The Role of Cytokine-Inducible SH2 Domain-Containing Protein (CISH) in the Regulation of Basal and Cytokine-Mediated Myelopoiesis

Cytokine-inducible SH2 domain-containing protein (CISH) is a member of the suppressor of cytokine signaling (SOCS) family of negative feedback regulators shown to play crucial roles in lymphoid cell development and function as well as appetite regulation. It has also been implicated in the control of signaling downstream of the receptors for the cytokines granulocyte/macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) in myeloid cells. To investigate the physiological role of CISH in myelopoiesis, mice deficient in CISH were analyzed basally and in response to administration of these cytokines. CISH knockout (KO) mice possessed basally elevated neutrophils in the blood, bone marrow, and spleen compared to wild-type (WT) mice. During GM-CSF-induced myelopoiesis, the frequency of neutrophils, myeloid dendritic cells (DCs), and CFU-M in the bone marrow was higher in the KO, as were the neutrophils and CFU-G in the spleen. In contrast, no differences were observed between KO and WT mice during G-CSF-induced myelopoiesis apart from an elevated frequency of CFU-G and CFU-M in the spleen. This work has identified a role for CISH in the negative regulation of granulopoiesis, including that mediated by GM-CSF.

Previous studies have demonstrated that CISH plays a number of key roles in lymphoid cells. For instance, transgenic mice expressing CISH from the β-actin promoter showed altered T and NK cell responses due to partial suppression of IL-2R signaling [7]. CISH knockout (KO) mice showed preferential T cell differentiation into Th2 and Th9 cells [4] and Treg cell differentiation into Th2 cells [8] due to the dysregulation of IL-4R signaling, resulting in the spontaneous development of allergic pulmonary disease with excessive eosinophil influx [4,8]. CISH deletion in NK cells resulted in an increase in NK cell proliferation, differentiation, and cytotoxicity against tumors due to enhanced Int. J. Mol. Sci. 2023, 24, 12757 2 of 10 IL-15R signaling [5]. CISH ablation also resulted in altered CD8+ T cell functionality due to dysregulated TCR signaling [9].
However, our understanding of the physiological role of CISH in the context of myelopoiesis remains incomplete. Studies ex vivo have demonstrated that GM-CSF can substantially induce CISH expression [3,10] and also implicated CISH in the negative regulation of GM-CSF-mediated bone marrow-derived dendritic cell (BMDC) development [11]. Additional in vivo work has revealed a key role for CISH in limiting GM-CSF receptor signaling in inflammatory states [10]. Other in vitro studies have shown that G-CSF can also induce modest CISH expression, with CISH able to interact indirectly with the G-CSF receptor, suggesting that CISH may regulate G-CSFR signaling [12]. To further explore the regulation of myelopoiesis by CISH a recently described Cish KO mouse line [13] was analyzed with respect to basal and cytokine-induced myelopoiesis.

