H-X.Y. and H-Y.W. contributed equally to this work.
Liver Biology and Pathobiology
Negative regulation of hepatocellular carcinoma cell growth by signal regulatory protein α1
Article first published online: 30 AUG 2004
Copyright © 2004 American Association for the Study of Liver Diseases
Volume 40, Issue 3, pages 618–628, September 2004
How to Cite
Yan, H.-X., Wang, H.-Y., Zhang, R., Chen, L., Li, B.-A., Liu, S.-Q., Cao, H.-F., Qiu, X.-H., Shan, Y.-F., Yan, Z.-H., Wu, H.-P., Tan, Y.-X. and Wu, M.-C. (2004), Negative regulation of hepatocellular carcinoma cell growth by signal regulatory protein α1. Hepatology, 40: 618–628. doi: 10.1002/hep.20360
- Issue published online: 30 AUG 2004
- Article first published online: 30 AUG 2004
- Manuscript Accepted: 21 MAY 2004
- Manuscript Received: 11 NOV 2003
- High-Tech Research and Development Program of China. Grant Number: 2001AA221021
- China Key Basic Research Program. Grant Number: 2002BA711A02-3
- National Natural Science Foundation of China. Grant Numbers: 30270686, 30370740
Signal regulatory protein (SIRP) α1 is a member of the SIRP family that undergoes tyrosine phosphorylation and binds SHP-2 tyrosine phosphatase in response to various mitogens. The expression levels of SIRPα1 were decreased in HCC tissues, compared with the matched normal tissues. Exogenous expression of wild type SIRPα1, but not of a mutant SIRPα1 lacking the tyrosine phosphorylation sites, in SIRPα1-negative Huh7 human HCC cells resulted in suppression of tumor cell growth both in vitro and in vivo. Treatment of Huh7 transfectants with EGF or HGF induced tyrosine phosphorylation of SIRPα1 and its association with SHP-2, which were accompanied by reduced ERK1 activation. Expression of SIRPα1 significantly suppressed activation of NF-κB and also sensitized Huh7 cells to TNFα or cisplatin-induced cell death. In addition, SIRPα1-transfected Huh7 cells displayed reduced cell migration and cell spreading in a fashion that was dependent on SIRPα1/SHP-2 complex formation. In conclusion, a negative regulatory effect of SIRPα1 on hepatocarcinogenesis is exerted, at least in part, through inhibition of ERK and NF-κB pathways. (HEPATOLOGY 2004;40:618–628.)
Signal regulatory protein (SIRP) α1, also known as Src homology-containing phosphotyrosine phosphatase substrate (SHPS) 1, is a member of the immunoglobulin-like receptor superfamily proteins. The putative extracellular region of SIRPα1 possesses 3 immunoglobulin-like domains with multiple N-linked glycosylation sites. The cytoplasmic region of SIRPα1 contains 2 immunoreceptor tyrosine-based inhibitory motifs (ITIMs) with 4 tyrosine residues that are phosphorylated in response to a variety of growth factors and integrin-mediated cell adhesion.1–3 This phosphorylation enables recruitment and activation of Src homology-containing phosphotyrosine phosphatase 2 (SHP-2) that in turn dephosphorylates specific protein substrates involved in mediating various physiological effects.
SHP-2, a widely expressed cytoplasmic tyrosine phosphatase with 2 Src homology-2 domains, has been implicated in growth factor-induced cell proliferation probably through activation of the Ras–mitogen-activated protein kinase (MAPK) cascade.4 SIRPα1/SHP-2 complex negatively or positively regulates intracellular signaling initiated either by tyrosine kinase-coupled receptors for growth factors or by cell adhesion to extracellular matrix proteins. For example, SIRPα1 overexpression in NIH3T3 cells inhibited DNA synthesis and MAPK phosphorylation following endothelial growth factor (EGF) or insulin stimulation.1 In contrast, overexpression of wild-type SHPS-1 (the murine homolog of SIRPα1) was reported to increase MAPK activity in response to insulin or integrin stimulation.5 SHPS-1 overexpression suppressed anchorage-independent cell growth and cancer dissemination; however, phosphorylation states of MAPK and c-jun NH2-terminal kinase (JNK) were almost similar.6 In fibroblasts expressing an SHPS-1 mutant lacking most of the cytoplasmic region, growth factor–induced JNK activation was enhanced, whereas extracellular signal-regulated kinase (ERK) activation was dependent on the kind of growth factor.7 Moreover, overexpression of SIRPα1 was reported to negatively regulate EGF-induced PI3-K activation and modulate nuclear factor (NF)-κB signaling.8, 9 Thus, the precise mechanism by which SIRPα1/SHP-2 complex regulates MAPK and other signaling pathways remains unclear.
