The STAT3 inhibitor NSC 74859 is effective in hepatocellular cancers with disrupted TGF-β signaling

doi:10.1038/onc.2008.448

Oncogene. Author manuscript; available in PMC 2009 June 29.

Published in final edited form as:

Oncogene. 2009 February 19; 28(7): 961–972.

Published online 2009 January 12. doi: 10.1038/onc.2008.448

PMCID: PMC2703464

NIHMSID: NIHMS115563

The STAT3 inhibitor NSC 74859 is effective in hepatocellular cancers with disrupted TGF-β signaling

L Lin,^1,^2,⁷ R Amin,^1,⁷ GI Gallicano,³ E Glasgow,^1,² W Jogunoori,¹ JM Jessup,⁴ M Zasloff,¹ JL Marshall,² K Shetty,^1,⁵ L Johnson,^1,⁵ L Mishra,^1,⁶ and AR He^1,²

¹ Cancer Genetics, Digestive Diseases, and Developmental Molecular Biology, Department of Surgery, Georgetown University, Washington, DC, USA

² Department of Medicine and Oncology, Lombardi Comprehensive Cancer Center, Georgetown University, Washington, DC, USA

³ Department of Biochemistry and Molecular and Cellular Biology, Georgetown University Medical Center, Washington, DC, USA

⁴ Cancer Diagnosis Program, National Cancer Institute, Rockville, MD, USA

⁵ Department of Surgery, Institute of Transplant Surgery, Georgetown University, Washington, DC, USA

⁶ Department of Veterans Affairs Medical Center, Washington, DC, USA

Correspondence: Dr AR He, Department of Medicine and Oncology, Lombardi Comprehensive Cancer Center, Georgetown University, 3800 Reservoir Road, NW, Washington, DC 20007, USA. E-mail: aiwu.r.he/at/gunet.georgetown.edu or Dr L Mishra, Cancer Genetics, Department of Surgery, Lombardi Comprehensive Cancer Center, Georgetown University, 3900 Reservoir Road, NW, Medical/Dental Building NW212, Washington, DC 20007, USA. E-mail: lm229/at/georgetown.edu

⁷Author contributions: LL and RA contributed equally to this work. ARH, LM, GIG, LJ, KS, RA, and LL designed research; ARH, LL, and RA performed research; GIG, LL, RA, and JJ contributed new reagents/analytical tools; ARH, LL, GIG, JM, KS, MZ, LJ, and LM analyzed data; ARH, EG, LL, RA, GIG, JM, KS, MZ, LJ, and LM wrote the paper.

The publisher's final edited version of this article is available at Oncogene

See other articles in PMC that cite the published article.

Abstract

Hepatocellular carcinoma (HCC) is the third leading cause of cancer deaths worldwide, with few effective therapeutic options for advanced disease. At least 40% of HCCs are clonal, potentially arising from STAT3 +, NANOG + and OCT3/4 + liver progenitor/stem cell transformation, along with inactivation of transforming growth factor-beta (TGF-β) signaling. Here we report significantly greater signal transducer and activator of transcription 3 (STAT3) and tyrosine phosphorylated STAT3 in human HCC tissues (P<0.0030 and P<0.0455, respectively) than in human normal liver. Further, in HCC cells with loss of response to TGF-β, NSC 74859, a STAT3-specific inhibitor, markedly suppresses growth. In contrast, CD133⁺ status did not affect the response to STAT3 inhibition: both CD133⁺ Huh-7 cells and CD133⁻ Huh-7 cells are equally sensitive to NSC 74859 treatment and STAT3 inhibition, with an IC₅₀ of 100 μM. Thus, the TGF-β/beta2 spectrin (β2SP) pathway may reflect a more functional ‘stem/progenitor’ state than CD133. Furthermore, NSC 74859 treatment of Huh-7 xenografts in nude mice significantly retarded tumor growth, with an effective dose of only 5 mg/kg. Moreover, NSC 74859 inhibited tyrosine phosphorylation of STAT3 in HCC cells in vivo. We conclude that inhibiting interleukin 6 (IL6)/STAT3 in HCCs with inactivation of the TGF-β/β2SP pathway is an effective approach in management of HCCs. Thus, IL6/STAT3, a major signaling pathway in HCC stem cell renewal and proliferation, can provide a novel approach to the treatment of specific HCCs.

Keywords: IL6, stem cell, TGF-β, STAT3, β2SP, hepatocellular carcinoma

Introduction

Hepatocellular carcinoma (HCC) is the fifth most common cancer and third most frequent cause of cancer deaths worldwide, with 600 000 new cases diagnosed every year. Seventy per cent of HCC cases are found to be ineligible for potentially curative surgical therapy because of the disease reaching an advanced stage at the time of diagnosis. At present chemotherapy is for the most part ineffective and patients often have significant liver dysfunction; the median survival is from 6 to 16 months (Kew et al., 1971; Investigators TCotLIPC, 1998). One potential mechanism of HCC resistance to chemotherapy may lie in the plasticity of the cell of origin, which is often a dysfunctional progenitor or stem cell. Up to 40% of HCCs are clonal and thus are considered to originate from progenitor/stem cells (Alison, 2005; Roskams, 2006; Zender et al., 2006; Tang et al., 2008). In addition, several signaling pathways, such as signal transducer and activator of transcription 3 (STAT3), NOTCH, hedgehog and transforming growth factor-beta (TGF-β), which are involved in stem cell renewal, differentiation and survival, are commonly deregulated in HCC (Dando et al., 2005; Sicklick et al., 2006a, b; Kitisin et al., 2007; Nguyen et al., 2007; Wurmbach et al., 2007; Yeoh et al., 2007). We and others have identified TGF-β pathway inactivation in HCCs that have a stem cell phenotype (Tang et al., 1998; Im et al., 2001; Kanzler et al., 2001; Kitisin et al., 2007).

