See accompanying article by Collazo et al.
Two SHIPs passing in the middle of the immune system
Article first published online: 17 JUL 2012
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
European Journal of Immunology
Volume 42, Issue 7, pages 1681–1684, July 2012
How to Cite
Corey, S. J., Mehta, H. M. and Stein, P. L. (2012), Two SHIPs passing in the middle of the immune system. Eur. J. Immunol., 42: 1681–1684. doi: 10.1002/eji.201242706
- Issue published online: 17 JUL 2012
- Article first published online: 17 JUL 2012
- Accepted manuscript online: 14 JUN 2012 03:21AM EST
- Manuscript Accepted: 4 JUN 2012
- Manuscript Revised: 2 JUN 2012
- Manuscript Received: 28 MAY 2012
- Signal transduction;
- T cells;
- Innate immunity;
Immunity requires a complex, multiscale system of molecules, cells, and cytokines. In this issue of the European Journal of Immunology, Collazo et al. [Eur. J. Immunol. 2012. 42: 1785–1796] provide evidence that links the lipid phosphatase SHIP1 with the coordination of interactions between regulatory T (Treg) cells and myeloid-derived suppressor cells (MDSCs). Using conditional knockouts of SHIP1 in either the myeloid or T-cell-lineage of mice, the authors show that the regulated development of Treg cells is controlled directly by cell-intrinsic SHIP1, and indirectly by extrinsic SHIP1 control of an unknown myeloid cell. Regulation of MDSCs is also determined by SHIP1 in an extrinsic manner, again via an as-yet-unknown myeloid cell. Furthermore, this extrinsic control of Treg cells and MDSCs is mediated in part by increased production of G-CSF, a growth factor critical for the production of neutrophils, in SHIP1-deficient mice. Thus, a physiologically important implication of this report is the collaboration between the innate and adaptive immune systems in fine tuning of Treg cells as discussed in this commentary.
The vertebrate immune system has evolved to navigate constantly between the need to fight off the infections by viruses, bacteria, parasites, and fungi, and the need to keep inflammation in check through a highly co-ordinated and exceedingly complex network of cells, cytokines, and microenvironments. The components of this immune system are copious and diverse and vertebrate immunity may be viewed as a combination of the innate and adaptive immune systems.
Hematopoietic stem cells generate two distinct lineages of leukocytes, arising from either a common myeloid progenitor (CMP) or a common lymphoid progenitor (CLP) (reviewed in ). The CMP generates the granulocyte/macrophage progenitor, which in turn spawns a number of highly specialized leukocytes. Of those, the neutrophil serves as a major bulwark in innate immunity by phagocytosing, releasing lytic proteases and collagenases, generating reactive oxygen species, and producing either pro- or anti-inflammatory mediators (reviewed in ). Neutrophils have long been considered to reside and function at the bottom of the cellular hierarchy of the immune system, being constantly produced as a consequence of their short lifespan (∼2 days) resulting from their death by apoptosis or consumption during host defense against microbes. The T cell and all of its subtypes have been viewed as comprising the upper echelons, coordinating and regulating a wide range of host defense functions, and being long lived.
Multiple lines of evidence suggest that there are extensive interactions between the cellular components of the innate and adaptive immune systems. Neutrophils function side-by-side with Th17 cells, mediating neutrophil recruitment and function through their production of cytokines such as G-CSF, GM-CSF, and IL-8 . In the middle of the immune hierarchy, straddling both the innate and adaptive systems, reside dendritic cells and macrophages, the latter being a myeloid cell that both phagocytoses (innate) and produces a number of inflammatory mediators (adaptive). Another myeloid cell type, the myeloid-derived suppressor cell (MDSC) is heterogeneous in nature, and acts to suppress T cells and their functions (reviewed in ). In many respects MDSCs appear to be immature myeloid precursors with diverse functions. In mice, MDSCs are characterized by expression of Gr-1, CD11b, and other neutrophil or macrophage markers. By affecting both innate and adaptive immune systems, MDSCs suppress immunity through the production of cytokines, such as IL-10, and the production of reactive oxygen species, nitric oxide, and arginase (reviewed in ).