Cell migration is crucial for multiple biological functions including embryonic development and tumor metastasis. Formation of SIRPα1/SHP-2 complex was implicated in regulation of cell migration. However, the effects of SIRPα1 on cell motility were also controversially reported. Overexpression of wild type SHPS-1 promoted CHO cell migration, whereas expression of SHPS-4Y mutant, which lacks the phosphorylation sites required for SHP-2 binding, had no effect.10 Likewise, fibroblasts homozygous for expression of an SHPS-1 mutant lacking most of the cytoplasmic region of this protein exhibited defective migration associated with increased formation of actin stress fibers and focal adhesions.7 In contrast, ectopically expressed SHPS-1/SIRPα1 in v-Src–transformed BALB/c3T3 cells or U87MG cells led to marked impairment of cell migration and cancer dissemination.6, 8 These observations therefore suggest that the functional consequences of tyrosine phosphorylation of SIRPα1 and its association with SHP-2 in cell migration have not been established.
CD47 (or integrin-associated protein [IAP]), an important component of the signaling pathway triggered by integrins or cell-cell adhesions, has been implicated as a ligand for SIRPα1.11 The CD47/SHPS-1 system was recently shown to inhibit cell migration by cell-cell contact10 and to play novel regulatory roles on immune cells.3 SIRPα1 has also been shown to bind to various adaptor proteins such as FyB/SLAP130, SKAP55 and Grb2.12 Thus, SIRPα1 appears to function in a variety of cellular signaling systems.
Recently it has been reported that the expression level of SIRPα1 was decreased in breast cancer tissues and leukemia myeloid cells, indicating a role of SIRPα1 in oncology.6, 13 In this study, we observed that SIRPα1 expression in hepatocellular carcinoma (HCC) tissues frequently seemed lower than that of paired normal tissues. Exogenous expression of SIRPα1 in the HCC cell line led to inhibition of cell cycle progression and cell growth, which is associated with reduced ERK1 activation after EGF or hepatocyte growth factor (HGF) stimulation. Furthermore, SIRPα1 conferred enhanced apoptosis following tumor necrosis factor (TNF) α or cisplatin treatment in Huh7 cells by negatively regulating NF-κB signaling. In addition, SIRPα1 expression also led to reduced cell spreading and migration. In brief, we present evidence that SIRPα1 functioned as a tumor suppressor in human HCC cells.
Patients and Methods
Patients and Samples.
A total of 36 HCC specimens and adjacent nontumorous liver counterparts were studied. All samples were collected at the Eastern Hepatobiliary Surgery Hospital, Shanghai, China.
Complementary DNA (cDNA) Constructs, Antibodies, and Reagents.
The human SIRPα1 and SIRPα1-4Y cDNAs contained in pLXSN1 (kindly provided by A. Ullrich, Max –Planck Institute, Martinsried, Germany) were digested with EcoRI and BamHI and ligated into expression vector pcDNA3.0 (Invitrogen, Carlsbad, CA). The plasmids pcDNA3-SHP-2 (WT) and pcDNA3-mIκBα were kindly provided by G. S. Feng (The Burnham Institute, La Jolla, CA) and C. Scheidereit (Mark –Delbruck Center, Berlin, Germany), respectively. Antibodies specific for poly(ADP-ribose) polymerase (PARP), total MAPKs and the phosphorylated and active ERK (Thr202/Tyr204), JNK (Thr183/Tyr185), p38 (Thr180/Tyr182), Akt (S473), and phospho-inhibitor of κB (IκB) kinase (IKK) α (Ser180)/IKKβ (Ser181) were purchased from Cell Signaling Technology (Beverly, MA) and used as previously described.14 Anti–IκBα1 was from IMGENEX (San Diego, CA) and anti–Bcl-xL from Oncogene (Bayer, Cambridge, MA). A polyclonal antibody reactive with SIRPα1 was raised against a GST fusion protein containing the C-terminal part of SIRPα1. The monoclonal antibody against phosphotyrosine (4G10) was obtained from Transduction Laboratories (BD Biosciences, San Diego, CA). HGF and TNF-α were obtained from PeproTech (Rocky Hill, NJ) and cisplatin was purchased from Sigma (St. Louis, MO). The PI3K inhibitor wortmannin (Calbiochem, Boston, MA) was used at a 100nM concentration.
