Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

CD73 promotes hepatocellular carcinoma progression and metastasis via activating PI3K/AKT signaling by inducing Rap1-mediated membrane localization of P110β and predicts poor prognosis

CD73 promotes hepatocellular carcinoma progression and metastasis via activating PI3K/AKT... Background: Hepatocellular carcinoma (HCC) is one of the most prevalent malignancies worldwide because of rapid progression and high incidence of metastasis or recurrence. Accumulating evidence shows that CD73-expressing tumor cell is implicated in development of several types of cancer. However, the role of CD73 in HCC cell has not been systematically investigated and its underlying mechanism remains elusive. Methods: CD73 expression in HCC cell was determined by RT-PCR, Western blot, and immunohistochemistry staining. Clinical significance of CD73 was evaluated by Cox regression analysis. Cell counting kit-8 and colony formation assays were used for proliferation evaluation. Transwell assays were used for motility evaluations. Co-immunoprecipitation, cytosolic and plasma membrane fractionation separation, and ELISA were applied for evaluating membrane localization of P110β and its catalytic activity. NOD/SCID/γc(null) (NOG) mice model was used to investigate the in vivo functions of CD73. Results: In the present study, we demonstrate that CD73 was crucial for epithelial-mesenchymal transition (EMT), progression and metastasis in HCC. CD73 expression is increased in HCC cells and correlated with aggressive clinicopathological characteristics. Clinically, CD73 is identified as an independent poor prognostic indicator for both time to recurrence and overall survival. CD73 knockdown dramatically inhibits HCC cells proliferation, migration, invasion, and EMT in vitro and hinders tumor growth and metastasis in vivo. Opposite results could be observed when CD73 is overexpressed. Mechanistically, adenosine produced by CD73 binds to adenosine A2A receptor (A2AR) and activates Rap1, which recruits P110β to the plasma membrane and triggers PIP3 production, thereby promoting AKT phosphorylation in HCC cells. Notably, a combination of anti-CD73 and anti-A2AR achieves synergistic depression effects on HCC growth and metastasis than single agent alone. (Continued on next page) * Correspondence: yang.xinrong@zs-hospital.sh.cn; guo.wei@zs-hospital.sh.cn Xiao-Lu Ma, Min-Na Shen and Bo Hu contributed equally to this work. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai 200032, People’s Republic of China Department of Laboratory Medicine, Zhongshan Hospital, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, People’s Republic of China Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 2 of 17 (Continued from previous page) Conclusions: CD73 promotes progression and metastasis through activating PI3K/AKT signaling, indicating a novel prognostic biomarker for HCC. Our data demonstrate the importance of CD73 in HCC in addition to its immunosuppressive functions and revealed that co-targeting CD73 and A2AR strategy may be a promising novel therapeutic strategy for future HCC management. Keywords: Hepatocellular carcinoma, CD73, Epithelial-mesenchymal-transition,Prognosis,PI3K/AKT Introduction vivo. Adenosine produced by CD73 binds to adenosine Hepatocellular carcinoma (HCC) is one of the most A2A receptor (A2AR) and activates Rap1, which recruits prevalent malignancies worldwide and the second most P110β to the plasma membrane and triggers PIP3 pro- common cause of cancer-related death [1, 2]. Although duction, thereby promoting AKT phosphorylation. Im- great improvements in treatment have been made, the portantly, the combination of CD73 and A2AR prognosis of HCC remains unfavorable, with an approxi- inhibitors provided more synergistic tumor inhibition mate 30% overall 5-year survival rate even after curative than either regimen alone in HCC. resection [3–5]. Recurrence or metastasis is the major reason for the poor survival of HCC patients [6, 7], but the precise regulatory mechanisms of invasion and me- Methods and materials tastasis remain elusive. Therefore, identification of crit- Clinical specimens and follow-up ical molecules that contribute to the invasive phenotype Four groups of patients were recruited in the present of HCC and clarification of the underlying mechanism study. In group I, frozen tumor tissues with paired para- are urgently needed to improve HCC prognosis. tumor normal tissues from 25 HCC patients receiving CD73, an AMP hydrolyzing enzyme that regulates the curative resection in October 2010 in Zhongshan conversion of extracellular ATP into adenosine, functions Hospital were collected for comparisons of CD73 ex- as a powerful immunosuppressor for maintaining tissue pression between matched tumor and paratumor tissues homeostasis and preventing immune responses during in- by RT-PCR and Western blot (WB) assays. In group II, flammation [8]. Accumulating evidence demonstrates the eight recurrent HCC patients receiving curative resec- vital role of CD73 in tumor [9, 10]. Hematopoietic CD73 tion from April to November 2014 in Zhongshan Hos- overexpression within the tumor microenvironment is ob- pital were enrolled, and matched frozen paratumor, served in a variety of cancers, and these CD73-expressing primary tumor, and recurrent tumor tissues were col- cells hamper the immune reaction towards cancer cells lected for further WB assays. In group III, 10 HCC pa- and disable the cytotoxic antitumor immune response by tients suffered lung metastasis after curative resection producing high levels of adenosine [11–14]. CD73 from 2011 to 2015 in Zhongshan hospital were enrolled, expressed on hematopoietic cells can also function as a and matched paraffin sections of paratumor, primary co-stimulatory molecule in human T cells to block CD8 tumor, and metastatic tumor tissues were collected for T cell activation [15]. Recent reports also demonstrated further immunohistochemical staining. In group IV, 189 that CD73 overexpression is involved in migration, inva- HCC patients receiving curative resection in Zhongshan sion, and angiogenesis [15–18]. Moreover, clinical data Hospital from March 2012 to September 2013 were en- demonstrated that CD73 is a biomarker of poor prognosis rolled, and paratumor and tumor specimens were col- in solid cancers [19–21]. These investigations suggested lected for tissue microarrays (TMA) establishment. that CD73 could be a critical regulator that promotes Enrollment criteria were according to a previous study tumor progression in an immune-independent manner. A [7]. HCC diagnosis was based on histopathology accord- preliminary study reported that CD73 was overexpressed ing to the American Association for Study of Liver in HCC cells and was positively correlated with EGFR ex- Disease guidelines. The Barcelona Clinic Liver Cancer pression [22]. However, whether CD73 could promote (BCLC) staging