Role of CISH in Basal Myelopoiesis
To investigate a potential role for CISH in basal myelopoiesis, the blood, bone marrow, and spleen of 11-week-old Balb/c Cish +/+ (wild-type, WT) and Cish −/− (KO) mice [13] were examined. Differential counts revealed a significant increase in total neutrophils in the peripheral blood of Cish −/− compared to Cish +/+ mice ( Figure 1A), but no other changes were observed, including in the bone marrow ( Figure 1B) and spleen ( Figure 1E). However, analysis of specific myeloid populations by FACS [14] identified an increased frequency of neutrophils and total myeloid cells in Cish −/− mice in both the bone marrow ( Figure 1C) and spleen ( Figure 1F). Colony-forming assays showed no statistically significant differences in the bone marrow ( Figure 1D) or spleen ( Figure 1E) between genotypes, nor were changes observed in the overall cellularity of the bone marrow (WT: 4.36 ± 0.25 × 10 7 ; KO: 4.41 ± 0.27 × 10 7 ; p = 0.882) or spleen (WT: 6.21 ± 0.66 × 10 7 ; KO: 6.69 ± 0.71 × 10 7 ; p = 0.625). proliferation, differentiation, and cytotoxicity against tumors due to enhanced IL-15R signaling [5]. CISH ablation also resulted in altered CD8+ T cell functionality due to dysregulated TCR signaling [9]. However, our understanding of the physiological role of CISH in the context of myelopoiesis remains incomplete. Studies ex vivo have demonstrated that GM-CSF can substantially induce CISH expression [3,10] and also implicated CISH in the negative regulation of GM-CSF-mediated bone marrow-derived dendritic cell (BMDC) development [11]. Additional in vivo work has revealed a key role for CISH in limiting GM-CSF receptor signaling in inflammatory states [10]. Other in vitro studies have shown that G-CSF can also induce modest CISH expression, with CISH able to interact indirectly with the G-CSF receptor, suggesting that CISH may regulate G-CSFR signaling [12]. To further explore the regulation of myelopoiesis by CISH a recently described Cish KO mouse line [13] was analyzed with respect to basal and cytokine-induced myelopoiesis.
In the bone marrow, GM-CSF injection caused a significant increase in total and mature neutrophils and a reduction in lymphocytes and normoblasts in both Cish +/+ and Cish −/− mice, and band cells just in Cish +/+ mice, although there were no differences between genotypes with either vehicle or GM-CSF treatment ( Figure 2C,D). FACS analysis revealed a significant increase in the frequency of neutrophils and myeloid DCs, along with total DCs and total myeloid cells, in both genotypes, but they were significantly higher in each case in GM-CSF-treated Cish −/− mice compared to similarly treated Cish +/+ mice ( Figure 2E). Colony-forming assays revealed that GM-CSF administration resulted in a significant increase in the frequency of CFU-G and CFU-M in Cish −/− mice only, with the frequency of CFU-M in GM-CSF-treated Cish −/− mice elevated in comparison to Cish +/+ mice treated in the same manner ( Figure 2F). There was no statistically significant change in total cellularity in either Cish +/+ or Cish −/− mice following GM-CSF injection ( Figure 2G).
In the spleen, GM-CSF injection caused significant increases in the frequency of total and mature neutrophil populations in both Cish +/+ and Cish −/− mice, with a significant reduction in lymphocytes and normoblasts ( Figure 2H,I). However, no significant differences were observed between genotypes in response to either vehicle or GM-CSF treatment ( Figure 2H,I). FACS analysis showed that GM-CSF injection caused statistically significant increases in the frequency of myeloid DCs and total DCs in both genotypes, with a significant increase in total myeloid cells in Cish +/+ mice only, but again, no significant differences between genotypes were seen ( Figure 2J). Colony-forming assays revealed no significant changes in Cish +/+ mice in response to GM-CSF administration, but a statistically significant increase in the frequency of CFU-G in GM-CSF-treated Cish −/− mice was observed that was also significantly higher compared to similarly treated Cish +/+ mice ( Figure 2K). In the spleen, GM-CSF injection caused an increase in splenic cellularity in both Cish +/+ and Cish −/− mice, although this reached statistical significance only in Cish −/− mice ( Figure 2L). This resulted in the total number of spleen neutrophils being significantly elevated in response to GM-CSF in Cish −/− compared to Cish +/+ mice (WT: 1.44 ± 0.06 × 10 6 ; KO: 2.09 ± 0.17 × 10 6 ; p = 0.037).