Cell Culture and Transfection.
All cell lines were grown in Dulbecco's modified eagle medium (DMEM) containing 10% fetal bovine serum (FBS). Stable transfection was performed as previously described.15 Established clones were isolated and screened by Western blot analysis with anti-SIRPα1 antibody. Positive clones expressing similar levels of SIRPα1 or mutant protein were pooled and cultured in the presence of 200 μg/mL G418 for maintenance of the transgene expression.
Total RNA was prepared using TRIzol (Invitrogen), RNA (40 μg) was separated and transferred to a nitrocellulose membrane and probed using SIRPα1 cDNA labeled by random oligonucleotide priming (Promega, Madison, WI).
Immunoprecipitation and Immunoblotting.
Immunoprecipitation and immunoblotting analysis with whole-cell extracts were performed essentially as previously described.16
NF-κB transcriptional activity was determined using NF-κB–driven luciferase plasmid. pRL-TK (Promega) was used as internal control. A dual luciferase reporter assay was carried out according to the manufacturer's suggestions.
Cell Cycle Analysis.
Huh7 cells were synchronized in G0 by serum starvation for 3 days (in DMEM without serum), followed by stimulation in DMEM supplemented with 10% FBS. Progression through the cell cycle was monitored by detection of the DNA content as previously described.17
Cell Spreading and Migration Assay.
Growth Curves, Viability, and Focus Formation Assay.
Huh7 cells (105) were plated per well of 6-well plates. At each time point, the cells were trypsinized and counted. Each data point was performed in triplicate. The measurement of viable cell mass was performed with a Cell Counting Kit (Dojin Laboratories, Kumamoto, Japan) to count living cells by WST-8. To visualize the nuclei of Huh7 cells, DNA was stained with DAPI (1 μg/mL) and determined using fluorescence microscopy. The luciferase-based in vitro cell viability assay was performed as described.19 For assays of focus formation of stably transfected cells, previously described methods were employed.8
In Vivo Tumorigenicity Assay.
Huh7 cells were released from tissue culture dishes and washed in serum-free medium. Tumor cells were diluted with phosphate-buffered solution and injected into the mid-dorsum of BALB/c nude mice (4–6 weeks old) in a total volume of 0.1 mL (5× 106). Animals were inspected weekly for tumor development. Growing tumors were measured using vernier calipers, and tumor volume was calculated by the formula length × width2 × 0.52, which approximates the volume of an elliptical solid. Statistical analysis was performed by Student t test (2-tailed). The criterion for statistical significance was taken as P < .05.
Expression of SIRPα1 Was Decreased in HCC.
SIRPα1 has been shown to negatively regulate cellular responses induced by growth factors. To search for the involvement of SIRPα1 in tumorigenesis, we evaluated SIRPα1 expression in HCC tissues from 36 patients. Northern blot analysis revealed 2 major RNA transcripts of 2.5 and 3.9 kilobase (kb) in the normal tissues. Quantified data of tumors relative to paired normal tissues demonstrate that the expression of SIRPα1, especially the 2.5 kb transcript, was frequently down-regulated in cancerous tissues (Fig. 1A). Of 36 HCC tissues, 19 appeared to have a decreased expression of SIRPα1. These results were consistent with the 2 previous reports concerning SIRPα1 expression in human HCC tissue.20, 21 Immunohistochemical analysis showed that SIRPα1 was down-regulated in HCC tissues and a loss of SIRPα1 expression was observed in some tumor cells and areas (Fig. 1B). Of 4 established HCC cell lines detected (HepG2, Huh7, Sk-Hep1, and Changliver), the expression of SIRPα1 was significantly decreased in Huh7 and SK-Hep1 cells (Fig. 1C).These results suggest that SIRPα1 expression might be involved in some steps of tumor development.