system was used to assess tumor stage HCC progression and metastasis and the underlying regu- [23]. Approval for the use of human subjects was ob- latory mechanism still needs to be elucidated. tained from the research ethics committee of Zhongshan Here, we report that CD73 expression is positively cor- Hospital, and informed consent was obtained from each related with metastasis in HCC and is an independent individual. Follow-up ended on December 2016. Time to indicator for predicting prognosis. CD73 overexpression recurrence (TTR) was defined as the interval between enhanced HCC progression and metastasis in the ab- treatment and intrahepatic recurrence or extrahepatic sence of an immunological environment in vitro and in metastasis. Overall survival (OS) was defined as the Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 3 of 17 interval between treatment and death of any cause or multivariate analysis demonstrated high CD73 expres- the last observation date. sion on HCC cell and was an independent indicator for predicting both TTR [HR 2.82 (1.87–4.26), P <0.001] and OS [HR 3.35 (2.01–5.52), P < 0.001, Tables 2 and 3]. Statistical analysis Intriguingly, patients with high CD73 had significantly Statistical analyses were performed using SPSS 20.0 soft- higher pre-treatment circulating tumor cell (CTC) load ware (IBM, Chicago, IL, USA). Experimental values for (P < 0.001, Fig. 1h). Using 0.80 as the cutoff value for continuous variables were expressed as the mean ± stand- CTC load [24], a higher proportion of patients in the ard error of the mean. The chi-squared test, Fisher’sexact high CD73 group had high CTC loads (Fig. 1h). Notably, probability tests, and the Student’s t test were used as ap- CD73 showed elevated expression in MHCC97H and propriate to evaluate the significance of differences in data HCCLM3 cells, which have higher invasive and meta- between groups. If variances within groups were not static potentials (Fig. 1i). In comparison, cell lines with homogeneous, a nonparametric Mann–Whitney test or a lower invasive and metastatic capacities such as Wilcoxon signed-rank test was used. Prognostic value was SMMC7721 and MHCC97L showed relatively low CD73 evaluated by Kaplan–Meier survival curves, log-rank tests, levels. Collectively, our data demonstrated the clinical and Cox proportional hazards models. A P value less than significance of CD73 and suggested it as a potential pro- 0.05 was considered significant. moter for HCC metastasis/recurrence. Further details of materials and methods are described in Additional file 1. CD73 promotes HCC progression in vitro Results CD73 expression in HCC cell lines was manipulated by CD73 is overexpressed in HCC tissues and correlates with ectopic expression or short hairpin RNA (shRNA) poor prognosis knockdown. Two distinct shRNAs were designed to high CD73 expression was markedly higher in 55% of HCC knock down CD73 expression in two CD73 HCC cell tissues than paired adjacent normal liver tissues lines, HCCLM3 and Hep3B cells. Knockdown effects (Additional file 2: Figure S1A). WB assays confirmed the were validated by RT-PCR and WB assays (Fig. 2a, RT-PCR findings (Fig. 1b). Moreover, immunohisto- Additional file 3: Figure S2A). Stable ectopic expression low chemistry (IHC) staining was conducted with 189 HCC of CD73 in a CD73 HCC cell line, SMMC7721, was patients receiving curative resection, and results showed also confirmed by RT-PCR and WB assays. We found KD that HCC tissues expressed higher level of CD73 than CD73 HCCLM3 cells exhibited significantly decreased adjacent non-cancerous liver tissues, according to the proliferation potential compared with controls according criteria of CD73 expression levels (Fig. 1c). These data to CCK-8 and colony formation assays (Fig. 2b and c). revealed the potential oncogenic role of CD73 in HCC. Similar reduced proliferation was observed in Hep3B Next, the correlation between CD73 and HCC progno- cells (Additional file 3: Figure S2B). In contrast, overex- sis was evaluated. Samples from eight recurrent patients pression of CD73 in SMMC7721 cells increased their were collected. WB assays showed that expression levels proliferation in CCK-8 assays and clonogenicity capacity of CD73 in matched non-tumor, primary tumor, and re- (Fig. 2b and c). Cell cycle assays demonstrated that currence lesions exhibited an escalating pattern in most CD73 knockdown resulted in G0/G1 arrest in HCC cells patients (Fig. 1d). For patients with metastasis, the CD73 (Fig. 2d, Additional file 3: Figure S2C), while CD73 over- expression level of metastasis foci was also significantly expression accelerated the cell cycle in SMMC7721 cells elevated according to IHC results (n = 10, Fig. 1e). More- (Fig. 2d). Apoptosis assays further showed that CD73 over, we found patients with high CD73 were prone to knockdown greatly induced apoptosis in HCC cells, show incomplete tumor encapsulation, microvascular in- while CD73 overexpression protected SMMC7721 cells vasion, and poor differentiation (all P < 0.050, Table 1). from a serum-free environment (Fig. 2e, Additional file 3: Median TTR was significantly shorter in patients with Figure S2D). Transwell assays showed that CD73 knock- high CD73 expression (median 25.73 months vs. not down HCC cells had significantly fewer migrating and OE reached, P < 0.001, Fig. 1f), and the recurrence rate was invading cells than parental cells, while CD73 also higher in these patients (70.45% vs. 45.54%, Fig. 1g). SMMC7721 significantly higher migratory and invasive Similarly, patients with high CD73 expression had sig- capacities (Fig. 2f, Additional file 4: Figure S3A and B). nificantly shorter OS (median 38.88 months vs. not Moreover, wound healing assays confirmed the findings reached, P < 0.001) and a higher death rate (57.95% vs. of Transwell assays that knockdown of CD73 greatly 21.78%). Furthermore, CD73 retained its prognostic hindered HCC migration, while overexpression of CD73 prediction value in early-stage and low-AFP subgroups promoted migratory potentials (Fig. 2g, Additional file 4: (all P < 0.050, Additional file 2: Figure S1A–D). Notably, Figure S3C). Importantly, we also conducted CD73 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 4 of 17 Fig. 1 CD73 is overexpressed in HCC and indicates poor prognosis. a RT-PCR analysis of CD73 expression levels in 25 HCC tissues and paired non-tumor liver tissues. b Western blot (WB) analysis of CD73 expression levels in 25 HCC tissues and paired non-tumor liver tissues. n = 25; paired t test was used. c Representative immunohistochemistry (IHC) staining of CD73 expression, and comparison of the IHC score between cancerous and non-cancerous tissues (asterisk indicated P < 0.050, n = 189, non-parametric test was used). d WB analysis of CD73 expression levels in matched paratumor, primary-tumor, and recurrent tumor tissues from the same case (n = 8). e Representative IHC staining of CD73 