doi:10.1111/j.1432-1033.2004.04424.x

Eur. J. Biochem. 271, 4613–4620 (2004) (cid:1) FEBS 2004

M I N I R E V I E W

Regulation of STAT signalling by proteolytic processing

Lisa Hendry and Susan John

Peter Gorer Department of Immunobiology, Programme in Infection and Immunity, King’s College London, UK

Interaction of cytokines with their cognate receptors leads to the activation of latent transcription factors, the signal transducer and activator of transcription (STAT) proteins. Numerous studies have identified the critical roles played by STAT proteins in regulating cell proliferation, differ- entiation and survival. Consequently, the activity of STAT proteins is negatively regulated by a variety of different mechanisms, which include alternative splicing, covalent modifications, protein–protein interactions with negative regulatory proteins and proteolytic processing by pro- teases. Cleavage of STAT proteins by proteases results in the generation of C-terminally truncated proteins, called STATc, which lack the transactivation domain and behave functional dominant-negative proteins. Currently, as

STATc isoforms have been identified for Stat3, Stat5a, Stat5b and Stat6 in different cellular contexts and biolo- gical processes. Evidence is mounting for the role of as yet unidentified serine proteases in the proteolytic processing of STAT proteins, although at least one cysteine protease, calpain is also known to cleave these STATs in platelets and mast cells. Recently, studies of acute myeloid leukae- mia and cutaneous T cell lymphoma patients have revealed important roles for the aberrant expression of Stat3c and Stat5c proteins in the pathology of these diseases. To- gether, these findings indicate that proteolytic processing is an important mechanism in the regulation of STAT pro- tein biological activity and provides a fertile area for future studies.

Introduction

The Janus kinase-signal transducer and activator of tran- scription (JAK-STAT) signalling pathway, first identified for the interferon-a/b and c receptors, is now known to be employed by many cytokine and growth factor receptors and to be evolutionarily conserved [1,2]. STAT proteins have a common overall structure and are organized into distinct functional modular domains (Fig. 1).

In addition to these direct protein–protein interaction methods of negative regulation, STATs are also regulated at the level of alternative splicing. The STATb forms, gener- ated by alternative splicing, possess an altered carboxy- terminal (C-terminal) lacking the natural transactivation domain and behave as functional dominant-negative pro- teins when overexpressed in cells [4–6]. However, recent evidence from transgenic mice indicates that STATb proteins are not strict dominant-negatives, and actually contribute to transcriptional activation of selective target genes, despite the absence of the natural transactivation domain [7–9]. The mechanism by which STATb isoforms achieve transactivation remains to be elucidated, but probably involves the differential interaction with other transcription factors.

After a decade of intense investigation into the structure and biological functions of STAT proteins, their essential roles in cell proliferation, differentiation and survival have been firmly established [2]. A number of studies have iden- tified important negative regulatory mechanisms that exist to curtail the activity of STAT proteins (Fig. 2). These include the activities of phosphatases, suppressors of cytokine signalling (SOCS), interaction of inhibitory proteins such as protein inhibitor of activated STATs (PIAS), and targeted proteasome-dependent degradation of active STATs [2,3].

Another mechanism by which STAT signalling is regu- lated occurs at the level of limited proteolytic processing in cellular contexts where there is no evidence for alternative splicing [10]. Proteolytic processing of STAT proteins also results in the generation of C-terminally truncated STAT proteins, referred to as STATc, but these proteins lack the transactivation domain, without the addition of any extra amino acid sequences at their C-termini. Thus, multiple functional forms of STAT proteins, generated by distinct mechanisms exist in different cell lineages. Here we review the generation and function of STATc proteins and their role in human diseases.

Processing of Stat5 in haematopoietic progenitor cells

Stat5 is activated by a wide variety of haematological and nonhaematological cytokines and growth factors including those which regulate the proliferation and differentiation of

Correspondence to S. John, Peter Gorer Department of Immunobiol- ogy, Programme in Infection and Immunity, King’s College London, 2nd floor New Guy’s House, St. Thomas Street, London SE1 9RT, UK. E-mail: susan.john@kcl.ac.uk Abbreviations: AML, acute myeloid leukaemia; BMMC, bone mar- row-derived mast cells; CTCL, cutaneous T cell lymphoma; G-CSF, granulocyte colony stimulating factor; GM-CSF, granulocyte- macrophage colony-stimulating factor; IL, interleukin; JAK, Janus kinase; PBMC, peripheral blood mononuclear cell; PIAS, protein inhibitor of activated STATs; PMSF, phenylmethanesulfonyl fluoride; SOCS, suppressors of cytokine signalling; SS, Sezary syndrome; STAT, signal transducer and activator of transcription. (Received 17 August 2004, accepted 7 October 2004)