Multiscale analysis has become popular within the field of systems biology as this approach seeks to understand the complex behavior by revealing systems’ properties due to components and their interactions at a number of different levels: molecular, cellular, tissue, organismal, and environmental (Fig 1). In this issue of the European Journal of Immunology, Collazo et al.  extend, at the molecular level, our understanding of the interactions between the innate and adaptive immune systems by determining the role played by SHIP1 (or rather the consequences of its absence in specific cell types) in regulating MDSC and Treg-cell numbers. The authors show by using mice deficient in SHIP1 in either myeloid- or T-cell-lineages that the development of Treg-cell numbers requires both T-cell-lineage intrinsic and extrinsic expression of SHIP1, whereas the development of MDSC numbers requires myeloid-lineage intrinsic expression, but not cell-intrinsic since SHIP1 is not expressed in MDSCs. The authors show that cell-extrinsic effects of SHIP1 on both MDSC and Treg-cell numbers involve an unidentified myeloid cell perhaps via increased expression of cytokines and growth factors such as G-CSF. Myelopoietic growth factors such as G-CSF influence increased production of MDSCs, whereas control of Treg-cell numbers may involve reduced stromal-derived factor (SDF-1) production and an unknown factor produced by the unknown myeloid cell.
The exquisite control of inositol phospholipids is of paramount importance for the regulation of cellular growth. These lipids are produced by the phosphatidylinositol-3′ kinase (PI-3′ kinase) pathway upon engagement of multiple receptor subclasses that include but are not limited to receptor tyrosine kinases, cytokine receptors, G-protein coupled receptors (GPCRs) and B-cell receptors (reviewed in ). This pathway is counteracted by two important phosphatases, phosphatase and tensin homolog (PTEN) and SHIP1/SHIP2 . While PTEN and SHIP2 are ubiquitously expressed in all cell types, SHIP1 appears to be restricted to hematopoietic cells in humans and mice. PTEN cleaves the 3′-phosphate of PI-(3,4,5)P3 to produce PI-(4,5)P2 and PTEN is inactivated in a variety of tumors, demonstrating that the level of PIP3 is crucial for cellular homeostasis. In contrast, evidence suggesting that SHIP1 could behave as a tumor suppressor gene has been scant, but targeted inactivation of SHIP1 in the mouse yields a phenotype that mimics many aspects of myeloproliferative disease. In addition, SHIP1-deficient mice exhibit lupus-like autoimmunity and osteoporosis . We have recently reported that bone marrow cells from patients with high-risk myelodysplastic syndromes or acute myeloid leukemia express only very low levels of SHIP1 protein but high levels of activated Akt , which is consistent with findings in mouse models of myeloproliferative disease.
SHIP is inositol (Ins) polyphosphate-5-phosphatase that degrades two polyphosphoinositides, products of PI-3′ kinase activity: PI-(3,4,5)P3 to PI-(3,4)P2 and Ins-(1,3,4,5,)P4 to Ins-(1,3,4)P3. PI-(3,4,5)P3 remains in the plasma membrane, serving as a docking site for proteins that contain lipid-binding motifs, such as the pleckstrin homology (PH) domain. Among the proteins that contain PH domains are the important signal transduction effectors, Akt and phospholipase Cγ. One of the products of SHIP, PI-(3,4)P2 appears to have a higher affinity than the PTEN degradation product PI-(4,5)P2 for the PH domain of Akt. Inositol-(1,3,4,5)P4 may be found in the aqueous compartment, where it may regulate intracellular calcium levels or serve to negatively regulate PI-3′ kinase signaling. In addition to its inositol phosphatase activity, SHIP contains other signaling motifs that confer upon it scaffolding properties. Its SH2 domain binds Shc, SHP-2, DOK, and the Gab adapter proteins. SHIP can recruit other signaling proteins through its two NPXY motifs, which when phosphorylated, bind to the phosphotyrosine binding (PTB) domains of Shc and Dok2. At its C-terminus, SHIP contains a polyproline motif that binds to SH3 motifs (reviewed in ).
When SHIP1 was ablated in the myeloid compartment through selective Cre expression, Collazo et al. found that the MDSC population was expanded in comparison with wild-type (WT) controls. To their surprise, when SHIP1 expression was deleted within the T-cell lineage, the number of MDSCs was unaffected despite increased numbers of Treg cells. The combination of lineage intrinsic and extrinsic effects mediated by total SHIP deficiency resulted in the highest increase in MDSC and Treg-cell numbers. What was even more striking was that the level of SHIP1 in MDSCs was undetectable. The investigators conclude that another SHIP1-expressing myeloid cell is acting through the production of an as-yet-unidentified factor(s) to exert this effect. One possible candidate factor is G-CSF, the most important cytokine for the production of granulocytes, and a cytokine found to be markedly elevated in the serum of SHIP1-deficient mice . G-CSF drives the proliferation of myeloid progenitor cells and their differentiation into granulocytes through the recruitment of protein tyrosine kinases and PI-3′ kinase . Collazo et al. conclude that downregulation of SHIP1 might serve as a switch to promote an immunosuppressive state in the peripheral lymphoid tissues. Such downregulation might occur either through translational suppression by miR-155, which targets SHIP1, or post-translational modification of SHIP1 by either protein tyrosine phosphorylation or ubiquitination [14, 15].