expression in matched paratumor, primary-tumor, and metastatic tissues from the same case (left); comparisons of the IHC score were shown at the right panel (n = 10; non-parametric test was used). f Kaplan–Meier analysis of TTR (left) and overall survival (right) of HCC patients after curative resection according to CD73 expression level. g Recurrence (left) and death (right) incidences of HCC patients according to CD73 expression level. h Circulating tumor cell (CTC) load level of HCC patients according to CD73 expression level (left), and incidence of high CTC load in patients according to CD73 expression levels (right). (i) Expression level of CD73 in different HCC cell lines overexpression in HCCLM3 cells to confirm that the dif- migration, and invasion capacities in HCCLM3 cells ference observed above was not due to the variability (Additional file 5:Figure S4B–D). However, no significant between cell lines. CD73 expression was greatly in- alteration of apoptosis and cell cycle was observed after creased after plasmid transfection in HCCLM3 cells CD73 overexpression (Additional file 5: Figure S4E and F). (Additional file 5: Figure S4A). Similar to the effects of Taken together, our data indicated that CD73 promotes CD73 overexpression on SMMC7721 cells, forced expres- HCC cell proliferation, migration, and invasion capacities sion of CD73 also significantly promoted proliferation, and prevents apoptosis in vitro. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 5 of 17 Table 1 Correlation between clinicopathological parameters of Table 1 Correlation between clinicopathological parameters of patients enrolled patients enrolled (Continued) Clinical characteristics No. of patients CD73 low CD73 high P Clinical characteristics No. of patients CD73 low CD73 high P (n = 189) (n = 189) Age, years BCLC stage ≤ 50 79 35 44 0.033 0 + A 158 86 72 0.537 > 50 110 66 44 B+C 31 15 16 ALT alanine aminotransferase, AST aspartate transaminase, AFP α-fetoprotein, Sex BCLC Barcelona Clinic Liver Cancer Female 38 16 22 0.117 Bold indicated statistical significance Male 151 85 66 Child–Pugh score CD73 promotes HCC growth and metastasis in vivo A 179 96 83 0.823 Since shCD73-1 exhibited more efficiency in knocking down CD73 expression and resulting in more significant biological B10 5 5 alterations, further in vivo experiments and mechanism in- Liver cirrhosis vestigation were conducted using this shRNA. To minimize No 42 20 22 0.391 the impact of the immune system, NOG mice were used. Yes 147 81 66 Model mice were sacrificed after 6 weeks, and analyses of ALT, U/L liver orthotopic xenograft tumors showed that CD73 knock- ≤ 40 134 71 63 0.845 down greatly inhibited tumor growth, whereas CD73 over- expression promoted tumor growth in vivo (Fig. 2h). >40 55 30 25 Moreover, tumor tissues with high CD73 expression also ex- AST, U/L hibited high proliferating cell nuclear antigen levels (Add- ≤ 40 135 73 62 0.782 itional file 6: Figure S5A). Metastatic foci were examined >40 54 28 26 based on their tissue structure and cell morphology by two AFP, ng/ml senior histopathologists to evaluate the effects of CD73 on ≤ 400 140 74 66 0.786 HCC metastasis. Both the intrahepatic and pulmonary me- KD tastasis rates in mice with tumors generated from CD73 > 400 49 27 22 HCCLM3 cells were lower than in mice with tumors de- No. of tumors rived from parental HCCLM3 cells (Fig. 2i, Additional file 6: Single 169 92 77 0.424 Figure S5B and C). In contrast, both intrahepatic and pul- Multiple 20 9 11 monary metastasis rates were increased in tumors derived OE Tumor size, cm from CD73 SMMC7721 cells compared with parental ≤ 5 119 62 57 0.631 SMMC7721 cells. Collectively, our data demonstrated that CD73 promotes HCC proliferation and metastasis in vivo. > 5 70 39 31 Tumor encapsulation CD73 triggers epithelial-mesenchymal transition in HCC Complete 123 59 64 0.040 We further examined whether CD73 contributes to EMT. None 66 42 24 KD Through phalloidin staining, we found CD73 HCCLM3 Satellite lesion cells showed an epithelial cobblestone phenotype with less OE No 171 88 83 0.093 pseudopod, whereas CD73 SMMC7721 cells trans- formed into a spindle-like shape with more pseudopods Yes 18 13 5 (Fig. 3a). Moreover, CD73 knockdown resulted in an Macrovascular invasion epithelial-like molecular phenotype, while CD73 overex- No 175 94 81 0.784 pression induced a mesenchymal-like molecular phenotype Yes 14 7 7 according to RT-PCR and WB results (Fig. 3bandc).Con- Microvascular invasion sistently, immunofluorescence analysis confirmed these No 107 64 43 0.045 findings (Fig. 3d). Moreover, xenograft tumors generated KD from CD73 HCCLM3 cells exhibited an epithelial-like Yes 82 37 45 OE phenotype, while tumors derived from CD73 Edmondson stage SMMC7721 cells showed a mesenchymal-like phenotype I-II 126 74 52 0.039 according to IHC staining (Fig. 3e). We next examined the III-IV 63 27 36 correlation between CD73 and EMT in a clinical cohort. IHC staining in HCC consecutive sections indicated that Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 6 of 17 Table 2 Univariate Cox proportional regression analysis of factors associated with recurrence and overall survival Variables Recurrence Overall survival HR (95% CI) P HR (95% CI) P Age (> 50 years versus ≤ 50 years) 0.73 (0.50–1.07) 0.104 0.64 (0.41–1.02) 0.060 Sex (male versus female) 0.76 (0.48–1.21) 0.249 1.16 (0.69–1.95) 0.563 Liver cirrhosis (yes versus no) 1.52 (0.92–2.49) 0.099 1.11 (0.63–1.96) 0.712 ALT (> 40 U/L versus ≤ 40 U/L) 1.57 (1.06–2.34) 0.025 1.8 (0.85–2.24) 0.190 AST (> 40 U/L versus ≤40 U/L) 1.78 (1.20–2.64) 0.004 1.49 (0.92–2.41) 0.101 AFP (> 400 ng/ml versus ≤ 400 ng/ml) 1.86 (1.25–2.78) 0.002 1.49 (0.91–2.43) 0.117 No. of tumors (multi versus single) 1.81 (1.06–3.08) 0.030 1.57 (0.84–3.02) 0.157 Tumor size (> 5 cm versus ≤ 5 cm) 2.35 (1.60–3.43) < 0.001 1.59 (0.99–2.48) 0.056 Tumor encapsulation (none versus complete) 1.15 (0.78–1.69) 0.452 1.03 (0.64–1.67) 0.903 Satellite lesions (yes versus no) 1.67 (0.96–2.88) 0.069 0.78 (0.33–1.76) 0.528 Macrovascular invasion (yes versus no) 2.30 (1.28–4.12) 0.005 2.19 (1.13–4.28) 0.021 Microvascular invasion (yes versus no) 2.08 (1.42–3.04) < 0.001 1.91 (1.21–3.04) 0.006 Edmondson stage (III–IV versus I–II) 1.78 (1.21–2.61) 0.003 1.50 (0.94–2.39) 0.087 BCLC stage (B + C versus 0 + A) 1.98 (1.27–3.09) 0.003 1.72 (1.01–2.94) 0.045 ALBI grade (II versus I) 1.23 (0.80–1.89) 0.344 1.86 (1.12–3.09) 0.017 CD73 (high versus low) 2.33 (1.59–3.42) < 0.001 3.64 (2.20–6.01) < 0.001 Bold indicated statistical significance high CD73 expression was correlated with low level of was used and results showed that PI3K/AKT signaling ex- E-Cadherin but high level of N-Cadherin, and vice versa in hibited the greatest fold changes due to CD73 expression HCC tissue with low CD73 expression (Fig. 3f). Notably, manipulation (Fig. 4a). We therefore examined the PI3K/ correlation analysis revealed that high CD73 expression