Transforming growth factor-beta pathway proteins are critical regulators of neuronal, hematopoietic, mesenchymal and epithelial cell lineages, as well as suppressors of carcinogenesis (Massague et al., 2000). The TGF-β signaling pathway is activated upon ligand binding to the type I and II transmembrane receptor serine–threonine kinases; TGF-β receptor I (TGFBR1) and TGF-β receptor II (TGFBR2), respectively. Activated TGFBR1 subsequently phosphorylates SMAD transcription factors, such as receptor-regulated SMAD3, resulting in heterodimerization with the common mediator SMAD4, SMAD3/SMAD4 nuclear translocation and activation (or repression) of TGF-β target genes. The non-pleckstrin homology (PH) domain β-general-spectrin, β2SP (also known as Embryonic Liver Fodrin isoform, ELF or Spectrin β, non-erythrocytic 1 isoform 2), is crucial for SMAD regulation and cell specificity. β2SP (Embryonic liver fodrin) associates with SMAD3, presenting it to the cytoplasmic domain of the TGF-β receptor complex. Further, β2SP facilitates nuclear translocation and target gene activation by SMAD3 (Tang et al., 2003). Dysfunction of TGF-β pathway members, such as TGFBR2, SMAD3, SMAD4 and β2SP, may lead to progenitor/stem cell deregulation, and cancer could ensue.

Recently, we identified putative liver progenitor or stem cells from human living-donor liver transplant specimens. These normal human progenitor/stem cells expressed the stem cell markers STAT3, OCT4 (POU5F1) and NANOG, as well as the TGF-β signaling proteins TGFBR2 and β2SP (Kitisin et al., 2007). In contrast, we also showed that, in human HCC tissues, STAT3/OCT4-positive cells are strikingly negative for TGFBR2 and β2SP. Therefore, STAT3/OCT4-positive HCC cells, which have dysfunctional TGF-β signaling, are likely cancer progenitor cells that have the potential to give rise to HCCs, similar to the HCCs that develop in the Elf mutant mice with loss of response to TGF-β (Kitisin et al., 2007). Using mouse genetics, we were able to modify STAT3 signaling in the β2SP (Elf) mutant mice and markedly reduce HCC formation.

The STAT3 inhibitor, NSC 74859 (also known as S3I-201), was discovered through virtual screening as an inhibitor of the Src homology 2 (SH2) dimerization domain of STAT3. When NSC74859 inhibits STAT3 dimerization, it inhibits target gene activation (Siddiquee et al., 2007). In this study, we show that NSC 74859 suppresses growth of HCC cells in culture and in xenograft assays. Seven HCC cell lines were used in these studies. HepG2 cells were used as controls as TGF-β signaling is intact in this cell line (Cabibbo et al., 1998). We show that two cell lines in addition to HepG2 cells, PLC/PRF/5 and SUN-449, have intact TGF-β signaling and that four cell lines, Huh-7, SNU-398, SNU-475 and SNU-182, have defective TGF-β signaling. Although the levels of STAT3 and p^Y705STAT3 are similar among the seven cell lines, the levels of p^S727STAT3 are lowered by a factor of two and four, respectively, in cells with impaired TGF-β signaling compared with the cell lines with intact TGF-β signaling. In addition, the cells with impaired TGF-β signaling are four times as sensitive to the STAT3 inhibitor NSC 74859. Our data suggest that TGF-β signals influence serine phosphorylation of STAT3. Thus, HCC cells with dysfunctional TGF-β signaling may be particularly sensitive to STAT3 inhibition.

To determine whether CD133 represented a further functional marker for HCC ‘cancer stem cells (CSCs)’, we assessed the response of HCC cells to STAT3 inhibition using antibodies to the stem cell surface marker CD133. CD133 is highly upregulated during early and post liver regeneration (Suetsugu et al., 2006; Ma et al., 2007; Yin et al., 2007). CD133⁺ Huh-7 cells are also twice as resistant to doxorubicin compared with CD133⁻ cells (Ma et al., 2008). Surprisingly, we observed that CD133 status did not affect the response to treatment. Here, we show that Huh-7 cells do not express β2SP or TBGFR2 and are sensitive to STAT3 inhibition, with an IC₅₀ of 100 μM for NSC 74859, regardless of CD133⁺ status. Furthermore, low doses (5 mg/kg) of NSC 74859 suppress HCC cell proliferation in xenografts. It is noteworthy that NSC 74859 blocks STAT3 phosphorylation in HCC cells. For these reasons, we propose STAT3 as a promising therapeutic target for the treatment of HCC.

Results

Signal transducer and activator of transcription 3 (STAT3) and phosphorylated STAT3 expression in hepatocellular carcinoma tissue and cell lines

Immunohistochemistry for STAT3 and p^Y705STAT3 (activated STAT3) was performed on nine HCC tissues and five normal liver tissues to determine whether STAT3 expression is linked to HCC. Signal transducer and activator of transcription 3 and p^Y705STAT3 expression was dramatically increased in HCC tissues when compared with normal liver samples (Figure 1). In HCC tissues, strong STAT3 immunostaining was observed in the cytoplasm, and p^Y705STAT3 immunostaining was observed in the nucleus.

Figure 1

Signal transducer and activator of transcription 3 (STAT3) and p^Y705STAT3 levels are elevated in hepatocellular carcinoma (HCC) tissues. Immunohistochemical detection shows increased expression of STAT3 in human HCC tissues (b and d; arrows) compared (more ...)

Quantification of STAT3 and p^Y705STAT3 levels by western blot was precluded by the expression of these proteins in neutrophil and microphage, which invariably contaminates these samples. Therefore, semiquantitative analyses of the immunohistochemically stained samples were used to determine whether STAT3 and p^Y705STAT3 expression is significantly correlated with HCC. Intense (+ + +) and moderate (+ +) STAT3 immunostaining was observed in eight of nine HCC samples, whereas none of the five normal liver samples were more than weakly stained (Figures 1a–d, and Table 1). Nuclear p^Y705STAT3 immunostaining was observed in all HCC tissues, with intense (+ + +), moderate (+ +) and weak (+) staining in 28.5, 28.5 and 43% of HCC tissues, respectively. In contrast, only two of the five normal liver tissues showed weak nuclear p^Y705STAT3 staining (+), and three were negative (−) for nuclear p^Y705STAT3 (Figures 1e, f, and Table 1). Thus, in HCC tissues, STAT3 expression levels were significantly elevated (P =0.0030) and nuclear p^Y705STAT3 staining was significantly increased (P =0.0455), compared with normal liver samples.