4614 L. Hendry and S. John (Eur. J. Biochem. 271)

(cid:1) FEBS 2004

Fig. 1. Modular structure of STAT proteins. All STAT proteins share a common molecular topology and are organized into distinct functional domains. The NH2-terminus (N-domain) is involved in protein–protein interactions between adjacent STAT dimers on DNA, facilitating the formation of STAT tetramers. It is also involved in the formation of dimers between nonphosphorylated STAT monomers, which is important for receptor-mediated activation and nuclear translocation of certain STAT proteins. Interactions with STAT cofactors, which positively or negatively modulate their transcriptional activity, occur via the N-domain, the adjacent coiled-coil domain and the carboxy-terminal transactivation domain (TAD). The conserved serine residue (p-S), which is phosphorylated upon cytokine stimulation and is important for maximal transcriptional activation, is located within the transactivation domain. The conserved tyrosine residue (p-Y), that becomes phosphorylated upon activation is located immediately preceeding the transactivation domain.

effects on early haematopoietic progenitor cells [13]. The two Stat5 proteins, Stat5a and Stat5b, are encoded by separate genes and are expressed as both full-length (Stat5a) and shorter, C-terminally truncated proteins [5,14]. Although alternative splicing generates Stat5b in certain cellular contexts, the lack of abundance of the alternatively

myeloid [interleukin (IL)-3, IL-5, granulocyte-macrophage colony-stimulating factor (GM-CSF) and thrombopoietin], erythroid (erythropoietin) and lymphoid lineages (the gamma-c family of cytokines, IL-2, IL-7 and IL-15) [11,12]. Targeted deletions in mice of genes encoding Stat5 results in defects in myeloid cell differentiation through

Fig. 2. Negative regulation of STAT signalling. Cytokine-induced STAT activation can be inhibited by suppressors of cytokine signalling (SOCS) proteins, whose gene expression is regulated by STAT proteins, thus fulfilling a negative feedback loop. SOCS proteins inhibit STAT activation either by inhibition of the activating JAKs or by competition with STATs for receptor binding. Activated STAT proteins can be dephosphorylated by cytoplasmic and/or nuclear phosphatases. C-terminally truncated STAT proteins, STATb and STATc, behave as dominant-negative proteins to functionally compete with their full-length counterparts to alter or inhibit gene expression, respectively. Protein inhibitor of activated STATs (PIAS) proteins interact with STAT proteins to inhibit their DNA binding and/or potentially facilitate their covalent modification by sumoylation and subsequent degradation. Ub, ubiquitin; SUMO, small ubiquitin-like modifier.

STAT signalling by proteolytic processing (Eur. J. Biochem. 271) 4615

(cid:1) FEBS 2004

spliced message in haematopoetic progenitor cells led investigators to evaluate other mechanisms for the genera- tion of C-terminally truncated Stat5 proteins.

nucleus of naı¨ ve PBMC/T cells [21]. Analysis of the truncated Stat5 proteins using N- and C-terminal Stat5 antibodies revealed that the truncation is at the C-terminus of the Stat5 protein, as previously noted in myeloid cells. The expression of the truncated protein in the nucleus is independent of the phosphorylation state of Stat5a and Stat5b. Unlike myeloid progenitor cells, the cytoplasmic fraction expresses both the full-length and the truncated Stat5 protein, although at present we cannot exclude the possibility that the truncated protein is exclusively generated in the nucleus but is present in the cytoplasmic fraction due to protein shuttling, which has been shown to occur in a cytokine-dependent and independent manner for STAT proteins [22,23].

It was noted that distinct forms of Stat5 proteins were activated upon IL-3 treatment of specific myeloid cell lineages. Thus, in myeloid progenitor cell lineages stimula- tion with IL-3, GM-CSF or erythropoietin activates a shorter, C-terminally truncated isoform of Stat5a (77 kDa) and Stat5b (80 kDa), while full-length Stat5a (96 kDa) and Stat5b (94 kDa) are only activated in differentiated mature myeloid cells [10,15]. The Stat5c proteins in myeloid progenitor cells are generated by a putative Stat5 protease, which is primarily located in the nucleus and cleaves Stat5 proteins independently of their tyrosine-phosphorylation states [10,16]. The protease is an endopeptidase and is inhibited by the broad-spectrum serine protease inhibitor, phenylmethanesulfonyl fluoride (PMSF). Cellular fraction- ation and chromatography studies indicate that the protease has an approximate molecular mass of 25 kDa and cleaves murine Stat5a between amino acids 719 (tyrosine; Y) and 720 (methionine; M) and Stat5b between Y724 and M725 [17]. Mutant Stat5 proteins bearing amino acid substitutions at these positions were resistant to cleavage by the protease. Importantly, the Stat5-proteolytic activity was absent in mature myeloid cells suggesting that either the expression of protease is down-regulated or alternatively inactivated upon myeloid cell differentiation [10,16,17].

Upon activation of naı¨ ve T cells by mitogenic stimula- tion, the expression of truncated Stat5a and Stat5b proteins disappears and is replaced by the expression and activation of the full-length Stat5 proteins [21]. Significantly, the normal regulation of truncated vs. full-length Stat5 is lymphoma (CTCL) dysregulated in cutaneous T cell patients and will be described in a later section. Ongoing studies indicate that the truncated Stat5 protein is generated by the activity of a protease, which is down-regulated or inactivated upon mitogenic stimulation (Fig. 3). Future biochemical characterization and purification of the prote- ase(s) and the identification of the exact cleavage site on Stat5 will be important in enhancing our understanding of the regulation of Stat5 function by proteolytic cleavage in peripheral T cells.