AKT pathway as a candidate critical signaling pathway. was positively correlated with N-Cadherin but negatively WB assays in HCC cells with CD73 knockdown con- correlated with E-Cadherin expression in clinical HCC firmed that phosphorylation levels of AKT and GSK3β,a samples (Fig. 3g). Collectively, our results suggested that direct substrate of pAKT, were reduced (Fig. 4b). In CD73 is a critical regulator of EMT in HCC. addition, expression of FOXO3a, which would undergo degradation due to AKT activation, was also increased CD73 promotes HCC progression and EMT by activating after CD73 knockdown. In contrast, pAKT and pGSK3β the PI3K-AKT signaling pathway levels were and increased, while FOXO3a was decreased To identify the underlying signaling of CD73 in HCC, with ectopic expression of CD73 in SMMC7721 cells Cignal Finder RTK signaling 10-Pathway Reporter Array (Fig. 4b). We next examined whether the effects of CD73 Table 3 Multivariate cox proportional regression analysis of factors associated with recurrence and overall survival Variables Recurrence Overall survival HR (95% CI) P HR (95% CI) P ALT (> 40 U/L versus ≤ 40 U/L) 1.24 (0.78–1.96) 0.369 N.A. N.A. AST (> 40 U/L versus ≤ 40 U/L) 1.46 (0.93–2.27) 0.099 N.A. N.A. AFP (> 400 ng/ml versus ≤ 400 ng/ml) 1.94 (1.29–2.94) 0.002 N.A. N.A. No. of tumors (multi versus single) 4.71 (1.28–17.42) 0.020 N.A. N.A. Tumor size (> 5 cm versus ≤ 5 cm) 2.01 (1.29–3.13) 0.002 N.A. N.A. Macrovascular invasion (yes versus no) 3.19 (0.83–12.29) 0.092 1.59 (0.95–2.67) 0.078 Microvascular invasion (yes versus no) 1.18 (0.75–1.85) 0.471 1.86 (1.16–2.96) 0.009 Edmondson stage (III–IV versus I–II) 1.33 (0.88–2.02) 0.182 0.98 (0.60–1.61) 0.932 ALBI grade (II versus I) 1.51 (0.63–3.60) 0.355 1.89 (1.09–3.18) 0.021 CD73 (high versus low) 2.82 (1.87–4.26) < 0.001 3.35 (2.01–5.52) < 0.001 Bold indicated statistical significance Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 7 of 17 Fig. 2 Biological functions of CD73 in HCC. a Efficiencies of CD73 knockdown and overexpression were validated by RT-PCR (left) and WB (right) assays. b Evaluation of the influence of CD73 on HCC proliferation by CCK-8. c Evaluations of the influence of CD73 on HCC proliferation by colony formation assays. d Evaluations of the influence of CD73 on cell cycle in HCC cells by flow cytometry. e Evaluations of the influence of CD73 on apoptosis in HCC cells by flow cytometry. f Evaluations of the influence of CD73 on migration and invasion activities of HCC cells by Transwell assays. g Evaluations of the influence of CD73 on migration activities of HCC cells by wound healing assays. h Establishment of the orthotopic xenograft model with NOG mice. Tumor volumes are shown in the right panel. i Incidences of intrahepatic (left) or lung (right) metastasis in indicated orthotopic xenograft model groups. All in vitro experiments were performed in triplicate; asterisk indicated P <0.050, and t tests were used Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 8 of 17 Fig. 3 CD73 triggers EMT in HCC. a Representative images of HCC cells with CD73 downregulation or overexpression. Cells were stained by FITC- phalloidin; scale bar 20 μm. b Expressions of EMT-related markers in HCC cells with CD73 downregulation or overexpression by RT-PCR assays. c Expressions of EMT-related markers in HCC cells with CD73 downregulation or overexpression by WB assays. d Expressions of CD73, E-Cadherin, N-Cadherin, and Vimentin in HCC cell lines with CD73 downregulation or overexpression by immunofluorescence staining; scale bar 200 μm. e Representative IHC images of CD73, E-Cadherin, and N-Cadherin expressions in tumor tissues derived from HCC cell lines as indicated. f Representative IHC images of CD73, E-Cadherin, and N-Cadherin expressions in consecutive tissue sections of clinical HCC samples. g Correlations between CD73 and E-Cadherin or N-Cadherin in clinical HCC samples were analyzed by Spearman’s rank correlation test. PCR assays were performed in triplicate. “N.S.” indicated not significant; asterisk indicated P <0.050, t tests were used on HCC cell activities were dependent on PI3K/AKT sig- HCCLM3 cells effectively attenuated the epithelial-induc- OE naling. Treatment of HCCLM3 cells with MK-2206, an tion effect, while inhibition of AKT signaling in CD73 AKT inhibitor, resulted in an epithelial-like molecular SMMC7721 cells prevented the mesenchymal-induction KD phenotype, which resembled the phenotype of CD73 effect (Fig. 4c). Also, biological experiments showed that HCCLM3 cells (Fig. 4c, Additional file 7: Figure S6A). AKT inhibition hindered proliferation and invasion cap- high Treatment of SMMC7721 cells with SC-79, an AKT acti- acities and promoted apoptosis in CD73 HCCLM3 vator, resulted in a mesenchymal-like phenotype that cells to levels observed in CD73 knockdown cells, while OE resembled the phenotype of CD73 SMMC7721 cells. AKT activation promoted proliferation and invasion cap- KD low Importantly, reactivation of AKT signaling in CD73 acities and prevented apoptosis in CD73 SMMC7721 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 9 of 17 Fig. 4 CD73 require its enzymatic activity and occur through adenosine receptor A2A to activate the PI3K/AKT pathway. a Cignal Finder RTK signaling 10-Pathway Reporter Array results demonstrate signaling changes in the indicated HCC cells. b Phosphorylation levels of AKT, GSK3β, and FOXO3a expression levels in the indicated HCC cells as determined by WB assays. c Expressions of EMT-related markers in the indicated HCC cells were detected by WB assays. d Proliferation in the indicated HCC cell lines was evaluated by CCK-8 and colony formation assays. “N.S.” indicated not significant; asterisk indicated P <0.050, and t tests were used. e Apoptosis in the indicated HCC cell lines was evaluated by flow cytometry assays; asterisk indicated P < 0.050 when compared with control group, and t tests were used. f Invasion in the indicated HCC cell lines was evaluated by Transwell assays; asterisk indicated P < 0.050 when compared with control group, and t tests were used. g Phosphorylation levels of AKT in HCCLM3 cells under different concentrations of APCP (left) or in SMMC7721 cells under different concentrations of adenosine treatments (right) as detectedby WB assays. h Phosphorylation levels of AKT in the indicated HCCLM3 (left) or SMMC7721 cells (right) as detected by WB assays. i Phosphorylation levels of AKT in HCCLM3 (left) or Hep3B cells (right) treated with antagonists targeting specific adenosine receptors (A1R, DPCPX; A2AR, KW6002; A2BR, CVT6883; A3R, Reversine). j Phosphorylation levels of AKT in indicated HCCLM3 (left) or SMMC7721 cells (right) as detected by WB assays. All in vitro experiments were performed in triplicate cells to levels observed in cells with CD73 overexpression CD73 function in HCC mainly depends on its enzymatic (Fig. 4d–f, Additional file 7:Figure S6B–D). Notably, AKT activity