Table 1

Immunohistochemical labeling results of liver tissues

The levels of STAT3 and phosphorylated STAT3 (p^Y705STAT3 and p^S727STAT3) were also examined in seven human HCC cell lines: PLC/PRF/5, Huh-7, SNU-398, SNU-449, SNU-182, SNU-475 and HepG2 cells (Figure 2a). All HCC cell lines showed comparable STAT3 and p^Y705STAT3 expression levels. In contrast, levels of p^S727STAT3 varied among the cell lines. In HepG2, SNU-449 and PLC/PRF/5 cells, p^S727STAT3 levels were two to four times those in Huh-7, SNU-182 and SNU-475 cells (Figures 2a).

Figure 2

(a) Western blots in hepatocellular carcinoma (HCC) cell lines of signal transducer and activator of transcription 3 (STAT3), p^Y705STAT3 and p^S727STAT3 (left), and transforming growth factor-beta (TGF-β) pathway proteins (right). (b) Quantitative (more ...)

Frequent loss of transforming growth factor-beta signaling in human hepatocellular carcinoma cell lines

To evaluate the status of the TGF-β signaling pathway in HCC cell lines, we determined expression levels of five TGF-β pathway proteins, including β2SP, SMAD3, SMAD4, TGFBR1 and TBGFR2 (Figure 2b). Among the seven HCC cell lines, four (Huh-7, SNU-398, SNU-475 and SNU-182) showed reduction or loss of one to two TGF-β pathway proteins when compared with HepG2 cells (Figure 2b). In Huh-7 cells, β2SP was diminished. In SNU-398 cells, there was less β2SP by a factor of 9 and less TGFBR2 by a factor of 13. SNU-475 cells showed less TGFBR2 and TGFBR1 by factors of 13 and 19, respectively, and SNU-182 cells lacked SMAD4 expression. On the other hand, two HCC cell lines (SNU-449 and PLC/PRF/5) showed TGF-β pathway protein levels comparable to those of HepG2 cells, which are known to have normal TGF-β signaling (Figure 2b). It is to be noted that Huh7, SNU-398, SNU-182 and SNU-475 cells, which had reduced levels of TGF-β pathway proteins, also had reduced p^S727STAT3 levels (Figure 2a).

We further tested the status of TGF-β signaling in HCC cells by examining growth inhibition by TGF-β treatment using the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) assay for cell proliferation. Transforming growth factor-beta treatment inhibited proliferation of HepG2, Huh-7 and SNU-449 cells, whereas SNU-398 cells were resistant to TGF-β treatment up to 8 ng/ml, the highest dose tested (Figure 2c). From the TGF-β pathway protein levels and MTT proliferation assays, we conclude that HepG2 and SNU-449 cells have intact TGF-β signaling, that SNU-398 cells have impaired TGF-β signaling and that Huh-7 cells are sensitive to TGF-β, but that the TGF-β pathway is altered because of low levels of β2SP in these cells.

Inhibition of hepatocellular carcinoma cell proliferation with the STAT3 inhibitor NSC 74859

To evaluate the effect of STAT3 inhibition on HCC cell proliferation, HCC cell lines were treated with NSC 74859 at increasing concentrations (Figure 3). NSC 74859 inhibited HCC cell proliferation, with the most potent effects observed in those lines with decreased p^S727STAT3 levels and reduced levels of TGF-β pathway proteins (Figures 3a–d). The IC₅₀ of NSC 74859 was 150 μM for Huh-7 and SNU-398 cells, 15 μM for SNU-475 cells and 200 μM for SNU-182 cells. Conversely, cells with elevated p^S727STAT3 and intact TGF-β pathway proteins (HepG2, PLC/PRF/5 and SNU-449) were less sensitive to NSC 74859 treatment, and the IC₅₀ was not reached at 250 μM, the highest dose tested (Figures 3e–g). At 250 μM NSC 74859, proliferation of HCC cells with low p^S727STAT3 levels and aberrant TGF-β pathway proteins decreased by 80%. In contrast, in NSC 74859-treated HCC cells with high p^S727STAT3 levels and intact TGF-β pathway proteins, proliferation decreased by only 20% (Figure 3h).

Figure 3

Viable cells from 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) assay (measured as absorbance at 590 nm with an enzyme-linked immunosorbent assay reader) to show the effects of the signal transducer and activator of transcription (more ...)

Isolation and inhibition of liver cancer stem cells

A progenitor/stem cell side population was isolated from the Huh-7 cell line by CD133 selection using MACS MicroBeads Cell Separation Technology (Ma et al., 2007). As few as five of these CD133⁺ cells could form a colony on soft agar, whereas at least 100 unsorted Huh-7 cells were required to form one colony. Therefore, CD133⁺ cells are enriched in progenitor/stem cells (Figure 4d). In addition, CD133⁺ Huh-7 cells expressed a level of the stem cell marker CD44 that was four times those of CD133⁻ cells, and diminished levels of TGFBR2 (Figure 4b).

Figure 4

(a) 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide assay for NSC 74859 inhibition of cell proliferation in CD133⁺, CD133⁻ cells and unsorted Huh-7 cells. (b) Western blot of CD133, tumor growth factor-β receptor I (TGFBR2), (more ...)