To elucidate the effects of SIRPα1 on liver cell functions, Huh7 HCC-derived cell line was stably transfected with either wild type SIRPα1, or a carboxyl terminal SIRPα1 mutant (SIRPα1-4Y) in which 4 carboxyl terminal tyrosines have been mutated to phenylalanine, making this mutant incapable of binding to SHP-2. Positive clones expressing similar levels of SIRPα1 or mutant protein were pooled to rule out clone-specific results. As shown in Fig. 1D, SIRPα1 became tyrosine-phosphorylated upon pervanadate treatment, but SIRPα1-4Y did not, even though their expression was comparable.
Expression of SIRPα1 Reduced Cell Proliferation in Human HCC Cells.
To characterize the distribution of Huh7/pcDNA3, Huh7/SIRPα1 and Huh7/SIRPα1-4Y cells in the cell cycle, the progression of synchronized cells through the cell cycle was determined by DNA/flow cytometry analysis. As shown in Fig. 2A, SIRPα1 expression strongly delayed cell cycle progression. Growth curves also demonstrated a significant difference between the Huh7/SIRPα1 cells and Huh7/pcDNA3 or Huh7/SIRPα1-4Y cells (Fig. 2B), indicating that the alterations in cell cycle distribution resulted in decreased growth and reduced saturation densities. Furthermore, SIRPα1 expression nearly abolished the colony formation ability of Huh7 cells (Fig. 2C). These data suggest that SIRPα1 exerts an inhibitory effect on cell proliferation that was dependent on its carboxyl terminal sites mutated in the SIRPα1-4Y.
SIRPα1 Sensitized Huh7 Cells to TNF-α and Chemotherapeutic Agent-Induced Apoptosis.
To examine the effect of SIRPα1 on cell survival, an in vitro cell viability assay was performed using a luciferase assay.19 Following transfection with the luciferase reporter gene (pCMV-luc) and treatment with TNF-α, the luciferase activity in the Huh7/SIRPα1 cells was strongly reduced compared with that in Huh7/pcDNA3 and Huh7/SIRPα1-4Y cells (Fig. 3A). This suggested that SIRPα1 sensitized Huh7 cells to TNF-α–induced cell death. Furthermore, DAPI staining showed clear evidence of condensed and fragmented nuclei in Huh7/SIRPα1 cells, confirming the SIRPα1-mediated sensitization of TNF-α–induced apoptosis (Fig. 3B).
Cisplatin is a DNA-reactive agent commonly used in chemotherapy protocols for the treatment of HCC. An assessment of growth effects of Huh7 cells subjected to 24 hour exposure to cisplatin was performed using WST-8 assay to measure viable cell mass. As shown in Fig. 3C, cells harboring wild type SIRPα1 were more sensitive than the empty vector control and the mutant counterpart to the growth inhibition effects of cisplatin.
SIRPα1 Negatively Regulates Growth Factor–Induced ERK1 Activation.
Several groups have reported a modulation of MAPK activities by SIRPα1 or SHPS-1. To characterize the molecular mechanism responsible for the inhibitory effect of SIRPα1 on cell growth, we examined the effects of EGF and HGF, 2 major inducers of liver cell growth and proliferation, on the activity of ERK. As shown in Fig. 4A, EGF/HGF stimulation induced tyrosine phosphorylation of SIRPα1 and binding to SHP-2. Although both growth factors induced marked activation of P44 (ERK1) and P42 (ERK2) isoforms of ERK in Huh7/pcDNA3 and Huh7/SIRPα1-4Y cells, their effects on ERK1 activity was substantially lower in the Huh7/SIRPα1 cells (Figs. 4B and C). Interestingly, TNF-α–induced ERK1 activation was similarly suppressed in Huh7/SIRPα1 cells (Fig. 4D), suggesting that growth factor- or cytokine-induced ERK1 activation was constitutively inhibited by SIRPα1 in Huh7 cells.