and occurs through adenosine receptor A2A inhibition abolished the effects of CD73 overexpression We next investigated whether the enzymatic activity of on HCC proliferation and invasion, whereas AKT activa- CD73 was required for its function in promoting progres- tion attenuated the inhibitory effects of CD73 knockdown sion and metastasis in HCC by using APCP, a CD73 en- high on HCC proliferation and invasion. These results indi- zyme activity inhibitor. APCP treatment of CD73 cated that CD73 promotes HCC progression and EMT by HCCLM3 cells reduced pAKT levels in a dose-dependent activating PI3K/AKT signaling. manner (Fig. 4g). Contrarily, exogenous adenosine Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 10 of 17 low treatment of CD73 SMMC7721 cells increased pAKT We next determined the role of Rap1 activation in in a dose-dependent manner. Notably, adenosine could re- AKT phosphorylation via manipulating Rap1 expression store the reduction of pAKT caused by CD73 knockdown, (Fig. 5b). In HCCLM3 cells, Rap1 knockdown dramatically while APCP treatment could partly abolish the increase of decreased pAKT levels, which could not be rescued by pAKT caused by CD73 overexpression (Fig. 4h). Function- CGS21680 treatment (Fig. 5c).However,transfectionof ally, APCP produced similar inhibitory effects on prolifer- Rap1G12V, a constitutively active mutant, attenuated the ation and invasion, induced an epithelial-like phenotype in reduction of pAKT level caused by CD73 knockdown or HCCLM3 cells as observed in CD73 knockdown cells, A2AR antagonist. Moreover, transfection of Rap1S17N, a and abolished the effects of CD73 overexpression (Add- dominant negative mutant, failed to restore the inhibitory itional file 8: Figure S7A-C), whereas adenosine promoted effects of Rap1 knockdown on pAKT levels (Fig. 5c). In proliferation and invasion, induced a mesenchymal-like SMMC7721 cells, transfection of Rap1G12V greatly in- phenotype in SMMC7721 cells as observed in CD73 over- creased pAKT levels, which could not be abolished by expression cells, and attenuated the effects of CD73 A2AR antagonist. However, CD73 overexpression or A2AR knockdown (Additional file 8:Figure S7D–F). Collectively, agonist showed no effects on pAKT level when Rap1 was our data demonstrated that CD73 functions in HCC re- knocked down. In addition, Rap1G12V expression rescued quire its enzymatic activity. the inhibition of AKT activation caused by Rap1 knock- We next searched for the adenosine receptor involved down (Fig. 5c). Importantly, through analyzing cytosolic in mediating the functions of CD73 in HCC using select- and plasma membrane fractions, we found that the majority ive inhibitors targeting adenosine A1 (DPCPX), A2A of Rap1 was recruited to the membrane after activation by (KW6002), A2B (CVT6883), and A3 receptors (Rever- the CD73-A2AR axis (Fig. 5d), which was consistent with a sine). Only KW6002 reduced AKT phosphorylation in previous study [28]. These results indicate that Rap1 serves high both CD73 HCC cell lines (Fig. 4i). Furthermore, as a key downstream regulator of the CD73-A2AR axis and KW6002 inhibited HCC proliferation and invasion to the that activation of Rap1 is essential for CD73-A2AR axis me- maximum extent and induced an epithelial-like phenotype diating AKT phosphorylation in HCC. (Additional file 9: Figure S8A–F). We thus focused on A2A receptor (A2AR) for further investigation. Inhibition The CD73-A2AR axis promotes Rap1-P110β interaction effects of A2AR on pAKT level could not be rescued by Previous studies reported that Rap1 could bind to Class I exogenous adenosine, while APCP failed to abolish the in- PI3Ks to promote PIP3 production, resulting in AKT acti- crease of pAKT resulting from using CGS21650, a select- vation [26, 27]. Thus we hypothesized that CD73-A2AR ive A2AR agonist (Fig. 4j). Importantly, A2AR agonist axis activates AKT signaling via promoting interaction be- could effectively restore the reduction of pAKT caused by tween Class I PI3K member and Rap1. First, to identify CD73 knockdown, while A2AR antagonist greatly abol- the key involved Class I PI3K member, we silenced two ished the increase of pAKT resulting from CD73 overex- major Class I PI3Ks, P110α, and P110β, in HCCLM3 cells pression (Fig. 4j). These data demonstrated that CD73 and found that P110β knockdown resulted in a dramatic exerts its function in HCC through A2AR. reduction of pAKT levels (Additional file 10:Figure S9A). Similar results were observed in Hep3B cells (Additional file 10: Figure S9B). These results suggested a The CD73-A2AR axis induces Rap1 activation and promotes critical role for P110β in CD73-mediated AKT activation its membrane recruitment in HCC. Next, Co-IP experiments in 293 T cells Adenosine A2AR belongs to the G-protein-coupled recep- co-transfected with HA-P110β and either Flag-Rap1G12V tor family, which is closely associated with the activation or Flag-Rap1S12N showed that only activated form Rap1 of Rap1 [25]. Previous studies showed that Rap1 is a crit- could interact with P110β (Fig. 5e). Notably, level of intra- ical regulator of PI3K activity [26, 27]. We therefore hy- cellular PIP3 and pAKT only elevated when Rap1 binding pothesized that the CD73-A2AR axis might activate PI3K/ to P110β, implying a requirement for this interaction in AKT signaling by inducing Rap1 activation. We first promoting catalytic activity of P110β (Additional file 10: assessed the effects of the CD73-A2AR axis on Rap1 acti- Figure S9C, D). vation. CD73 knockdown or A2AR antagonist greatly re- Above data demonstrated the significance of the inter- duced the level of Rap1-GTP, an active form of Rap1, action between Rap1 and P110β in regulating AKT acti- while A2AR agonist restored the reduction of Rap1-GTP vation. Next, the role of CD73-A2AR axis in promoting level caused by CD73 knockdown in HCCLM3 cells endogenous Rap1-P110β interaction was further vali- (Fig. 5a). Conversely, CD73 overexpression or A2AR dated by co-IP assays. CD73 knockdown or A2AR antag- agonist increased Rap1-GTP, while A2AR antagonist abol- onist greatly hindered the Rap1-P110β interaction, ished the increase of Rap1-GTP level resulting from CD73 whereas A2AR agonist successfully rescued the inhib- overexpression. ition of Rap1-P110β interaction caused by CD73 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 11 of 17 Fig. 5 CD73-A2AR axis activates PI3K/AKT signaling by inducing Rap1-mediated membrane localization of P110β. a Rap1 activation status was detected in the indicated HCC cells. b Rap1 expression was silenced via using shRNA in HCCLM3 and SMMC7721 cells and was confirmed by WB assays. c pAKT levels in the indicated HCCLM3 (upper) or SMMC7721 cells (lower) by WB assays. d Cytosolic and plasma membrane fractions were + + isolated from the indicated HCCLM3 (left) or SMMC7721 cells (right), followed by WB assays with antibody against Rap1. Na -K ATPase was used as an internal control for