Role of CISH in G-CSF-Induced Myelopoiesis
To directly investigate the potential in vivo role for CISH in G-CSF-mediated myelopoiesis, Cish +/+ and Cish −/− mice were injected with G-CSF, and the blood, bone marrow, and spleen analyzed. G-CSF injection resulted in an obvious and statistically significant increase in blood neutrophils-total and mature-in both Cish +/+ and Cish −/− mice, with a concomitant decrease in lymphocytes ( Figure 3A,B). However, G-CSF treatment failed to elicit significant differences between Cish +/+ and Cish −/− mice ( Figure 3A,B).   In the bone marrow, G-CSF injection resulted in significant increases in the percentage of total neutrophils, metamyelocytes, and mature neutrophils as well as decreases in lymphocytes and normoblasts for both genotypes, with band cells increasing just in Cish −/− mice ( Figure 3C,D). However, no significant differences between Cish +/+ and Cish −/− mice were observed ( Figure 3C,D). FACS analysis demonstrated a significant increase in the frequency of macrophages, neutrophils, and total myeloid cells in both Cish +/+ and Cish −/− mice following G-CSF treatment, but no statistically significant differences were observed between genotypes apart from an increased frequency of myeloid cells and neutrophils in vehicle-treated Cish −/− compared to Cish +/+ mice ( Figure 3E). Colony-forming assays failed to reveal significant changes between G-CSF-treated and vehicle-treated mice of either genotype, nor between Cish +/+ and Cish −/− mice, with either treatment ( Figure 3F). G-CSF injection resulted in a reduction in bone marrow cellularity in both Cish +/+ and Cish −/− mice, although this only reached statistical significance in Cish −/− mice ( Figure 3G).
In the spleen, G-CSF treatment caused increases across most neutrophil populations of both genotypes, reaching significance for total neutrophils, promyeloblasts, and mature neutrophils, with lymphocytes significantly decreasing, but again no differences between genotypes were seen ( Figure 3H,I). FACS analysis further demonstrated that G-CSF injection resulted in a significant increase in the frequency of neutrophils and total myeloid cells in both Cish +/+ and Cish −/− mice, although there were no differences between genotypes ( Figure 3J). Colony-forming assays also revealed a statistically significant increase in the frequency of CFU-G, CFU-M, and CFU-GM following G-CSF treatment in both Cish +/+ and Cish −/− mice, with the frequency of CFU-G and CFU-M significantly elevated in G-CSF-treated Cish −/− compared to similarly treated Cish +/+ mice ( Figure 3K). G-CSF treatment caused a statistically significant increase in splenic cellularity in both Cish +/+ and Cish −/− mice, but there was no difference between genotypes ( Figure 3L).