Proteolytic regulation of Stat5 and Stat3 in mature human neutrophils

Stat3 and Stat5 isoforms have been identified in differen- tiated human peripheral blood monocytes and polymor-

Consistent with the distinct function of truncated Stat5 proteins in immature myeloid progenitors, they fail to activate several known IL-3-induced target genes that are activated by the full-length proteins in differentiated mature myeloid cells [10]. The functional significance of truncated Stat5 proteins in maintaining an undifferentiated immature phenotype of myeloid cells was convincingly demonstrated by studies using stable enforced expression of mutant, noncleavable forms of Stat5 in undifferentiated myeloid cells [18]. The mutant cell lines developed a partially differentiated phenotype and were resistant to further differentiation by cytokine treatment. Thus, proteolytic cleavage of Stat5 is an important physiological mechanism in regulating myeloid cell differentiation.

Proteolytic cleavage of Stat5 in peripheral T cells

Despite the clear role of proteases in regulating myeloid cell development, Schindler and colleagues were unable to demonstrate an analogous situation for lymphoid cell development in murine thymic T cells [17]. However, studies of human peripheral blood mononuclear cells (PBMCs) indicate that naı¨ ve T cells in the peripheral immune system possess a similar mechanism for regulating Stat5 as myeloid progenitor cells. Activation of naı¨ ve T cells by antigenic or mitogenic stimulation leads to cell proliferation and differ- entiation into effector T cells, mediated by the action of immunologically important cytokines, which signal via Type I and Type II cytokine receptors. Stat5 activation, mediated by IL-2 signalling upon T cell activation, is an important regulator of cell proliferation and survival [19,20]. Recently, studies on Stat5 expression and activation in normal human PBMC and peripheral T cells revealed that Stat5 is expressed exclusively as a truncated protein in the

Fig. 3. Stat5a protein is cleaved to Stat5c by the activity of a protease present in peripheral blood mononuclear cell (PBMC) extract. The presence of Stat5-proteolytic activity was evaluated by coincubation assay. Extracts prepared from either PBMC (lane 2) or PBMC mito- genically stimulated with phytohaemagglutinin (PHA-Blasts, lane 3), or a buffer control containing no cell extract (lane 1), were incubated with FLAG-tagged Stat5a protein at 37 (cid:2)C for 15 min. Samples were then analyzed by Western blot analysis using an anti-FLAG IgG. Cleavage of the FLAG-Stat5a input protein was obtained specifically with fresh PBMC extrcats and not with extracts made from PHA- Blasts.

4616 L. Hendry and S. John (Eur. J. Biochem. 271)

(cid:1) FEBS 2004

the main STAT that

phonuclear neutrophils [24–27]. During terminal differen- induced by granulocyte colony tiation of neutrophils, stimulating factor (G-CSF), is activated is Stat3 and it is predominantly expressed as Stat3c, generated by proteolytic cleavage of Stat3a [25,28]. Unlike the progenitor myeloid Stat5 protease, the Stat3 protease, activated by G-CSF can only cleave the active, phosphorylated form of Stat3a [25]. The exact specificity of the Stat3 protease appears to be less clear, as the proteolytic activity was shown to be inhibited by di- isopropylfluorophosphate and not PMSF in living cells, but neither was effective at inhibiting the protease in vitro [25]. The relationship between the Stat3 protease from mature neutrophils and the Stat5 protease from immature myeloid cells is also unknown at present, but the activation of these proteases in different developmental contexts may suggest that they are distinct proteases.

The activity of the murine Stat6 protease is exclusively nuclear and can be inhibited by the serine protease inhibitors, PMSF and 4-(2-aminoethyl)-benzenesulfonyl- fluoride [35]. Moreover, the activity of the Stat6 protease is not dependent on the expression of Stat6, as Stat6-deficient BMMC also contained Stat6-specific proteolytic activity [35,36]. More recently Iwamoto and colleagues have further characterized the serine protease to be inhibited by an elastase inhibitor ONO-5046, suggesting that this protease may belong to an elastase family [36,37]. The Stat6 protease cleaves Stat6 between amino acids 685 (aspartic acid; D) and 686 (methionine; M). The amino acid sequences surrounding the cleavage site are not conserved in the human Stat6 protein, providing an explanation for the lack of observation of truncated Stat6 in human mast cells. While cleavage-resistant point mutants of Stat6 (Stat6 D685A and M686A) have similar transcriptional activity as their wild-type counterpart in cell transfection assays, the stable expression of these Stat6 mutants in cell lines results in prolonged nuclear accumulation of Stat6 and enhanced IL-4-induced apoptosis and growth inhibition of the mutant mast cell lines [35]. Furthermore, enforced coexpression of truncated Stat6 with Stat6 D685A reverses the functional effect of the latter mutant indicating that the truncated Stat6 protein can potentially function as a dominant-negative in BMMC [35].