membrane fractions, and β-actin was used as an internal control for cytosolic fractions. e Co-immunoprecipitation of HA-P110β and Flag-Rap1G12V or Flag-Rap1S12N in SMMC7721 cells. f Effects of the CD73-A2AR axis on promoting Rap1-P110β interaction in co-IP assays. g Cytosolic and plasma membrane fractions were isolated from the indicated HCCLM3 (left) or SMMC7721 cells (right), followed by WB assays with antibody against P110β. pAKT level and PIP3 concentrations were also detected. ELISA experiments for PIP3 determinationwereconductedintriplicate. h Immunofluorescence staining of Rap1 (green) and P110β (red) with specific antibodies in indicated HCCLM3 (left) or SMMC7721 cells (right). DAPI was used for nuclear staining; scale bar 50 μm. Asterisk indicated P < 0.050 when compared with control group, t tests were used knockdown in HCCLM3 cells (Fig. 5f). Contrarily, CD73 the effects of CD73 overexpression on promoting overexpression or A2AR agonist promoted the Rap1-P110β Rap1-P110β binding in SMMC7721 cells. Our findings in- interaction, while A2AR antagonist dramatically inhibited dicate that CD73-A2AR serves as an upstream promoter of Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 12 of 17 the Rap1-P110β interaction, resulting in PIP3 production (Fig. 6c). Moreover, miR-193b expression was signifi- and AKT phosphorylation. cantly downregulated in HCC tissues compared to paratumor tissues (Fig. 6d) and reversely correlated The CD73-A2AR axis promotes Rap1-mediated membrane with CD73 mRNA in clinical HCC samples (r = − 0.83, localization of P110β and PIP3 production P <0.001, Fig. 6e). Functional assays demonstrated that As our results showed that activated Rap1 was recruited transfection of miR-193b could significantly inhibit HCC to the membrane, and Rap1 binds to P110β, we further proliferation, migration, and invasion, and these inhibitory evaluated whether CD73-A2AR could induce P110β effects could be rescued by CD73 re-expression (Fig. 6f–h, membrane localization, a critical process for its catalytic Additional file 11: Figure S10C and D). Additionally, in- function [29], in a Rap1-dependent manner. Through duction of epithelial-like phenotype and downregulation analyzing cytosolic and plasma membrane fractions by of phosphorylation of AKT caused by miR-193b mimic WB assays, we found that, similar to the effect of Rap1 transfection could also be reversed by re-expression of knockdown, CD73 knockdown or A2AR antagonist CD73 (Fig. 6i). Taken together, our data identified greatly reduced the plasma membrane localization of miR-193b as a novel negative regulator of CD73 expres- P110β in HCCLM3 cells, and this could be rescued by sion in HCC. Rap1G12V expression (Fig. 5g). On the other hand, con- sistent with the results of Rap1G12V transfection, CD73 Co-inhibition of CD73 and A2AR shows therapeutic overexpression or A2AR agonist promoted plasma mem- potential in vitro and in immunodeficient mice brane localization of P110β in SMMC7721 cells, and this We examined whether targeting CD73 in combination was greatly abolished by Rap1 knockdown. Notably, with A2AR inhibition would inhibit HCC growth inde- higher PIP3 concentrations and pAKT levels were ob- pendent of the immune system. Colony formation assays served when the majority of P110β localized on plasma showed that co-treatment with inhibitors against CD73 membrane (Fig. 5g), indicating abundant catalytic activ- and A2AR dramatically reduced cell proliferation in high ity of P110β under these circumstances. Assays were fur- CD73 HCCLM3 cells, while inhibiting CD73 or ther performed and results further confirmed the A2AR alone showed moderate inhibition effects (Fig. 7a). membranous co-localization of Rap1-P110β due to Moreover, co-targeting strategy induced more apoptosis CD73-A2AR axis regulation, which was consistent with in HCCLM3 cells than targeting CD73 or A2AR alone the results of WB assays (Fig. 5h). These data demon- (Fig. 7b). Targeting CD73 or A2AR alone in NOG mice strated that the CD73-A2AR axis activates AKT by indu- caused moderate suppression of tumor growth in vivo. cing Rap1-mediated membrane localization of P110β. However, dramatic and more durable responses were ob- served in mice co-treated with both CD73 and A2AR in- CD73 is negatively regulated by miR-193b hibitors (Fig. 7c). Similarly, Targeting CD73 or A2AR Previous studies revealed miRNA was responsible for alone caused moderate suppression of lung metastasis, the dysregulation of CD73 expression [30, 31]. We fur- while co-targeting strategy exerted a synergistic effect, ther investigated the potential upstream miRNA of resulting in maximal suppression of lung metastasis. CD73 in HCC. Seven microRNAs were predicted by two (Fig. 7d). Our findings revealed co-targeting CD73 and bioinformatics algorithms (TargetScan and miRanda) to A2AR exhibited potentials for preventing HCC progres- be potential upstream regulators of CD73 according to sion and metastasis. StarBase database [32] (Additional file 11: Figure S10A). Among these potential miRNAs, two homologous miR- Discussion NAs, miR-193a and miR-193b were selected for further CD73 plays a vital role in adenosinergic signaling by cata- study, as previous studies reported them as tumor sup- lyzing AMP into adenosine [10]. Hence, CD73 has been pressors in HCC [33, 34]. A putative binding site of mainly studied for its immunosuppression functions and miR-193a/miR-193b and 3′UTR of CD73 was identified was identified as a novel immune checkpoint target with (Fig. 6a). Transfection of miR-193b mimics significantly promising potential for suppressing tumor development high reduced CD73 expression in two CD73 cell lines [19, 35]. Meanwhile, studies also showed CD73 promotes (Fig. 6b). However, miR-193a mimics failed to exert tumor growth in an immune-independent manner [16, 17, similar effects (Additional file 11: Figure S10B), and 30, 31]. Here, functional experiments demonstrated that thus, miR-193b was identified as the potential regulator CD73 could promote HCC progression and metastasis. for CD73 expression in our study. To validate its regula- We also confirmed the clinical significance of CD73 as an tory role, dual luciferase reporter assays were firstly per- independent prognostic indicator of TTR and OS for formed and results revealed that miR-193b markedly HCC patients after curative resection. Moreover, CD73 inhibited the activity of a luciferase vector containing retained significant prognostic prediction value in the wild-type CD73 3′UTR, but not the mutant 3′UTR early-HCC and AFP-low patients, whose clinical outcomes Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 13 of 17 Fig. 6 CD73 is regulated by miR-193b. a Putative binding site of miR-193a/miR-193b and CD73. b Expression of CD73 in Hep3B and HCCLM3 cells after miR-193b mimic transfection as detected by RT-PCR and WB assays; asterisk indicated P < 0.050 when compared with parental group, and t tests were