Because tyrosine-phosphorylated SIRPα1 was preferentially associated with SHP-2 that has been shown to positively regulate ERK activation, it is conceivable that SIRPα1-mediated ERK repression might be due to SHP-2 sequestration. To test this possibility, we transiently transfected the Huh7/SIRPα1 cells with empty vector or wildtype SHP-2 and examined ERK activity following EGF treatment. As expected, overexpression of SHP-2 restored ERK1 activation in Huh7/SIRPα1 cells, indicating that ERK activation was regulated by SIRPα1 via SHP-2.
SIRPα1 Suppressed NF-κB Transcriptional Activation.
TNF-α and cisplatin have been reported to activate both the JNK and NF-κB pathways that transduced antiapoptotic signals in Huh7 cells.22 To explore the mechanism of SIRPα1-mediated apoptosis induction, we first measured JNK and p38 activity in cells at different time points after stimulation with TNF-α or cisplatin. However, no significant difference was found in JNK or p38 activation after TNF-α or cisplatin stimulation (Fig. 5).
Having obtained these results, we shifted our attention to the NF-κB pathway. We monitored the NF-κB activity in stable Huh7 transfectants by measuring NF-κB–driven luciferase reporter activity. As shown in Fig. 6A, Huh7/SIRPα1 cells indeed displayed an evident decrease in basal NF-κB activity with respect to Huh7/pcDNA3 and Huh7/SIRPα1-4Y cells. Interestingly, TNF-α failed to further enhance the NF-κB–driven reporter activity. Introduction of a super suppressor form of IκBα alone led to high levels of cell death (Fig. 6B). These observations indicated that the basal activity of endogenous NF-κB was important in maintaining the survival of Huh7 cells. Notablely, NF-κB–mediated transcription of Bcl-xL, an antiapoptotic member of Bcl-2 family, was also reduced in Huh7/SIRPα1 cells compared with controls (Fig. 6C). Since the antiapoptotic effect of NF-κB as well as that of Bcl-xL involves prevention of caspase activation,23 we studied the cleavage of PARP, a caspase-3 substrate, by Western blot analysis. In Huh7/pcDNA3 and Huh7/SIRPα1-4Y cells (Fig. 6D), the cleavage product was first detected 4 hours after TNF/CHX stimulation. In Huh7/SIRPα1 cells, in contrast, PARP cleavage had already started after 2 hours, indicating an early induction of caspase-3–like activity in the SIRPα1-expressing cells.
SHP-2 Was Involved in SIRPα1-Mediated NF-κB Suppression.
To understand the molecular basis of SIRPα1-mediated NF-κB suppression, we assayed Akt and IKKα/β activities following the treatment of cells with TNF-α because both of them were required for TNF-α–induced NF-κB activation.24 However, SIRPα1 expression did not affect their activation kinetics in response to TNF-α. Furthermore, IκBα proteolysis was also independent of SIRPα1 on TNF-α treatment (Fig. 7A). In addition, although wortmannin completely inhibited NF-κB–mediated transcription caused by addition of serum, the expression of SIRPα1 did not inhibit serum-induced NF-κB activation. The phosphorylation status of Akt after treatment with serum was consistently unchanged on SIRPα1 expression (Fig. 7B). These findings placed the defect in NF-κB activation regardless of the canonical PI3K/Akt pathway.
Because the SIRPα1-4Y mutant was not tyrosine phosphorylated and thus was incapable of binding the SHP-2 implicated in NF-κB activation,14, 25 we speculated that SIRPα1-mediated NF-κB suppression may also be due to SHP-2 sequestration by SIRPα1. To test this hypothesis, we analyzed NF-κB activity and survival of Huh7/SIRPα1 cells transiently overexpressing SHP-2. As shown in Fig. 7C, introduction of SHP-2 augmented NF-κB–driven reporter activity and, importantly, diminished TNF-induced cell death (Fig. 7D). These results indicated that SHP-2 was required for protection from SIRPα1-mediated TNF-α cytotoxicity.
SIRPα1 Regulated HCC Cell Spreading and Migration.
Cell adhesion to extracellular matrix could induce tyrosine phosphorylation of SIRPα1 and its subsequent association with SHP-2, suggesting that this protein also functioned in integrin-mediated signaling.26, 27 Next we examined the role of SIRPα1 in spreading of Huh7 cells on the extracellular matrix, a process promoted by integrins. As depicted in Fig. 8A, Huh7/SIRPα1 cells spread at a much reduced rate, with only about 30% of the cells spreading well by 60 minutes when more than 80% of cells containing empty vector or SIRPα1-4Y were well extended and displayed a flat morphology.
