pierce
K. L.
,
Laporte
S.
,
Oakley
R. H.
,
Caron
M. G.
,
Lefkowitz
R. J.
,
Luttrell
L. M.
.
2003
.
The stability of the G protein-coupled receptor-beta-arrestin interaction determines the mechanism and functional consequence of ERK activation.
J. Biol. Chem.
278
:
6258
6267
.
54
Ohnishi
H.
,
Chao
J. T.
,
Lin
K. K.
,
Lee
H.
,
Chu
T. M.
.
1989
.
Role of the kidney in metabolic change of interleukin-2.
Tumour Biol.
10
:
202
214
.
55
Rafei
M.
,
Hsieh
J.
,
Fortier
S.
,
Li
M.
,
Yuan
S.
,
Birman
E.
,
Forner
K.
,
Boivin
M. N.
,
Doody
K.
,
Tremblay
M.
, et al
.
2008
.
Mesenchymal stromal cell-derived CCL2 suppresses plasma cell immunoglobulin production via STAT3 inactivation and PAX5 induction.
Blood
112
:
4991
4998
.
56
McQuibban
G. A.
,
Gong
J. H.
,
Wong
J. P.
,
Wallace
J. L.
,
Clark-Lewis
I.
,
Overall
C. M.
.
2002
.
Matrix metalloproteinase processing of monocyte chemoattractant proteins generates CC chemokine receptor antagonists with anti-inflammatory properties in vivo.
Blood
100
:
1160
1167
.
57
Huang
D. R.
,
Wang
J.
,
Kivisakk
P.
,
Rollins
B. J.
,
Ransohoff
R. M.
.
2001
.
Absence of monocyte chemoattractant protein 1 in mice leads to decreased local macrophage recruitment and antigen-specific T helper cell type 1 immune response in experimental autoimmune encephalomyelitis.
J. Exp. Med.
193
:
713
726
.
58
Jee
Y.
,
Yoon
W. K.
,
Okura
Y.
,
Tanuma
N.
,
Matsumoto
Y.
.
2002
.
Upregulation of monocyte chemotactic protein-1 and CC chemokine receptor 2 in the central nervous system is closely associated with relapse of autoimmune encephalomyelitis in Lewis rats.
J. Neuroimmunol.
128
:
49
57
.
59
Izikson
L.
,
Klein
R. S.
,
Charo
I. F.
,
Weiner
H. L.
,
Luster
A. D.
.
2000
.
Resistance to experimental autoimmune encephalomyelitis in mice lacking the CC chemokine receptor (CCR)2.
J. Exp. Med.
192
:
1075
1080
.
60
Zhang
J.
,
Patel
L.
,
Pienta
K. J.
.
2010
.
CC chemokine ligand 2 (CCL2) promotes prostate cancer tumorigenesis and metastasis.
Cytokine Growth Factor Rev.
21
:
41
48
.
61
Charo
I. F.
1999
.
CCR2: from cloning to the creation of knockout mice.
Chem. Immunol.
72
:
30
41
.
62
Rafei
M.
,
Berchiche
Y. A.
,
Birman
E.
,
Boivin
M. N.
,
Young
Y. K.
,
Wu
J. H.
,
Heveker
N.
,
Galipeau
J.
.
2009
.
An engineered GM-CSF-CCL2 fusokine is a potent inhibitor of CCR2-driven inflammation as demonstrated in a murine model of inflammatory arthritis.
J. Immunol.
183
:
1759
1766
.
63
Rafei
M.
,
Campeau
P. M.
,
Wu
J. H.
,
Birman
E.
,
Forner
K.
,
Boivin
M. N.
,
Galipeau
J.
.
2009
.
Selective inhibition of CCR2 expressing lymphomyeloid cells in experimental autoimmune encephalomyelitis by a GM-CSF-MCP1 fusokine.
J. Immunol.
182
:
2620
2627
.
64
DeNardo
D. G.
,
Barreto
J. B.
,
Andreu
P.
,
Vasquez
L.
,
Tawfik
D.
,
Kolhatkar
N.
,
Coussens
L. M.
.
2009
.
CD4(+) T cells regulate pulmonary metastasis of mammary carcinomas by enhancing protumor properties of macrophages.
Cancer Cell
16
:
91
102
.
65
Blanpain
C.
,
Migeotte
I.
,
Lee
B.
,
Vakili
J.
,
Doranz
B. J.
,
Govaerts
C.
,
Vassart
G.
,
Doms
R. W.
,
Parmentier
M.
.
1999
.
CCR5 binds multiple CC-chemokines: MCP-3 acts as a natural antagonist.
Blood
94
:
1899
1905
.
66
Edman
L. C.
,
Mira
H.
,
Arenas
E.
.
2008
.
The beta-chemokines CCL2 and CCL7 are two novel differentiation factors for midbrain dopaminergic precursors and neurons.
Exp. Cell Res.
314
:
2123
2130
.
67
Szymczak
W. A.
,
Deepe
G. S.
 Jr.
2009
.
The CCL7-CCL2-CCR2 axis regulates IL-4 production in lungs and fungal immunity.
J. Immunol.
183
:
1964
1974
.

The authors are inventors on pending United States provisional patents.