This work by Collazo et al. raises a number of intriguing questions. Are the increased Treg-cell numbers found in the spleen in Lys-SHIP (myeloid-specific deletion of SHIP1) mice thymically derived natural Treg (nTreg) cells or do the MDSCs induce Treg cell formation (inducible [iTreg] cells) from the naïve T-cell pool? In general, iTreg cells are thought to develop in response to a TGF-β signal. While MDSCs have been reported to produce TGF-β , in vitro culture work from other groups has demonstrated the existence of an alternative pathway where arginase activity rather than TGF-β is required for Treg-cell expansion . Moreover, it appears that the MDSCs exert their control on the nTreg-cell population . Normally, Treg cells express not only Foxp3 but also CD25. As previously reported by others , Collazo et al. also see a significant increase in an unusual Treg-cell population that lacks CD25 in mice with SHIP1-deficient T cells, raising questions as to the origin and function of this atypical Treg-cell population. SHIP1-deficient CD4+ T cells produce elevated levels of TGF-β , so it is possible that this unusual population are in fact iTreg cells. CD25− Foxp3+ T cells have been noted in patients with autoimmune diseases, such as systemic lupus erythermatosus (SLE), and in infliximab-treated arthritis patients . At present it is controversial as to whether this CD25− Foxp3+ T-cell subset is capable of suppressing T-cell responses. Nevertheless, it is important to gain a better understanding of the origin and functional characteristics of these Treg cells given the interest in expanding them ex vivo for therapeutic purposes.
The work by Collazo et al. also raises some important questions regarding the mechanism by which SHIP1 affects nTreg-cell development. A previous study using CD4-Cre mice suggested that loss of SHIP1 had no effect on nTreg-cell development . In contrast, Collazo et al. show that deletion of SHIP1 earlier in thymocyte ontogeny using a Lck-Cre transgene results in expansion of this lineage. This indicates that the nTreg cells most likely are committed by the time CD4 is activated; it will be of considerable interest to understand how SHIP1 affects Treg-cell lineage choice during early thymic development. The differential effect on Treg development caused by deletion of SHIP-1 at various times during T-cell development may be instructive in light of previous work on PTEN deletion in the T-cell lineage. Normally, nTreg cells are hyporesponsive to IL-2 because of weak activation of the Akt pathway . While nTreg-cell development appears unaffected when PTEN is excised using CD4-Cre, the cells are rendered extremely sensitive to IL-2. It will be of interest to determine whether the SHIP1-deficient nTreg cells behave similarly. While PTEN has been deleted using Lck-Cre, Treg cells were not examined . Given the stark contrast between Lck-Cre and CD4-Cre mediated excision of SHIP1, a similar comparison between PTEN deleted at different points in T-cell development will provide some much needed contrasts with the work on SHIP1 mutants.
Why are MDSCs so enriched following myeloid-specific deletion of SHIP1? Previous work showed that a variety of cytokines were elevated in the absence of SHIP1, including G-CSF . While neutralization of G-CSF clearly reduces MDSC numbers, it may not be the only mechanism. A previous report showed that neutrophils depleted of SHIP1 had a longer half-life , suggesting another possible mechanism for the increased MDSC numbers. Integrin cross-linking in neutrophils specifically results in elevated levels of PI-(3,4)P2, the product of SHIP1. Moreover, SHIP1, but not PTEN, plays a critical role in neutrophil migration; and, in the absence of SHIP1, neutrophils become “stuck” and migrate poorly . This could result in a buildup of MDSCs in secondary lymphoid sites, and by extension, create an environment rich in mediators that promote Treg-cell expansion. While SHIP1 is nearly undetectable in naïve MDSCs, perhaps a low level of this phosphatase is sufficient to control chemotaxis.
A therapeutic consideration arising from the work by Collazo et al. is how does this advance our understanding of the immunological consequences of pharmacological manipulation of SHIP activity? Pharmacologic targeting of SHIP has focused on either stimulating or abrogating its activity. Small molecular activators of SHIP1 targeting macrophage and mast cells resulted in better survival shock or anaphylactoid reactions in mice . SHIP inhibitors have also been pursued as a strategy to dampen the immune system either in vivo or ex vivo by increasing MDSCs and Treg-cell numbers [27, 28]. Thus, along with the finding from the study by Collazo et al. that SHIP1 is a critical regulator of MDSCs and Treg-cell numbers, the SHIP inhibitors used might see potential applications in transplantation and autoimmunity.
Conflict of interest
The authors declare no financial or commercial conflict of interest.
myeloid-derived suppressor cell