Despite the finding that both the Stat5 protease and the Stat6 protease are serine proteases, the similarity apparently does not extend any further. The Stat5 protease from myeloid cells does not cleave Stat6 and is not inhibited by ONO-5046, and the Stat6 protease from BMMC does not cleave Stat5 [35–37]. Thus, the serine proteases that regulate STAT activity show STAT and cell-type specificity.

More recently, investigators have shown that Stat5 is also similarly regulated by proteolytic processing in mature human neutrophils [26]. Stat5 is activated in human neutrophils by the cytokines IL-2 and GM-CSF, which are both potent modulators of neutrophil activity [29]. In a now familiar theme, these cytokines activate nuclear expression of a C-terminally truncated form of Stat5 in neutrophils, which results in a failure to induce expression of known Stat5-regulated genes, such as osm and pim-1, consistent with the inability of these cytokines to induce proliferation of these cells [26]. No evidence was found for alternative splicing of Stat5 in these cells and instead truncated Stat5 proteins were generated by the activity of a nuclear, PMSF-sensitive serine protease. The exact rela- tionship between the various Stat5-serine proteases derived from myeloid progenitors, human PBMC and mature neutrophils awaits identification by future molecular clo- ning studies.

Processing of Stat3, Stat5 and Stat6 by calpain

Regulation of Stat6 activity by proteolytic cleavage in mast cells

While the most common mechanism of proteolytic process- ing of STAT proteins is mediated by the action of serine proteases, at least one other cellular protease is known to specifically cleave certain STAT proteins. The calcium- dependent cysteine protease calpain was demonstrated to cleave Stat3 and Stat5 in platelets and Stat6 in mast cells to generate C-terminally truncated proteins [37,38]. Activation of intracellular calpain by thrombin treatment of platelets resulted in a significant increase in the levels of C-terminally truncated Stat3 and Stat5 [38]. Similarly, Stat6 was cleaved upon activation of calpain by dibucaine treatment of BMMC [37]. However, the truncated Stat6 protein that is generated as a result of cleavage by calpain is a 70 kDa protein as compared to the 65 kDa protein generated by the Stat6 protease. Furthermore, the generation of the 70 kDa but not the 65 kDa Stat6 protein is inhibited by the calpain inhibitor calpeptin [37]. Thus, multiple different STATc isoforms can be generated by the activation of different cellular proteases in BMMCs. It is unclear whether the calpain cleaved Stat5 in platelets is identical in size and function to the Stat5c proteins generated by proteolytic processing by the Stat5 proteases from myeloid progenitor or mature neutrophil cells. The physiological importance of STATc isoforms generated by calpain is unknown at present but, as calpain is potently activated by increased following cellular intracellular calcium concentrations

Unlike Stat3 and Stat5, which are activated by a wide variety of cytokines and growth factors, Stat6 is very selectively activated by IL-4 and the related cytokine, IL-13 [30]. Stat6 deficient mice reveal defects in such crucial aspects of normal immune function as Th2 cell differenti- ation, B cell isotype switching and the loss of contact hypersensitivity [30]. While IL-4 induced Stat6 signalling is an activating signal in murine B and T cells, its role in bone marrow-derived mast cells (BMMC) is less clear [31,32]. Analysis of Stat6 expression in murine BMMC provided a clue to these apparent cellular differences in response to IL-4. Brown and colleagues first observed that, whereas Stat6 is expressed as a 100 kDa full-length protein in B and T cells, it is expressed as a 65 kDa protein in murine BMMC [33]. A similarly truncated Stat6 protein has not been identified in human mast cells and it is possible that this mechanism of regulation of Stat6 has been lost during evolution. Studies revealed that Stat6 is truncated at its C- terminus and is lacking the transactivation domain in murine BMMCs [33]. While no evidence for alternative splicing of Stat6 was obtained in mast cells, several groups have established that truncated Stat6 protein is generated by proteolytic processing in these cells [33–35].

STAT signalling by proteolytic processing (Eur. J. Biochem. 271) 4617

(cid:1) FEBS 2004

activation, it is plausible that calpain mediated processing of STAT proteins may be an important mechanism for regulating STAT-dependent gene expression [39].

Dysregulated expression of proteolytically processed STAT proteins in human diseases

The constitutive activation of full-length Stat3 and Stat5 is a common feature of many primary human tumours of haematopoietic and nonhaematopoeitic origins and is extensively reviewed elsewhere [40,41]. Recent studies of acute myeloid leukaemia (AML) and CTCL patients indicate that C-terminally truncated STAT proteins also contribute to the pathology of these diseases.

molecular mass of 40 kDa. Thus, the active protease in AML blasts may either represent yet another member of the STAT-serine protease family or alternatively may be aberrantly post-translationally modified. Given the clearly established dominant-negative functions of C-terminally truncated STAT proteins, the aberrant constitutive expres- sion of truncated Stat3 and Stat5 proteins in AML blasts has important physiological implications for the pathology of the disease. As cleavage-resistant mutant Stat5 proteins induce differentiation and apoptosis of myeloid cells when artificially expressed, it is plausible to speculate that the selective expression of truncated Stat5 and Stat3 proteins may enhance survival of leukaemic blasts cells in AML, while at the same time preventing cellular differentiation [18].