We determined the sensitivity of CD133⁺ and CD133⁻ cell populations to growth inhibition by NSC 74859 using the MTT cell proliferation assay. CD133⁺ cells were as sensitive as CD133⁻ cells to STAT3 inhibition by NSC 74859. Both CD133⁺ and CD133⁻ Huh-7 cells were inhibited by NSC 74859, and with comparable IC₅₀ concentrations of 100 μM (Figure 4a). These results are significant as cancer progenitor/stem cells are thought to be more resistant to chemotherapy in general, and yet a hitherto well-described CSC marker, CD133⁺, did not reflect or predict the response of these HCC cells to treatment, whereas the TGF-β pathway inactivation was a positive predictive marker. Potentially, loss of the tumor suppressor TGF-β pathway indicates a key functional aspect of such CSCs and their response to therapeutics that target stem cells such as STAT3 inhibitors.

Inhibition of signal transducer and activator of transcription 3 with NSC 74859 results in HCC tumor regression

As an in vivo test of the STAT3 inhibitor, NSC 74859, on late-stage tumors, we generated HCC tumor xenografts by injecting Huh-7 cells into the rear hindquarters of nude mice. Once the tumor size reached 0.176±0.076 cm³ for the control group and 0.164± 0.065 cm³ for the treatment group (no significant size difference between tumors at day zero), NSC 74859 or vehicle alone was injected intraperitoneally at 5 mg/kg, and tumor measurements were taken every 2–3 days after drug injection. As early as 6 days after injection, tumors from treated mice were significantly smaller (P<0.01) than tumors from untreated mice (0.171± 0.038 and 0.43±0.199 cm³, respectively; Figure 5). At 21 days after the start of NSC 74859 treatments, tumors from the treated mice were significantly smaller (P<0.01) than tumors from untreated mice (0.594± 0.076 and 1.246±0.434 cm³, respectively; Figure 5a). In fact, tumor growth was significantly retarded for the duration of the experiment (Figure 5a). Importantly, all treated mice tolerated NSC 74859 well, showing no apparent signs of ill health. There is no difference in mouse weight when mice treated with NSC 74859 were compared with those that were untreated (Figure 5b). Expression of p^Y705STAT3 in the xenograft tumors was determined by immunohistochemistry. Strong nuclear p^Y705STAT3 staining of cells in untreated tumors was observed (Figures 5c and e, quantified in Figure 5g from >1200 cells). In contrast, the cells within NSC74859-treated tumors had few p^Y705STAT3-positive tumor cells (Figures 5d and f, quantified in Figure 5g from >1200 cells).

Figure 5

Inhibiting signal transducer and activator of transcription 3 (STAT3) retards tumor growth. p^Y705STAT3 expression was analysed by immunohistochemistry in untreated and NSC 74859-treated tumors. Panel a shows the sizes of the tumors over time in days after (more ...)

Discussion

A major challenge in the systemic treatment of HCC is cellular resistance to conventional cytotoxic agents, which may be attributed to heterogeneity of genetic abnormalities acquired during the course of hepatocarcinogenesis, and/or chemoresistance of liver CSCs. Cancer cells that have stem-like characteristics and are capable of tumor initiation show resistance to chemo-and radiotherapy (Frank et al., 2005; Bao et al., 2006; Hambardzumyan et al., 2006). During cancer therapy, a small number of cells with stem cell-like features may escape from being killed by established anticancer drugs. Therefore, targeting signaling pathways critical for the proliferation and survival of CSCs could present a powerful therapeutic strategy. We reported earlier that mice spontaneously develop HCC with disruption of TGF-β signaling (Kitisin et al., 2007). Several TGF-β signaling components are bona fide tumor suppressors with the ability to constrain cell growth and inhibit cancer development at its early stages. Inactivation of at least one of these components, such as TGFBR2, SMAD2 or the common mediator SMAD4, occurs in almost all gastrointestinal tumors (Massague et al., 2000; Weinstein et al., 2000; Wallner et al., 2006). Given the important role of TGF-β signaling in liver development, as well as in suppression of hepatocarcinogenesis, searching for signaling pathways that interact with TGF-β signaling may reveal mechanisms of CSC self-renewal, differentiation and apoptosis. In our earlier gene array analysis, disruption of TGF-β signaling resulted in upregulation of interleukin 6 (IL6)/STAT3 activity. Therefore, STAT3 is an attractive target for therapeutics developed against CSCs.

Owing to their low abundance (1 in 40–100 000), normal (non-embryonic) stem cells are difficult to isolate. Although the mechanism of CSC development remains in question, cells that have an increased ability to initiate tumors and to express stem cell markers have been isolated. Side-population analysis and cell sorting with stem cell surface markers CD133, CD44 and CD24 have been used for enriching CSC populations, which allowed phenotypic studies of these progenitor/stem-like cells. CD133⁺ cells in glioblastoma have been found to be resistant to etoposide, paclitaxel, temozolomide and carboplatin (Liu et al., 2006), and they express high levels of markers of neural precursors CD90, CD44, CXCR4, NESTIN, MSII and MELK compared with their CD133 counterparts. We find that HCC cells independent of CD133⁺ status, but with loss of β2SP, are sensitive to NSC 74859, reflecting that β2SP status may reflect a more functional CSC phenotype than CD133 in HCC. Therefore, STAT3 may be an effective target for disrupting HCC progenitor/stem cells with inactivation of the TGF-β/β2SP pathway irrespective of CD133. Genetic knock-in models show that CD133 is in fact expressed on a multitude of differentiated epithelial cells in adult mouse tissues and on spontaneous primary colon tumors in mice. In primary human colon tumors, all of the epithelial cells also expressed CD133, and, surprisingly, CD133⁺ and CD133⁻ populations were equally capable of tumor initiation in xenografts. In light of our findings that are supported by others, the role of CD133 as a marker of liver and gastrointestinal ‘cancer stem cells’ may need to be revised (LaBarge and Bissell, 2008; Shmelkov et al., 2008).