To further explore the effect of SIRPα1 on cell motility, we examined cell migration by using “wound healing” and Boyden chamber assays. As shown in Figs. 8B and C, Huh7/SIRPα1 cells exhibited a reduced rate of cell migration compared to control cells. Taken together, these data suggest that functional phosphorylated SIRPα1 negatively regulates cell migration and cell spreading.
Expression of Wild Type SIRPα1 Inhibited Tumorigenesis In Vivo.
On the basis of the inhibitory effects of SIRPα1 on the proliferation, viability, and invasiveness of human HCC Huh7 cells in vitro, we then compared growth of the Huh7 clones after injection into athymic mice. Compared with empty vector or the SIRPα1 mutant, introduction of wild type SIRPα1 significantly inhibited tumor growth and prolonged the survival rate of tumor-bearing mice (Figs. 9A and B). Although the Huh7/SIRPα1-4Y cells showed an intermediate suppression between Huh7/pcDNA3 and Huh7/SIRPα1 cells, the survival of tumor-bearing mice in Huh7/SIRPα1-4Y and Huh7/pcDNA3 groups was not statistically significant. This observation confirmed the role of SIRPα1 in tumor growth inhibition.
In this study, we found that expression levels of SIRPα1 were decreased in human HCC tissues and then characterized phenotypic alterations that resulted from SIRPα1 expression in the human HCC-derived Huh7 cell line. Effects were observed in the regulation of 3 important cellular activities: proliferation, apoptosis, and migration. We demonstrated that exogenous SIRPα1 expression significantly discouraged cell cycle progression and reduced the proliferation ability of these cells. Most strikingly, we demonstrated that SIRPα1 abrogated the colony formation ability of Huh7 cells and significantly suppressed tumorigenesis growth of xenografts in nude mice. In exploring the molecular mechanism responsible for SIRPα1-mediated cell growth inhibition, we found that ERK1 activity was diminished after EGF, HGF, or TNF-α stimulation in cells expressing SIRPα1. The mechanism of this selective effect is unknown; however, it is now clear that the cytoplasmic ITIM region of SIRPα1 is crucial for proper regulation of the Ras-ERK pathway by these growth factors. Although the role of tyrosine phosphorylation in the ITIM region of SIRPα1 is controversial with regard to the regulation of ERK activity, this tyrosine phosphorylation is generally indicative of receptor-mediated recruitment. In most cases, it is probable that SIRPα1 acts as a scaffolding molecule of SHP-2 and recruits it in the vicinity of the cell membrane. Thus, the nature of the SIRPα1 function may depend on the function of the phosphatase.3 It has been widely established that the PTP activity of SHP-2 mediated positive signaling in the Raf/MEK/ERK pathway.28–31 Our demonstration that overexpression of SHP-2 significantly up-regulated ERK1 activity in Huh7/SIRPα1 cells indicated that SHP-2 was the necessary component of the growth factor–mediated ERK activation loop, and SIRPα1 might negatively regulate mitogenic signaling by sequestration of this tyrosine phosphatase from growth factor receptors.1, 32 This idea is supported by the higher affinity of SHP-2 for tyrosine-phosphorylated SIRPα1 rather than for autophosphorylated insulin and EGF receptors and p85 (the regulatory subunit of PI3K).1, 8, 33
This study confirms the negative regulatory function of SIRPα1 on cell proliferation and transformation,1 but goes further in showing that expression of SIRPα1 may also be required for TNF-α– or chemotherapeutic agent-induced apoptosis in HCC cells. Previous studies suggested that activation of 2 signaling pathways, JNK and NF-κB, played a role in protecting Huh7 cells from apoptosis induced by TNF-α or a DNA-damaging agent.22, 34–37 Consistent with this, the present study suggests that the proapoptotic function of SIRPα1 is, at least in part, due to its inhibitory effect on NF-κB. Cytoprotection by NF-κB involves the activation of prosurvival genes, including Bcl-xL, that were previously shown to be up-regulated in HCC tissues, rendering HCC cells resistant to stress-induced apoptosis. Furthermore, down-modulation of Bcl-xL might be sufficient for the induction of apoptosis in response to cellular stress.38 In agreement with its antiapoptotic role in HCC, reduced Bcl-xL expression was observed in Huh7/SIRPα1 cells, compared to Huh7/pcDNA3 and Huh7/SIRPα1-4Y cells, accompanied by early induction of caspase-3 activity; this might explain their increased susceptibility to apoptosis. This observation was consistent with another report that overexpression of SIRPα1 in glioblastoma cells resulted in reduced levels of Bcl-xL protein.8