Acute myeloid leukaemia

Cutaneous T cell Lymphoma

AML is characterized by the clonal expansion of myeloid cells that have been arrested in their maturation. Like their normal counterparts, AML blasts can proliferate in response to haematopoietic cytokines such as GM-CSF, G-CSF, thrombopoietin and IL-3, which signal via the JAK-STAT pathway [42]. However, unlike normal myeloid cells, which undergo differentiation in response to specific cytokine treatment, the leukaemic cells proliferate but do not differentiate, suggesting that crucial signalling pathways that regulate cell proliferation and differentiation may be dysregulated in this disease. Analysis of a number of bone marrow samples from pretreatment AML patients revealed that (cid:1) 20–30% of AML blasts expressed constitutively activated full-length Stat3 and Stat5 proteins but a much higher proportion ((cid:1) 80%) expressed C-terminally trun- cated Stat3 and Stat5 proteins [43]. Moreover, 94% of patients in relapse expressed truncated STAT proteins compared to 35% of patients with constitutively active full- length STAT proteins, suggesting that the expression of truncated Stat3 and Stat5 proteins may contribute adversely the shortest to disease progression [44]. Nevertheless, disease-free survival rate and overall survival was seen in patients that had both constitutive activation of full-length Stat3 and concurrent aberrant expression of truncated Stat3 [45]. These studies suggest that the relative ratio of full- length : truncated STAT protein may influence the out- come of disease progression. Constitutive expression of C-terminally truncated Stat5 proteins have also been described previously in CD4 T cells from HIV patients undergoing antiretroviral monotherapy or IL-2 treatment and was associated with good response to therapy [46]. However, it is not known whether the truncated Stat5 protein in patient cells is generated by proteolytic activity or by alternative splicing.

Primary CTCLs are one of the most frequent extranodal lymphomas affecting the skin, and include mycosis fun- goides and its leukaemic variant Sezary syndrome (SS) [48]. Tumour cells are typically CD4 T cells, which display a memory activated phenotype, and express Th2-like cyto- kines (IL-4, IL-5 and IL-10) [49]. While the Jak3-Stat3 pathway is constitutively activated in SS, Stat5 activation is inducible [21,50]. Recently, a study of SS patients with advanced stage disease identified a different form of dysregulation of Stat5 [21]. As mentioned earlier, Stat5 is regulated by proteolytic processing in normal PBMC. Analysis of PBMC from SS patients showed that, unlike in healthy controls, there was elevated or exclusive expres- sion of the C-terminally truncated Stat5 protein even in potently activated cells. DNA binding studies revealed that in SS patients, truncated Stat5 proteins are activated upon IL-2 stimulation and preferentially bind to known Stat5 binding sites, even in patients where a mixture of full-length and truncated Stat5 proteins are expressed. Consistent with these findings, there was a loss of IL-2-induced Stat5- dependent gene expression of target genes such as pim-1, cis, and bcl-2 in patient samples. However, the Stat5-regulated gene CD25 was still inducible by IL-2, consistent with findings from other studies, which indicated that constitu- tively activated Stat3 aberrantly regulates CD25 expression in SS [50]. Thus, it seems likely that the regulation of other important target genes, which are shared by Stat5 and Stat3 may be similarly dysregulated in SS. Future studies investigating the repertoire of target genes activated by full-length vs. truncated Stat5 proteins in T cells will enable us to better understand the functional differences between the different forms of Stat5 and their potential dysregulation in SS. Nevertheless, the preferential DNA binding of truncated Stat5 proteins and the concomitant loss of Stat5-dependent gene expression in SS patients demon- strates that truncated Stat5 proteins can behave as physio- logical dominant-negatives.

Ongoing studies indicate that the dysregulated activity of a Stat5 protease may be responsible for the elevated expression of C-terminally truncated Stat5 proteins in SS (L. Hendry and S. John, unpublished observations). Given the critical role of IL-2 induced Stat5 signalling in normal immune homeostasis and the maintenance of peripheral important tolerance,

this pathway has

the loss of

Biochemical characterization of the AML samples that contained truncated STAT proteins, revealed that a pro- teolytic activity was expressed in these samples, which could selectively cleave Stat3 and Stat5, but not Stat6 [47]. The serine protease inhibitor PMSF was able to inhibit the activity of the Stat3/5 protease from AML blasts, as previously observed for progenitor myeloid cells. However, this serine protease differed from that present in immature myeloid cells in that it was present in both cytoplasmic and nuclear fractions and chromatographic analysis of the protease from AML blasts yielded a protein of approximate

4618 L. Hendry and S. John (Eur. J. Biochem. 271)

(cid:1) FEBS 2004

implications for the pathogenesis of SS [20,51]. Thus, sustained expression of C-terminally truncated Stat5 pro- teins may be one mechanism adopted by indolent malignant T cells in SS to escape apoptosis.