used. c Effect of miR-193b on the activity of CD73 mRNA 3′UTR was evaluated by luciferase reporter assays. d Comparison of miR-193b expression level between paratumor and tumor tissues determined by RT-PCR assays; asterisk indicated P < 0.050, n = 25, and t tests were used. e Correlation of miR-193b and CD73 mRNA in clinical HCC samples was analyzed by Spearman’s rank correlation test, n = 25. f Evaluations of proliferation capacities of indicated HCCLM3 (left) and Hep3B cells (right) by CCK-8 assays; “N.S.” indicated not significant; asterisk indicated P < 0.050, and t tests were used. g Evaluations of migration capacities of indicated HCCLM3 (left) and Hep3B cells (right) by wound healing assays; asterisk indicated P < 0.050, and t tests were used. h Evaluations of invasion capacities of indicated HCCLM3 (left) and Hep3B cells (right) by wound healing assays; asterisk indicated P < 0.050, t tests were used. i Evaluations of EMT-related genes and pAKT expression levels of indicated HCCLM3 (left) and Hep3B cells (right) by WB assays. In vitro experiments were performed in triplicate were difficult to predict by conventional indexes [24]. Our [36]. Previous studies reported CD73 as a critical regulator findings collaborated with the prognostic significance of for the maintenance of mesenchymal traits in mesenchymal CD73 in TCGA database (Additional file 12: Figure S11) stem cells and ovarian cancer cells [16, 37, 38]. In the and indicated that it might serve as an indicator to identify present study, we showed that CD73 expression was essen- HCC patients at high risk of recurrence that require fur- tial for maintaining the mesenchymal-like phenotype in ther interventions in addition to resection to improve HCC. Clinical samples confirmed our findings, showing their prognosis. that CD73 expression significantly correlated with EMT sta- EMT plays an important role in facilitating HCC metasta- tus. Therefore, we provide the first evidence demonstrating sis, despite its controversial function during carcinogenesis that CD73 serves as a trigger for EMT in HCC. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 14 of 17 Fig. 7 Co-targeting CD73 and A2AR shows therapeutic potential in HCC. a Colony formation assays in the indicated HCCLM3 cells. b Apoptosis assays in the indicated HCCLM3 cells. c Tumor growth kinetics in NOG model mice subcutaneously implanted with HCCLM3 cells treated as indicated. d Incidences of lung metastasis of different administered groups versus controls in NOG model mice intravenously injected with HCCLM3 cells. Asterisk indicated P < 0.050 when compared with controls; number sign indicated P < 0.050 when compared with APCP/KW5002 group alone; all in vitro experiments were performed in triplicate; and t tests were used Previous studies reported that CD73 exerted its functions speculated that co-targeting CD73 and A2AR might achieve by both enzymatic-dependent and enzymatic-independent satisfactory treatment effects in HCC. Indeed, CD73 inhib- ways [15]. Here, we found that CD73 exerts its tumor pro- ition in combination with A2AR blockade resulted in signifi- motion functions mainly through its enzymatic activity due cantly decreased tumor growth as well as metastasis to the significant changes in HCC cells after APCP or ad- compared with controls. Since our findings were observed enosine treatment. However, it should be noted that APCP in immunodeficient mouse models, we hypothesize that this showed a weaker effect on cell proliferation and invasion co-inhibition strategy might achieve more promising treat- than CD73, although this difference did not reach a statis- ment response by not only blocking progressive traits of tical significance. Similar results were observed when HCC but also restoring anti-tumor immune function within comparing adenosine with CD73 overexpression. These the tumor microenvironment. findings suggested the possibility of other regulatory Abnormal AKT activation is a hallmark of tumor pro- mechanisms of CD73 beyond producing adenosine. gression in various cancers, including HCC [39]. In the Thus, our future work will examine the non-enzymatic present study, Alterations of pGSK3β and pAKT levels functions of CD73 to elucidate the regulatory network were observed. More importantly, FOXO3a, a of CD73 in HCC. well-established tumor suppressor in HCC according to A2AR is a critical receptor that is responsible for the func- previous study [40], was also greatly repressed due to tion of CD73 in cancer immunity, and CD73-expressing CD73 expression, which served as solid evidence for AKT cancer cells are prone to form metastasis loci via activation activation. Our findings were consistent with previous of A2AR, which results in exhaustion of natural killer cells studies that CD73 activated AKT in breast and lung can- within the tumor microenvironment [8, 13, 14, 19]. Here, cer cells [17, 30]. However, the underlying mechanism of we provided the first evidence that A2AR serves as the key how CD73 triggered AKT activation had not been eluci- downstream mediator of CD73 to regulate the progressive dated clearly by previous studies. Here, we showed that phenotype of HCC, and inhibition of A2AR dramatically the CD73-A2AR axis served as a “switch” for Rap1 activa- abolished the effects of CD73 on HCC cells. Our data tion, thus resulting in membrane localization of P110β strongly demonstrate the non-immunosuppressive effects of and PIP3 production, which are critical steps for AKT ac- A2AR, thus expanding our understanding of the function of tivation. Rap1 is a central regulator of cell adhesion and A2AR in HCC development. Based on these findings, we motility, and aberrant activation of Rap1 can result in Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 15 of 17 carcinogenesis and tumor progression [27]. Importantly, Additional files Rap1 and Ras share similar binding partners, including PI3K [41]. Our study not only confirmed the importance Additional file 1: Supplementary methods and materials. (DOCX 28 kb) of Rap1 in HCC progression and metastasis, but also iden- Additional file 2: Figure S1. Prognostic value of CD73 in subgroups of HCC patients. (A) Kaplan–Meier analysis of TTR of early-stage (BCLC 0+A) tified Rap1 as a key downstream responder of patients after curative resection according to CD73 expression level. (B) CD73-A2AR that recruits P110β to the plasma mem- Kaplan–Meier analysis of OS of early-stage (BCLC 0+A) patients after cura- brane. It was reported that cAMP acted as a key down- tive resection according to CD73 expression level. (C) Kaplan–Meier ana- lysis of TTR of curative resection according to CD73 expression level. (D) stream molecule of GPCR and could induce Rap1 Kaplan–Meier analysis of OS of low-AFP patients after curative resection activation [42]. Based on our Cignal pathway screening re- according to CD73 expression level. (TIF 1853 kb) sults, we also found that cAMP/PKA signaling