Microarray and proteomic analyses of human and mouse HCC tissues with aberrant TGF-β signaling showed increased expression of the IL6/STAT3, wingless-type MMTV intergration site family (WNT) and cyclin-dependent kinase 4 signaling pathways (Kitisin et al., 2007). Here, we focus on the potential role of activated STAT3 in HCC. Signal transducer and activator of transcription 3 is activated by tyrosine phosphorylation at Tyr⁷⁰⁵ (p^Y705STAT3) in response to cytokines and growth factors (Aaronson and Horvath, 2002). Phosphorylation of Tyr⁷⁰⁵ is required for STAT3 activity, whereas Ser⁷²⁷ phosphorylation (p^S727STAT3) positively regulates transcriptional activity but negatively affects its DNA-binding activity (Wen et al., 1995; Chung et al., 1997; Lim and Cao, 1999, 2001). Importantly, STAT3 is a stem cell renewal factor, and hyperactive STAT3 signaling results in enhanced liver progenitor cell proliferation (Yeoh et al., 2007). In addition, overexpression of a constitutively active form of STAT3 in immortalized rat or mouse fibroblasts induced tumors in nude mice (Bromberg et al., 1999). Owing to its role in modulating stem cell survival, proliferation and transformation, STAT3 is thought to be critical for CSC survival in some tissues (Zhou et al., 2007). In this study, we found elevated STAT3 and p^Y705STAT3 expression in HCC tissue, which is consistent with recent data showing an association of pSTAT3 with the histological grade of HCC tissue from 67 patients (Yang et al., 2007). Hence, disruption of TGF-β signaling and activation of STAT3 are important molecular events in the transformation of normal liver stem cells to cancer progenitor/stem cells. We thus propose STAT3 as a promising therapeutic target for HCC.

In this study, we show upregulation of STAT3 and p^Y705STAT3 levels in HCC. In HCC cell lines, we show equal amounts of STAT3 and p^Y705STAT3. However, there is less p^S727STAT3 in cells with lower levels of the TGF-β pathway proteins TGFBR2 and/or β2SP. These data suggest a reciprocal relation between IL6/STAT3 and TGF-β signaling in tumorigenic transformation, although a direct interaction between these two pathways is yet to be defined. Signal transducer and activator of transcription 3 inhibition suppresses the proliferation of HCC cells, with the most potent effect observed in cells with a dysfunctional TGF-β pathway, which also correlates with decreased p^S727STAT3 levels in these cells. Regulation of STAT3 serine phosphorylation by TGF-β through the TGF-β-activated kinase 1-Nemolike kinase (TAK1-NLK) cascade has been shown in Xenopus mesoderm induction (Ohkawara et al., 2004). In addition, STAT3 interacts with TAK1 and Nemolike kinase as a scaffold, and this interaction leads to STAT3 phosphorylation at serine 727 and activation of Nemolike kinase (Kojima et al., 2005). These studies suggest a mechanism by which TGF-β regulates serine phosphorylation of STAT3 and provide some thoughts for the sensitivity of HCC progenitor/stem cells with dysfunctional TGF-β signaling to the STAT3 inhibitor NSC 74859.

It is interesting to note that NSC 74859 is far more potent in vivo than in vitro, with an IC₅₀ of 100 μM in vitro, but an effective dosage of only 5 mg/kg in vivo. The same differences between in vivo and in vitro effective doses were observed when NSC 74859 was used to treat breast cancer (Siddiquee et al., 2007). NSC 74859 inhibits breast carcinoma MDA-MB-435, MDA-MB-453 and MDA-MB-231 cell lines with an IC₅₀ close to 100 μM in vitro, but in vivo it effectively retards the growth of MDA-MB-231 cells at 5 mg/kg. One potential reason could be that NSC 74859 inhibits stroma and endothelial compartments in addition to the tumor cells, giving stronger in vivo and thus potentially realistic inhibition.

Recently, the US Food and Drug Administration approved sorafenib for the treatment of advanced HCC; it improves median overall survival by 3 months compared with placebo. As the first systemic agent to show survival benefits in patients with HCC, the clinical benefits remain modest. There is still an urgent need for therapies for this highly lethal disease. Modulation of STAT3 signaling in cancer progenitor cells may provide an important approach for new therapeutics in cancers with a poor prognosis such as HCC. Another feature offered by STAT3 inhibition for cancer therapy is its specificity—that it is only one of two possible molecular targets (STAT3 and STAT5) as opposed to a multitude of tyrosine kinases that could potentially serve a similar function, thus reducing toxicities. Indeed, β2SP expression may represent a good predictive marker for response of HCC to new therapeutics targeting CSCs such as STAT3 inhibitors. Activities against tumor vasculature and good tolerability in animals serve as additional advantages of the STAT3 inhibitor. Indeed, our tests using the STAT3-specific drug NSC 74859 might be the type of approach needed for improving advanced HCC outcome, particularly in hepatocellular carcinomas with loss of TGF-β signaling.

Materials and methods

Cancer specimens

Formalin-fixed and paraffin-embedded HCC and normal liver specimens were obtained from the Georgetown University Division of Transplant Surgery, Washington, DC, USA. In all, nine HCC samples were collected from patients with varying grades and stages of cancer. Two independent blinded pathologists evaluated the tumors used in this study. The control samples of normal liver tissue used in the present investigation were taken from the borders of the surgical specimens. All specimens were collected with informed patient consent and according to the protocol, and with the approval of the Institutional Review Board #2005-206, Georgetown University, Washington, DC, USA.

Immunohistochemistry

Immunohistochemical labeling of tissues was performed as described earlier (Tang et al., 2003). Normal human liver and HCC tissues were fixed with 4% paraformaldehyde, dehydrated, embedded in paraffin and sectioned at 6 μm. Sections were immunohistochemically stained using primary antibodies against STAT3 and p^Y705STAT3 (Cell Signaling Technology, Danvers, MA, USA).