Fig. 4. Proteolytic processing of STAT proteins. STAT proteins may be cleaved at the C-terminus by the action of nuclear (A1,A2; progenitor myeloid cells, mature neutrophils, murine BMMC) and/or cytoplasmic (B1,B2; AML blasts, human PBMC) proteases. The activities of the proteases are generally not dependent on STAT-phosphorylation and therefore the protease can cleave activated (A1,B1) or unactivated STAT proteins (A2,B2). Unactivated, full-length STAT and STATc proteins can also shuttle between the cytoplasm and the nucleus in the absence of cytokine stimulation. The truncated STATc protein lacks the transactivation domain and behaves as a dominant-negative protein to functionally compete with the full-length protein. Pr, protease; TAD, transactivation domain.

Conclusions and perspectives

mast cells, respectively, although the physiological import- ance of these findings are unknown. Truncated Stat3c and Stat5c proteins generated by proteases have been shown to contribute significantly to the pathology of AML and CTCL. Thus, future identification of the relevant serine proteases and their natural inhibitors from myeloid cells and T cells will enhance our understanding of these diseases and also provide potential targets for therapeutic intervention by the rational design of drugs based on these proteins.

References

1. Darnell, J.E. Jr, Kerr, I.M. & Stark, G.R. (1994) Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science 264, 1415–1421. 2. Levy, D.E. & Darnell, J.E. Jr (2002) Stats: transcriptional control and biological impact. Nat. Rev. Mol. Cell Biol. 3, 651–662. 3. Greenhalgh, C.J. & Hilton, D.J. (2001) Negative regulation of cytokine signaling. J. Leukoc Biol. 70, 348–356.

4. Moriggl, R., Gouilleux-Gruart, V., Jahne, R., Berchtold, S., Gartmann, C., Liu, X., Hennighausen, L., Sotiropoulos, A., Groner, B. & Gouilleux, F. (1996) Deletion of the carboxyl- terminal transactivation domain of MGF-Stat5 results in sus- tained DNA binding and a dominant negative phenotype. Mol. Cell. Biol. 16, 5691–5700.

5. Wang, D., Stravopodis, D., Teglund, S., Kitazawa, J. & Ihle, J.N. (1996) Naturally occurring dominant negative variants of Stat5. Mol. Cell. Biol. 16, 6141–6148.

Evidence is accumulating to suggest that proteolytic processing is a general mechanism for the negative regula- tion of STAT protein function (Fig. 4). Truncated forms of Stat3, Stat5a, Stat5b and Stat6, generated by the proteolytic cleavage of the C-termini, have been identified in progenitor myeloid cells, mature neutrophils, mast cells and peripheral T cells. STATc proteins have different C-termini than STATb proteins and behave as functional dominant- negative proteins. Of the STAT proteases that have been characterized most are serine proteases, whose activities are regulated by the developmental or activation state of cells depending on the cellular context. However, their expres- sion and functional activities are not dependent on the presence of the target STAT protein itself and it is likely that other cellular targets exist for these proteases. The exact identities and mechanism of action of the individual proteases are currently unknown but they show STAT and cell-type specificity. Future cloning of the proteases from the different cell sources will reveal whether they belong to a family of related serine proteases. In addition, the cysteine protease, calpain, has also been shown to process Stat3 and Stat5 in platelets and Stat6 proteins in

6. Mui, A.L., Wakao, H., Kinoshita, T., Kitamura, T. & Miyajima, A. (1996) Suppression of interleukin-3-induced gene expression by

STAT signalling by proteolytic processing (Eur. J. Biochem. 271) 4619

(cid:1) FEBS 2004

a C-terminal truncated Stat5: role of Stat5 in proliferation. EMBO J. 15, 2425–2433. functionally distinct 80-kDa STAT5 isoform by IL-5 and GM- CSF in human eosinophils and neutrophils. Mol. Cell Biol. Res. Commun. 1, 95–101.

25. Chakraborty, A. & Tweardy, D.J. (1998) Granulocyte colony- stimulating factor activates a 72-kDa isoform of STAT3 in human neutrophils. J. Leukoc. Biol. 64, 675–680. 7. Yoo, J.Y., Huso, D.L., Nathans, D. & Desiderio, S. (2002) Spe- cific ablation of Stat3beta distorts the pattern of Stat3-responsive gene expression and impairs recovery from endotoxic shock. Cell 108, 331–344.

26. Epling-Burnette, P.K., Garcia, R., Bai, F., Ismail, S., Loughran, T.P., Djeu, J.Y., Jove, R. & Wei, S. (2002) Carboxy-terminal truncated STAT5 is induced by interleukin-2 and GM-CSF in human neutrophils. Cell Immunol. 217, 1–11. 8. Maritano, D., Sugrue, M.L., Tininini, S., Dewilde, S., Strobl, B., Fu, X., Murray-Tait, V., Chiarle, R. & Poli, V. (2004) The STAT3 isoforms alpha and beta have unique and specific functions. Nat. Immunol. 5, 401–409.

27. Rosen, R.L., Winestock, K.D., Chen, G., Liu, X., Hennighausen, L. & Finbloom, D.S. (1996) Granulocyte-macrophage colony- stimulating factor preferentially activates the 94-kD STAT5A and an 80-kD STAT5A isoform in human peripheral blood mono- cytes. Blood 88, 1206–1214.