Discussion
CISH has been identified as a physiological regulator of immune cell development and function [2], particularly through its actions on the IL-4R [4] and T cell receptor [9] in specific T cell populations, the IL-15R in NK cells [5], and the GM-CSFR in DCs and other myeloid cells [10,11], with in vitro studies suggesting a potential role in regulating G-CSFR [12]. This study used a recently described Cish-deficient mouse line [13] to investigate the impact of CISH on myelopoiesis. The results indicate that CISH contributes to the regulation of both basal and cytokine-induced myelopoiesis mediated by GM-CSF, principally impacting neutrophil numbers (Table 1). Table 1. Summary of significant differences in specific myeloid cell populations between Cish +/+ (WT) and Cish −/− (KO) mice.
An analysis of basal myelopoiesis revealed small but significant differences in Cish −/− compared to Cish +/+ mice. There was a significant increase in total circulating neutrophils, with an increased frequency of CD11b + CD11c − Ly6G + neutrophils-and, indeed, total myeloid cells-in the bone marrow and spleen of Cish −/− mice (Figure 1). Not all of these parameters reached significance in the control mice used in the GM-CSF ( Figure 2) and G-CSF (Figure 3) studies, but the majority showed a similar trend, with the smaller numbers likely impacting the ability to demonstrate statistical significance. An analysis of mice on a C57/BL6 background [13] also confirmed a significant increase in bone marrow myeloid cells (WT: 10.14 ± 1.18%; KO: 13.90 ± 1.01%; p = 0.037) and neutrophils (WT: 7.37 ± 0.93%; KO: 10.34 ± 0.73%; p = 0.031). Overall, these results suggest a non-redundant role for CISH in the regulation of basal neutrophil levels. Amongst other SOCS proteins, SOCS3 has also been found to have a role in basal granulopoiesis, with hematopoietic-specific SOCS3 deletion resulting in neutrophilia and splenomegaly, although only in more mature mice, with this attributed to enhanced G-CSF signaling [15].
This led us to investigate potential cytokines that may have their myelopoietic actions regulated by CISH. GM-CSF represents a key regulator of myelopoiesis during inflammation as well as during DC development [16], with CISH implicated as a negative regulator of GM-CSFR signaling in the context of inflammation [10] and during GM-CSF-mediated bone marrow-derived dendritic cell (BMDC) development [11]. Therefore, Cish −/− and Cish +/+ mice were compared with respect to GM-CSF-mediated myelopoiesis. Following GM-CSF treatment, both Cish −/− and Cish +/+ mice exhibited a large increase in mature and total neutrophil populations as well as CD11b + CD11c + MHCII + myeloid DCs in the bone marrow and spleen, which is consistent with the results of other studies examining GM-CSF administration in mice [17,18]. GM-CSF treatment also caused a statistically significant increase in total blood neutrophils, CFU-G and CFU-M in the bone marrow, and CFU-G and overall cellularity in the spleen only in Cish −/− mice, with the frequency of bone marrow neutrophils, myeloid DCs, and CFU-M and spleen CFU-G significantly higher than in similarly treated Cish +/+ mice. However, because of the enhanced cellularity, the total number of splenic neutrophils in Cish −/− mice also significantly exceeded those in Cish +/+ mice. Collectively, this suggests that CISH plays a negative regulatory role in the GM-CSF-mediated production of both myeloid DCs and neutrophils in the bone marrow and spleen. The former is consistent with a previous study showing that CISH knockdown led to excessive GM-CSF-mediated DC production from isolated bone marrow cells [11]. Meanwhile, others have shown that the absence of CISH results in an enhanced responsiveness of neutrophils to GM-CSF [10]. These authors showed that Cish mRNA was strongly induced by GM-CSF, with CISH able to directly interact with the GM-CSFR beta-chain to regulate cell surface receptor levels, thereby impacting the length of activation of the associated JAK2 and the downstream STAT5 transcription factor, providing clear mechanistic details.
G-CSF is a critical regulator in both basal and emergency granulopoiesis [19,20], with G-CSF demonstrated to induce CISH expression and a suggestion that it may regulate G-CSFR signaling [12]. Therefore, the potential role of CISH in G-CSF-mediated myelopoiesis was also investigated. Following G-CSF treatment, Cish +/+ mice exhibited a dramatic increase in multiple neutrophil populations in the blood, bone marrow, and spleen, with the difference most evident in the latter, where the frequency of these and various CFU populations were markedly increased in concert with a substantial enhancement in overall cellularity, reflected in the overall size of the spleen (WT+PBS: 0.100 ± 0.05 g; WT+G-CSF: 0.283 ± 0.10 g; p = 0.000074), which is consistent with previous studies [21]. Cish −/− mice showed similar significant increases in the same populations following G-CSF treatment, with only CFU-G and CFU-M frequencies in the spleen significantly different from those in Cish +/+ mice. An analysis of G-CSF-mediated activation of STAT3 and STAT5 in bone marrow cells revealed no difference in extent or duration between Cish +/+ and Cish −/− mice ( Figure S1). These results suggest that CISH does not exert a major redundant negative regulatory role in G-CSFR signaling, including that involved in G-CSF-mediated neutrophil production. This is in stark contrast with SOCS3, with hematopoietic-specific Socs3 KO mice exhibiting a prolonged and excessive response to G-CSF injection, leading to a significant increase in cells along the neutrophil lineage [15]. This difference may relate to the G-CSF induction of CISH being less than that of SOCS3-and, indeed, much less than the GM-CSF induction of CISH [10]-as well as the inability of CISH to directly dock to the G-CSFR [12].
CISH has been implicated in the susceptibility of humans to a variety of infectious diseases [22,23]. Therefore, it remains intriguing as to how the differences in innate immunity in Cish −/− mice described here might contribute to this. Studies exploring viral and parasitic infections in Cish −/− mice are currently underway that aim to directly address the impacts of CISH on the pathogenesis of these agents.