The frequency of STAT3- and p^Y705STAT3-positive cells was determined by counting the total number of cells and total positively stained cells in randomly selected 40 × magnification fields, including at least 1000 cells. Average numbers from the field sets were then determined and reported as the percentage of positively stained cells to the total numbers of cells. Signal transducer and activator of transcription 3 and pSTAT3 labeling was measured in three different grades: + + +, intense labeling (>70%); + +, moderate labeling (16–69%); +, weak labeling (0–15%); and −, no labeling (0%).

Cell cultures

The HCC cell lines HepG2, PLC/PRF/5, Huh-7, SNU-398, SNU-449, SNU-182 and SNU-475 were obtained from American Type Culture Collection, Manassas, VA, USA. HepG2, PLC/PRF/5 and Huh-7 cell lines were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM). SNU-398, SNU-449, SNU-182 and SNU-475 were maintained in RPMI-1640. Both types of medium were supplemented with 10% fetal bovine serum.

Immunoblot analysis (western blot)

Cell lines were grown in a monolayer up to 70% confluence before harvesting for western blot analysis as described earlier (Tang et al., 2003). For western blots on sorted cells, cells were separated into CD133⁺ and CD133⁻ fractions by MACS MicroBeads Separation System (Miltenyi Biotec, Auburn, CA, USA) by using CD133 antibodies. Cells were lysed and denatured at 95 °C for 5 min in sample buffer (100 mM Tris, pH 6.8, 20% glycerol, 0.01% bromophenolblue, 10% β-mercaptoethanol and 5% SDS). Equal amounts of protein were separated on an SDS-polyacrylamide gel and transferred onto a nitrocellulose membrane (Invitrogen, Carlsbad, CA, USA). Membranes were blocked in 5% milk solution overnight and incubated with primary antibodies for STAT3 (Cell Signaling Technology, Danvers, MA, USA), phospho-STAT3 (Ser⁷²⁷) (Chemicon International, Temecula, CA, USA), phospho-STAT3 (Tyr⁷⁰⁵) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), β2SP, TGFBR1 (Santa Cruz Biotechnology, Inc.), TGFBR2 (Santa Cruz Biotechnology, Inc.), SMAD3 (Santa Cruz Biotechnology, Inc.), SMAD4 (Biocare Medical, Concord, CA, USA), CD133 (Santa Cruz Biotechnology, Inc.), CD44 (Santa Cruz Biotechnology, Inc.) and β-actin (Sigma, St Louis, MO, USA), followed by incubation with horseradish peroxidase-conjugated secondary antibodies (GE Healthcare, Milwaukee, WI, USA). Signals were visualized by enhanced chemiluminescence plus western blotting system (GE Healthcare).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay

The MTT assay is based on the conversion of the yellow tetrazolium salt MTT to purple formazan crystals by metabolically active cells (Vistica et al., 1991). The MTT assay provides a quantitative determination of viable cells. Cells (5 ×10³) were seeded in 96-well microplates in complete culture medium in the absence or presence of increasing serial dosages of NSC 74859 as indicated. At 72 h after culture, the number of viable cells was measured by adding 100 μl/well of 2 mg/ml MTT solution. After 2 h, the medium was removed, and the formazan crystals were dissolved by adding 100 μl dimethylsulfoxide per well. The absorbance was read at 590 nm with an enzyme-linked immunosorbent assay reader. Each treatment point was performed in 10 wells or sextuplicate (Sigma-Aldrich #M2128).

Magnetic sorting of CD133⁺ hepatocellular carcinoma cells

Sorting of CD133⁺ HCC cells was carried out as described earlier (Ma et al., 2007) and as per the manufacturer’s instructions. Cells were labeled with primary CD133/1 antibody (CD133 mouse IgG1, 10 μl per 10 million cells, conjugated with phycoerythrin, Miltenyi Biotec #130-080-801), and then magnetically labeled with anti-phycoerythrin multisort microbeads (rat anti-mouse, 20 μl per 10 million cells) (Miltenyi Biotec, #130-090-757), followed by separation on the MACS LS column (Miltenyi Biotec #130-042-202).

Soft-agar colony-formation assay

Colony formation assays in soft agar were carried out as described (Turkson et al., 2001). In brief, a base layer was made by mixing 1% soft agar with an equivalent 2 × medium and prepared in six-well dishes. CD133-sorted Huh-7 cells were harvested, suspended in medium containing 0.35% soft agar (Invitrogen), and seeded on the base layer at different densities from 5 to 5000 cells per well. All experiments were conducted in triplicate. Plates were maintained at 37 °C in a humidified incubator and were fed every 3 days with 0.1 ml of complete medium. After 2 weeks, the number of colonies formed was counted under the microscope.

Xenograft transplantation and in vivo tumor studies

Six-week-old female athymic nude mice were purchased from Harlan (Indianapolis, IN, USA) and maintained in the institutional animal facilities approved by the American Association for Accreditation of Laboratory Animal Care. Five athymic nude mice per group were injected in the left flank area subcutaneously with human HCC Huh-7 cells ranging from 5 ×10² to 5 ×10⁶ in 100 μl of phosphate-buffered saline. After 5–10 days, 10 tumors in the control or treatment group with a diameter of 0.175 mm³ were established. Animals were given NSC 74859 intraperitoneally at 5 mg/kg every other day for 3 weeks and monitored daily. Animals were stratified so that the mean tumor sizes in all treatment groups were nearly identical. Tumor volume was calculated according to the formula V =0.5 ×a² ×b, where a is the smallest superficial diameter and b is the largest superficial diameter. Student’s t-test was used to analyse in vivo growth patterns.

Statistical analysis

Global χ² test was used to test the hypothesis that the coefficient of each variable was equal to 0. Results from the immunohistochemical labeling of STAT3 and pSTAT3 in HCC samples as compared with normal liver samples were used to assess the statistical significance. P<0.05 was considered statistically significant, and all tests were two-sided. All tests were performed with SPSS 10.1 software (SPSS Inc., Chicago, IL, USA).