9. Hoey, T., Zhang, S. & Schmidt, N., YuQ., Ramchandani, S., Xu, X., Naeger, L.K., Sun, Y.L. & Kaplan, M.H. (2003) Distinct requirements for the naturally occurring splice forms Stat4alpha and Stat4beta in IL-12 responses. EMBO J. 22, 4237–4248. 10. Azam, M., Lee, C., Strehlow, I. & Schindler, C. (1997) Func- tionally distinct isoforms of STAT5 are generated by protein processing. Immunity 6, 691–701. 28. Chakraborty, A. & Tweardy, D.J. (1998) Stat3 and G-CSF- induced myeloid differentiation. Leuk. Lymphoma 30, 433–442. 29. Djeu, J.Y. (1992) Cytokines and anti-fungal immunity. Adv. Exp. Med. Biol. 319, 217–223. 30. Wurster, A.L., Tanaka, T. & Grusby, M.J. (2000) The biology of 11. Smithgall, T.E., Briggs, S.D., Schreiner, S., Lerner, E.C., Cheng, H. & Wilson, M.B. (2000) Control of myeloid differentiation and survival by Stats. Oncogene 19, 2612–2618. Stat4 and Stat6. Oncogene 19, 2577–2584. 31. Sherman, M.A. (2001) The role of STAT6 in mast cell IL-4 pro- duction. Immunol. Rev. 179, 48–56.

32. Sherman, M.A., Secor, V.H., Lee, S.K., Lopez, R.D. & Brown, M.A. (1999) STAT6-independent production of IL-4 by mast cells. Eur. J. Immunol. 29, 1235–1242. 12. Lin, J.X. & Leonard, W.J. (2000) The role of Stat5a and Stat5b in signaling by IL-2 family cytokines. Oncogene 19, 2566–2576. 13. Snow, J.W., Abraham, N., Ma, M.C., Abbey, N.W., Herndier, B. & Goldsmith, M.A. (2002) STAT5 promotes multilineage hema- tolymphoid development in vivo through effects on early hema- topoietic progenitor cells. Blood 99, 95–101.

33. Sherman, M.A., Secor, V.H. & Brown, M.A. (1999) IL-4 pre- ferentially activates a novel STAT6 isoform in mast cells. J. Immunol. 162, 2703–2708. 14. Lin, J.X., Mietz, J., Modi, W.S., John, S. & Leonard, W.J. (1996) Cloning of human Stat5B. Reconstitution of interleukin-2- induced Stat5A and Stat5B DNA binding activity in COS-7 cells. J. Biol. Chem. 271, 10738–10744.

34. Sherman, M.A., Powell, D.R. & Brown, M.A. (2002) IL-4 induces the proteolytic processing of mast cell STAT6. J. Immunol. 169, 3811–3818.

15. Azam, M., Erdjument-Bromage, H., Kreider, B.L., Xia, M., Quelle, F., Basu, R., Saris, C., Tempst, P., Ihle, J.N. & Schindler, C. (1995) Interleukin-3 signals through multiple isoforms of Stat5. EMBO J. 14, 1402–1411.

16. Meyer, J., Jucker, M., Ostertag, W. & Stocking, C. (1998) Car- boxyl-truncated STAT5beta is generated by a nucleus-associated serine protease in early hematopoietic progenitors. Blood 91, 1901–1908. 35. Suzuki, K., Nakajima, H., Kagami, S., Suto, A., Ikeda, K., Hirose, K., Hiwasa, T., Takeda, K., Saito, Y., Akira, S. & Iwa- moto, I. (2002) Proteolytic processing of Stat6 signaling in mast cells as a negative regulatory mechanism. J. Exp. Med. 196, 27–38. 36. Nakajima, H., Suzuki, K. & Iwamoto, I. (2003) Lineage-specific negative regulation of STAT-mediated signaling by proteolytic processing. Cytokine Growth Factor Rev. 14, 375–380.

17. Lee, C., Piazza, F., Brutsaert, S., Valens, J., Strehlow, I., Jaros- inski, M., Saris, C. & Schindler, C. (1999) Characterization of the Stat5 protease. J. Biol. Chem. 274, 26767–26775.

37. Suzuki, K., Nakajima, H., Ikeda, K., Tamachi, T., Hiwasa, T., Saito, Y. & Iwamoto, I. (2003) Stat6-protease but not Stat5-pro- tease is inhibited by an elastase inhibitor ONO-5046. Biochem. Biophys. Res. Commun. 309, 768–773. 18. Piazza, F., Valens, J., Lagasse, E. & Schindler, C. (2000) Myeloid differentiation of FdCP1 cells is dependent on Stat5 processing. Blood 96, 1358–1365.

38. Oda, A., Wakao, H. & Fujita, H. (2002) Calpain is a signal transducer and activator of transcription (STAT)3 and STAT5 protease. Blood 99, 1850–1852.