Animal Husbandry
This study used previously described Cish +/+ wild-type (WT) and Cish −/− knock-out (KO) mice, principally those on a Balb/c background but also some C57/BL6 [13]. Mice were maintained on a standard rodent chow diet and experienced a 12-hour light/dark cycle, with their genotype determined as described [13]. All animal work was carried out with the approval of the Deakin University Animal Ethics Committee, which is subject to the Australian Code for the Responsible Conduct of Research. ARRIVE 2.0 guidelines were followed throughout. The sample size was based on preliminary experiments, with no blinding or randomization performed. Mice showing signs of illness were excluded from the study. Cytokine administration involved intraperitoneal injection of 10-week-old female mice with 6 µg/kg recombinant human G-CSF (rhG-CSF) (Ristempa, Pegfilgrastim) in 150 µL sterile phosphate buffered saline (PBS) that exhibits equivalent binding as its murine counterpart [24] and 30 µg/kg recombinant mouse GM-CSF (rmGM-CSF) (Cat# 415-ML-050, R&D Systems, In Vitro Technologies, Noble Park North, Australia) in 100 µL PBS -or equivalent volumes of PBS alone -daily for 5 consecutive days (basal) with humane culling 1 day later for tissue collection.

Tissue Collection and Histochemistry
Blood was collected from the tail prior to euthanasia by cervical dislocation, with single-cell suspensions generated from aspirated bone marrow or dissected spleen and passed through a 40 µm cell strainer. An aliquot of each cell suspension was mixed with trypan blue, and vital cell counts were performed with a Countess™ Automated Cell Counter (Invitrogen Australia Pty. Ltd., Mount Waverley, Australia). Smears were prepared on poly-L-lysine-coated glass slides (ProSciTech, Kirwan, Australia) using a cytospin funnel (Thermo Fisher Scientific Australia Pty. Ltd., Scoresby, Australia) and fixed in absolute methanol for 1 min prior to staining with 10% (v/v) Giemsa (Sigma-Aldrich Pty. Ltd., Castle Hill, Australia) for 20 min. Slides were examined on a Leica DME stereomicroscope and differential counts performed, with images captured on a Leica DFC290 digital camera controlled by the Leica Application Suite for Windows (Leica Microsystems, Macquarie Park, Australia).

Colony-Forming Assays
A total of 5 × 10 5 cells from the bone marrow or spleen were added to 5 mL of methylcellulose media (R&D System), and 1 mL was poured into a 35 mm dish (Thermo Fisher Scientific). These were incubated in a humid atmosphere at 37 • C with 5% CO2, and the number of CFU-G, CFU-M, and CFU-GM was enumerated on day 14.

Electropheoretic Mobility Shift Assays
Isolated bone marrow cells were placed in a tissue culture flask for 1 h, with the non-adherent cells transferred to a fresh flask for 3 h before stimulation of 1 × 10 6 cells with rhG-CSF for up to 120 min, following which nuclear extracts were prepared and electrophoretic mobility shift assays (EMSA) performed to analyze activation of STAT3 and STAT5, as described [25].

Statistical Analysis
Physiological data were analyzed with GraphPad Prism 8.0 using a two-way analysis of variance (ANOVA)/Tukey's multiple comparison test, or multiple t-tests using the Bonferroni-Dunn method, with p < 0.05 considered significant in all cases.