Acknowledgments

We wish to thank Dr Said M Sebti for helpful discussions on the use of NSC 74859; Dr Joseph Cory for critical review and helpful suggestions with the manuscript; Tiffany Blake, Ed Flores, E Volpe and Sandra Acheampong for excellent technical expertise and manuscript preparation; Mr Carlos Benitez for xenograft production and injection of drugs; and Dr Richard Amdur for statistical analysis. Grant support was provided by NIH R01 CA106614A (LM), NIH R01 DK56111 (LM), NIH R01 CA4285718A (LM), VA Merit Award (LM) (Bao et al.), R Robert and Sally D Funderburg Research Scholar (LM), B Orr Foundation (LM), and NIH P30 CA51008-13 (transgenic shared resource; Cancer Center Support Grant).

Abbreviations


CSC	cancer stem cell

β2SP	embryonic liver fodrin

HCC	hepatocellular carcinoma

STAT3	signal transduction and activator of transcription

TGFBR1	transforming growth factor-β receptor I

TGFBR2	transforming growth factor-β receptor II

TGF-β	transforming growth factor-beta

References

Aaronson DS, Horvath CM. A road map for those who don’t know JAK-STAT. Science. 2002;296:1653–1655. [PubMed]
Alison MR. Liver stem cells: implications for hepatocarcinogenesis. Stem Cell Rev. 2005;1:253–260. [PubMed]
Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature. 2006;444:756–760. [PubMed]
Bromberg JF, Wrzeszczynska MH, Devgan G, Zhao Y, Pestell RG, Albanese C, et al. Stat3 as an oncogene. Cell. 1999;98:295–303. [PubMed]
Cabibbo A, Consalez GG, Sardella M, Sitia R, Rubartelli A. Changes in gene expression during the growth arrest of HepG2 hepatoma cells induced by reducing agents or TGFbeta1. Oncogene. 1998;16:2935–2943. [PubMed]
Chung J, Uchida E, Grammer TC, Blenis J. STAT3 serine phosphorylation by ERK-dependent and -independent pathways negatively modulates its tyrosine phosphorylation. Mol Cell Biol. 1997;17:6508–6516. [PMC free article] [PubMed]
Dando JS, Tavian M, Catelain C, Poirault S, Bennaceur-Griscelli A, Sainteny F, et al. Notch/Delta4 interaction in human embryonic liver CD34+ CD38− cells: positive influence on BFU-E production and LTC-IC potential maintenance. Stem Cells. 2005;23:550–560. [PubMed]
Frank NY, Margaryan A, Huang Y, Schatton T, Waaga-Gasser AM, Gasser M, et al. ABCB5-mediated doxorubicin transport and chemoresistance in human malignant melanoma. Cancer Res. 2005;65:4320–4333. [PubMed]
Hambardzumyan D, Squatrito M, Holland EC. Radiation resistance and stem-like cells in brain tumors. Cancer Cell. 2006;10:454–456. [PubMed]
Im YH, Kim HT, Kim IY, Factor VM, Hahm KB, Anzano M, et al. Heterozygous mice for the transforming growth factor-beta type II receptor gene have increased susceptibility to hepatocellular carcinogenesis. Cancer Res. 2001;61:6665–6668. [PubMed]
Investigators TCotLIPC. A new prognostic system for hepatocellular carcinoma: a retrospective study of 435 patients: the Cancer of the Liver Italian Program (CLIP) investigators. Hepatology. 1998;28:751–755. [PubMed]
Kanzler S, Meyer E, Lohse AW, Schirmacher P, Henninger J, Galle PR, et al. Hepatocellular expression of a dominant-negative mutant TGF-beta type II receptor accelerates chemically induced hepatocarcinogenesis. Oncogene. 2001;20:5015–5024. [PubMed]
Kew MC, Dos Santos HA, Sherlock S. Diagnosis of primary cancer of the liver. Br Med J. 1971;4:408–411. [PMC free article] [PubMed]
Kitisin K, Ganesan N, Tang Y, Jogunoori W, Volpe EA, Kim SS, et al. Disruption of transforming growth factor-beta signaling through beta-spectrin ELF leads to hepatocellular cancer through cyclin D1 activation. Oncogene. 2007;26:7103–7110. [PubMed]
Kojima H, Sasaki T, Ishitani T, Iemura S, Zhao H, Kaneko S, et al. STAT3 regulates Nemo-like kinase by mediating its interaction with IL-6-stimulated TGFbeta-activated kinase 1 for STAT3 Ser-727 phosphorylation. Proc Natl Acad Sci USA. 2005;102:4524–4529. [PubMed]
Kopf M, Baumann H, Freer G, Freudenberg M, Lamers M, Kishimoto T, et al. Impaired immune and acute-phase responses in interleukin-6-deficient mice. Nature. 1994;368:339–342. [PubMed]
LaBarge MA, Bissell MJ. Is CD133 a marker of metastatic colon cancer stem cells? J Clin Invest. 2008;118:2111–2120. [PMC free article] [PubMed]
Lim CP, Cao X. Serine phosphorylation and negative regulation of Stat3 by JNK. J Biol Chem. 1999;274:31055–31061. [PubMed]
Lim CP, Cao X. Regulation of Stat3 activation by MEK kinase 1. J Biol Chem. 2001;276:21004–21011. [PubMed]
Liu G, Yuan X, Zeng Z, Tunici P, Ng H, Abdulkadir IR, et al. Analysis of gene expression and chemoresistance of CD133+ cancer stem cells in glioblastoma. Mol Cancer. 2006;5:67. [PMC free article] [PubMed]
Ma S, Chan KW, Hu L, Lee TK, Wo JY, Ng IO, et al. Identification and characterization of tumorigenic liver cancer stem/progenitor cells. Gastroenterology. 2007;132:2542–2556. [PubMed]
Ma S, Lee TK, Zheng BJ, Chan KW, Guan XY. CD133+ HCC cancer stem cells confer chemoresistance by preferential expression of the Akt/PKB survival pathway. Oncogene. 2008;27:1749–1758. [PubMed]