39. Sato, K. & Kawashima, S. (2001) Calpain function in the mod- ulation of signal transduction molecules. Biol. Chem. 382, 743– 751. 19. Moriggl, R., Topham, D.J., Teglund, S., Sexl, V., McKay, C., Wang, D., Hoffmeyer, A., van Deursen, J., Sangster, M.Y., Bunting, K.D., Grosveld, G.C. & Ihle, J.N. (1999) Stat5 is required for IL-2-induced cell cycle progression of peripheral T cells. Immunity 10, 249–259. 40. Bowman, T., Garcia, R., Turkson, J. & Jove, R. (2000) STATs in oncogenesis. Oncogene 19, 2474–2488.

20. Lord, J.D., McIntosh, B.C., Greenberg, P.D. & Nelson, B.H. (2000) The IL-2 receptor promotes lymphocyte proliferation and induction of the c-myc, bcl-2, and bcl-x genes through the trans- activation domain of Stat5. J. Immunol. 164, 2533–2541.

41. Calo, V., Migliavacca, M., Bazan, V., Macaluso, M., Buscemi, M., Gebbia, N. & Russo, A. (2003) STAT proteins: from normal control of cellular events to tumorigenesis. J. Cell Physiol. 197, 157–168.

21. Mitchell, T.J., Whittaker, S.J. & John, S. (2003) Dysregulated expression of COOH-terminally truncated Stat5 and loss of IL2- inducible Stat5-dependent gene expression in Sezary Syndrome. Cancer Res. 63, 9048–9054. 42. Benekli, M., Baer, M.R., Baumann, H. & Wetzler, M. (2003) Signal transducer and activator of transcription proteins in leu- kemias. Blood 101, 2940–2954.

22. Meyer, T., Begitt, A., Lodige, I., van Rossum, M. & Vinkemeier, U. (2002) Constitutive and IFN-gamma-induced nuclear import of STAT1 proceed through independent pathways. EMBO J. 21, 344–354. 43. Xia, Z., Baer, M.R., Block, A.W., Baumann, H. & Wetzler, M. (1998) Expression of signal transducers and activators of tran- scription proteins in acute myeloid leukemia blasts. Cancer Res. 58, 3173–3180.

23. Zeng, R., Aoki, Y., Yoshida, M., Arai, K. & Watanabe, S. (2002) Stat5B shuttles between cytoplasm and nucleus in a cytokine- dependent and -independent manner. J. Immunol. 168, 4567–4575. 24. Caldenhoven, E., van Dijk, T.B., Raaijmakers, J.A., Lammers, J.W., Koenderman, L. & de Groot, R.P. (1999) Activation of a 44. Xia, Z., Sait, S.N., Baer, M.R., Barcos, M., Donohue, K.A., Lawrence, D., Ford, L.A., Block, A.M., Baumann, H. & Wetzler, M. (2001) Truncated STAT proteins are prevalent at relapse of acute myeloid leukemia. Leuk. Res. 25, 473–482.

4620 L. Hendry and S. John (Eur. J. Biochem. 271)

(cid:1) FEBS 2004

lymphoma: insights gained through cross-talk between the clinic and the bench. Leuk. Lymphoma 44, 1697–1703.

45. Benekli, M., Xia, Z., Donohue, K.A., Ford, L.A., Pixley, L.A., Baer, M.R., Baumann, H. & Wetzler, M. (2002) Constitutive activity of signal transducer and activator of transcription 3 pro- tein in acute myeloid leukemia blasts is associated with short disease-free survival. Blood 99, 252–257. 49. Rook, A.H., Vowels, B.R., Jaworsky, C., Singh, A. & Lessin, S.R. (1993) The immunopathogenesis of cutaneous T-cell lymphoma. Abnormal cytokine production by Sezary T cells. Arch. Dermatol. 129, 486–489.

50. Eriksen, K.W., Kaltoft, K., Mikkelsen, G., Nielsen, M., Zhang, Q., Geisler, C., Nissen, M.H., Ropke, C., Wasik, M.A. & Odum, N. (2001) Constitutive STAT3-activation in Sezary syndrome: tyrphostin AG490 inhibits STAT3-activation, interleukin-2 receptor expression and growth of leukemic Sezary cells. Leukemia 15, 787–793.

46. Bovolenta, C., Camorali, L., Mauri, M., Ghezzi, S., Nozza, S., Tambussi, G., Lazzarin, A. & Poli, G. (2001) Expression and activation of a C-terminal truncated isoform of STAT5 (STAT5 Delta) following interleukin 2 administration or AZT mono- therapy in HIV-infected individuals. Clin. Immunol. 99, 75–81. 47. Xia, Z., Salzler, R.R., Kunz, D.P., Baer, M.R., Kazim, L., Baumann, H. & Wetzler, M. (2001) A novel serine-dependent proteolytic activity is responsible for truncated signal transducer and activator of transcription proteins in acute myeloid leukemia blasts. Cancer Res. 61, 1747–1753. 51. Antov, A., Yang, L., Vig, M., Baltimore, D. & Van Parijs, L. (2003) Essential role for STAT5 signaling in CD25+CD4+ regu- latory T cell homeostasis and the maintenance of self-tolerance. J. Immunol. 171, 3435–3441. 48. Berger, C.L. & Edelson, R.L. (2003) Current concepts of the immunobiology and immunotherapy of cutaneous T cell