During the course of the present study, ectopic expression of the cytoplasmic region of SHPS-1/SIRPα1 was shown to increase NF-κB–dependent transcription, presumably through promoting Akt phosphorylation; however, full-length SIRPα1 was shown to have an effect opposite to the truncated one.9 The functional connection of SIRPα1 with the PI3K/Akt pathway was also previously reported.8 In our study, however, expression of SIRPα1 failed to inhibit TNF-α– or serum-induced Akt activation, suggesting that separate pathways might exist for SIRPα1-mediated NF-κB suppression. Recently, it has been shown that SHP-2 is critical for linking EGF receptor to NF-κB transcriptional activity.25 Furthermore, cells from mice lacking SHP-2 are unable to activate NF-κB in response to TNF and interleukin-1,14 establishing SHP-2 as a key component of NF-κB signaling. Our observation that SHP-2 overexpression in Huh7/SIRPα1 cells rescued NF-κB activation and led to increased viability indicates that SIRPα1 may function as an adapter molecule in titrating SHP-2 signalosome, thereby leading to impaired NF-κB activation. The mechanistic insight to this possibility is an area of ongoing study. Taken together, these results demonstrate that SHP-2 likely acts as the molecular transducer of the ERK repression and NF-κB suppression induced by wild type SIRPα1.
SIRPα1 also has a notable effect on cell migration and spreading. Huh7/SIRPα1 cells displayed significantly reduced ability to spread over fibronectin. Impaired cell migration on fibronectin was observed in Huh7/SIRPα1 cells but not in Huh7/SIRPα1-4Y cells, indicating that SIRPα1 functions to impair cell migration. Likewise, reduced cell spreading and migration was also noticed in glioblastoma U87MG cells overexpressing SIRPα1 by association between SIRPα1 and SHP-2.8 This phenotype is quite similar to SHP-2–deficient cells or cells overexpressing a catalytically inactive mutant SHP-2, which exhibited impaired ability in cell spreading and migration, and it would suggest that SIRPα1 might work in concert with SHP-2 in the control of cell motility.18, 39 Given that Ras-MAPK activity appears to be required for cell migration on the extracellular matrix,40, 41 the impaired motility of cells expressing SIRPα1 might result from the inhibition of adhesion-induced ERK activation. However, recent data demonstrated that expression of wild type SHPS-1 promoted the migration of CHO-IR cells in response to insulin, revealing increased biochemical complexity in the regulation of cell motility by SIRPα1.10 The possible explanation for this discrepancy might be the species difference between rat SHPS-1 and human SIRPα1 or the different cell line used. However, the mechanism by which SIRPα1 regulates cell motility remains unclear and warrants further investigation.
In summary, the data presented in this study substantiate the importance of SIRPα1 in the negative regulation of cell proliferation, survival, and migration in HCC cells. The heightened sensitivity of cells restoring SIRPα1 function could be exploited in the development of therapeutic regimens that may potentiate the antineoplastic effect of conventional cytokines or chemotherapeutic agents.
The authors thank Dr. A. Ulrich, Dr. Osamu Tetsu, Dr. G. S. Feng, Dr. C. Scheidereit, and Dr. Ya Cao for invaluable reagents, and Liang Tang for his technical assistance.
- 13Human signal-regulatory protein is expressed on normal, but not on subsets of leukemic myeloid cells and mediates cellular adhesion involving its counterreceptor CD47. Blood 1999; 11: 3633–3643., , , , , , et al.
- 30Role of SH-PTP2, a protein-tyrosine phosphatase with Src homology 2 domains, in insulin-stimulated Ras activation. Mol Cell Biol 1994; 4: 6674–6682., , , , .