Massague J, Blain SW, Lo RS. TGFbeta signaling in growth control, cancer, and heritable disorders. Cell. 2000;103:295–309. [PubMed]
Nguyen LN, Furuya MH, Wolfraim LA, Nguyen AP, Holdren MS, Campbell JS, et al. Transforming growth factor-beta differentially regulates oval cell and hepatocyte proliferation. Hepatology. 2007;45:31–41. [PubMed]
Ohkawara B, Shirakabe K, Hyodo-Miura J, Matsuo R, Ueno N, Matsumoto K, et al. Role of the TAK1-NLK-STAT3 pathway in TGF-beta-mediated mesoderm induction. Genes Dev. 2004;18:381–386. [PubMed]
Roskams T. Liver stem cells and their implication in hepatocellular and cholangiocarcinoma. Oncogene. 2006;25:3818–3822. [PubMed]
Shmelkov SV, Butler JM, Hooper AT, Hormigo A, Kushner J, Milde T, et al. CD133 expression is not restricted to stem cells, and both CD133+ and CD133− metastatic colon cancer cells initiate tumors. J Clin Invest. 2008;118:2111–2120. [PMC free article] [PubMed]
Sicklick JK, Li YX, Jayaraman A, Kannangai R, Qi Y, Vivekanandan P, et al. Dysregulation of the Hedgehog pathway in human hepatocarcinogenesis. Carcinogenesis. 2006a;27:748–757. [PubMed]
Sicklick JK, Li YX, Melhem A, Schmelzer E, Zdanowicz M, Huang J, et al. Hedgehog signaling maintains resident hepatic progenitors throughout life. Am J Physiol Gastrointest Liver Physiol. 2006b;290:G859–G870. [PubMed]
Siddiquee K, Zhang S, Guida WC, Blaskovich MA, Greedy B, Lawrence HR, et al. Selective chemical probe inhibitor of Stat3, identified through structure-based virtual screening, induces antitumor activity. Proc Natl Acad Sci USA. 2007;104:7391–7396. [PubMed]
Suetsugu A, Nagaki M, Aoki H, Motohashi T, Kunisada T, Moriwaki H. Characterization of CD133+ hepatocellular carcinoma cells as cancer stem/progenitor cells. Biochem Biophys Res Commun. 2006;351:820–824. [PubMed]
Tang B, Bottinger EP, Jakowlew SB, Bagnall KM, Mariano J, Anver MR, et al. Transforming growth factor-beta1 is a new form of tumor suppressor with true haploid insufficiency. Nat Med. 1998;4:802–807. [PubMed]
Tang Y, Katuri V, Dillner A, Mishra B, Deng CX, Mishra L. Disruption of transforming growth factor-beta signaling in ELF beta-spectrin-deficient mice. Science. 2003;299:574–577. [PubMed]
Tang Y, Kitisin K, Jogunoori W, Li C, Deng CX, Mueller SC, et al. Progenitor/stem cells give rise to liver cancer due to aberrant TGF-beta and IL-6 signaling. Proc Natl Acad Sci USA. 2008;105:2445–2450. [PubMed]
Turkson J, Ryan D, Kim JS, Zhang Y, Chen Z, Haura E, et al. Phosphotyrosyl peptides block Stat3-mediated DNA binding activity, gene regulation, and cell transformation. J Biol Chem. 2001;276:45443–45455. [PubMed]
Vistica DT, Skehan P, Scudiero D, Monks A, Pittman A, Boyd MR. Tetrazolium-based assays for cellular viability: a critical examination of selected parameters affecting formazan production. Cancer Res. 1991;51:2515–2520. [PubMed]
Wallner L, Dai J, Escara-Wilke J, Zhang J, Yao Z, Lu Y, et al. Inhibition of interleukin-6 with CNTO328, an anti-interleukin-6 monoclonal antibody, inhibits conversion of androgen-dependent prostate cancer to an androgen-independent phenotype in orchiectomized mice. Cancer Res. 2006;66:3087–3095. [PubMed]
Weinstein M, Yang X, Deng C. Functions of mammalian Smad genes as revealed by targeted gene disruption in mice. Cytokine Growth Factor Rev. 2000;11:49–58. [PubMed]
Wen Z, Zhong Z, Darnell JE., Jr Maximal activation of transcription by Stat1 and Stat3 requires both tyrosine and serine phosphorylation. Cell. 1995;82:241–250. [PubMed]
Wurmbach E, Chen YB, Khitrov G, Zhang W, Roayaie S, Schwartz M, et al. Genome-wide molecular profiles of HCV-induced dysplasia and hepatocellular carcinoma. Hepatology. 2007;45:938–947. [PubMed]
Yang SF, Wang SN, Wu CF, Yeh YT, Chai CY, Chunag SC, et al. Altered p-STAT3 (tyr705) expression is associated with histological grading and intratumour microvessel density in hepatocellular carcinoma. J Clin Pathol. 2007;60:642–648. [PMC free article] [PubMed]
Yeoh GC, Ernst M, Rose-John S, Akhurst B, Payne C, Long S, et al. Opposing roles of gp130-mediated STAT-3 and ERK-1/2 signaling in liver progenitor cell migration and proliferation. Hepatology. 2007;45:486–494. [PubMed]
Yin S, Li J, Hu C, Chen X, Yao M, Yan M, et al. CD133 positive hepatocellular carcinoma cells possess high capacity for tumorigenicity. Int J Cancer. 2007;120:1444–1450. [PubMed]
Zender L, Spector MS, Xue W, Flemming P, Cordon-Cardo C, Silke J, et al. Identification and validation of oncogenes in liver cancer using an integrative oncogenomic approach. Cell. 2006;125:1253–1267. [PMC free article] [PubMed]
Zhou J, Wulfkuhle J, Zhang H, Gu P, Yang Y, Deng J, et al. Activation of the PTEN/mTOR/STAT3 pathway in breast cancer stem-like cells is required for viability and maintenance. Proc Natl Acad Sci USA. 2007;104:16158–16163. [PubMed]