exhibited Additional file 3: Figure S2. Function of CD73 in Hep3B cell line. (A) significant changes due to CD73 alteration (Fig. 4a). Efficiencies of CD73 knockdown in Hep3B cells were evaluated by WB assays. (B) Effects of CD73 knockdown on proliferation in Hep3B cells Hence, A2AR might mediate Rap1 activation through pro- were evaluated by CCK-8 and colony formation assays. (C) Effects of moting cAMP production, and detailed mechanism is cur- CD73 knockdown on cell cycle in Hep3B cells were evaluated by flow rently ongoing in our lab. Previous reports also showed cytometry assays. (D) Effects of CD73 knockdown on apoptosis in Hep3B cells were evaluated by flow cytometry assays. (E) Effects of CD73 knock- that CD73 was positively correlated with EGFR [22], and down on migration and invasion in Hep3B cells were evaluated by future studies should also examine their potential inter- Transwell assays. (F) Effects of CD73 knockdown on migration were action in HCC. validated by wound healing assays. (G) Effects of CD73 knockdown on in vivo tumor growth. “N.S.” indicated not significant; Asterisk indicated Regulatory network of miRNA-mRNA interaction was P < 0.050, all in vitro experiments were performed in triplicate, t tests important for HCC progression. miR-193a was reported were used. (TIF 10457 kb) as a tumor suppressor and could be found in HCC Additional file 4: Figure S3. Effects of CD73 on cell motility. (A) cell-derived exosomes [33]. However, miR-193b did not Representative images of Transwell assays conducted with control or CD73 knockdown (KD) HCCLM3 cells. (B) Representative images of serve as a significant prognostic indicator from Transwell assays conducted with control or CD73 overexpression (OE) HCC-derived exosomes. The major explanation for this SMMC7721 cells. (C) Representative images of wound healing assays might be the different expression regulatory mechanism conducted with control or CD73-KD HCCLM3 cells. (D) Representative images of wound healing assays conducted with control or CD73-OE of these two miRNAs in HCC: low miR-193b expression SMMC7721 cells. “N.S.” indicated not significant; Asterisk indicated P <0.050, might due to inhibition of transcription while low all in vitro experiments were performed in triplicate, t tests were used. miR-193a level might result from selective packaging and (TIF 13202 kb) secreting through exosomes. Therefore, it was rational Additional file 5: Figure S4. Biological effects of CD73 overexpression in HCCLM3 cells. (A) Efficiencies of overexpression of CD73 in HCCLM3 that miR-193b could rarely be found in HCC-derived exo- cells were validated by RT-PCR (left) and WB assays (right). (B) Effects of somes. Meanwhile, CD73 expression showed little re- CD73 overexpression on proliferation were evaluated by CCK-8 (left) and sponse to miR-193a mimics transfection in our study. We colony formation (right) assay. (C) Effects of CD73 overexpression on migration and invasion were evaluated by Transwell assays. (D) Effects of speculated this phenomenon might result from different CD73 overexpression on migration were validated by wound healing clearance mechanism of exogenous miRNA between assays. (E) Effects of CD73 overexpression on cell survival were evaluated miR-193a and miR-193b. However, our findings demon- by flow cytometry. (F) Effects of CD73 overexpression on cell cycle were evaluated by flow cytometry. (G) Effects of CD73 overexpression on strated that miR-193b was the major negative regulator of E-cadherin, N-Cadherin, and pAKT level were evaluated by WB assays. CD73 in HCC, and our future work will aim at clarifying Asterisk indicated P < 0.050, all in vitro experiments were performed in the underlying mechanism of inhibitory difference be- triplicate, t tests were used. (TIF 5605 kb) tween miR-193a and miR-193b. Additional file 6: Figure S5. In vivo function of CD73 in HCC. (A) Representative IHC images of PCNA expression level in indicated HCCLM3 (left) or SMMC7721 cells (right). (B) Representative image of intrahepatic metastasis. (C) Representative images of lung metastasis in mice implanted Conclusions with HCCLM3 or SMMC7721 cells. (TIF 10892 kb) Our study demonstrated that CD73 was an independent Additional file 7: Figure S6. Role of PI3K/AKT signaling in CD73 mediating HCC progression. (A) Expressions of EMT-related markers in the prognostic indicator for HCC. Functional experiments indicated HCC cells were detected by RT-PCR assays; asterisk indicated established the essential role of CD73 in promoting HCC P < 0.050, experiments were performed in triplicate, and t tests were used. progression and metastasis through PI3K/AKT signaling (B) Representative images of colony formation assays of the indicated HCCLM3 (upper) and SMMC7721 cells (lower). (C) Representative results by inducing Rap1-mediated membrane localization of of apoptosis assays in the indicated HCCLM3 (upper) and SMMC7721 cells P110β. We also identified miR-193b as a potential up- (lower). (D) Representative images of Transwell assays of indicated stream regulator of CD73 in HCC. Notably, co-targeting HCCLM3 (upper) and SMMC7721 cells (lower). (TIF 9692 kb) CD73 and A2AR achieved satisfactory inhibitory effects Additional file 8: Figure S7. CD73 function in HCC depends on its enzymatic activity. (A) Proliferation capacity in the indicated HCCLM3 on HCC in vitro and vivo. Together these results demon- detected by CCK-8 assays. (B) Invasion capacity in the indicated HCCLM3 strate the importance of CD73 in HCC in addition to its detected by Transwell assays. (C) Expressions of EMT-related markers in immunosuppressive functions and revealed that a CD73 indicated HCCLM3 cells were detected by RT-PCR. (D) Proliferation capacity in the indicated SMMC7721 detected by CCK-8 assays. (E) targeting strategy may be a promising novel therapeutic Invasion capacity in the indicated SMMC7721 detected by Transwell strategy for future HCC management. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 16 of 17 Scheme (20130071140008), the Projects from Shanghai Science and assays. (F) Expressions of EMT-related markers of indicated SMMC7721 Technology Commission (14DZ1940300, 14411970200), The Strategic Priority cells were detected by RT-PCR. “N.S.” indicated not significant; asterisk Research Program of the Chinese Academy of Science (XDA12020105). indicated P < 0.050, all in vitro experiments were performed in triplicate, Bei-Li Wang was supported by the Projects from Shanghai Science and and t tests were used. (TIF 4551 kb) Technology Commission (16411952100). Additional file 9: Figure S8. CD73 function in HCC occurs through Yun-Fan Sun was supported by National Natural Science Foundation of adenosine receptor A2A. (A) Proliferation capacity in HCCLM3 cells China (81602543) and the Sailing Program from the Shanghai and Technology treated with indicated adenosine receptor antagonists detected by CCK-8 Commission (16YF1401400). assays. (B) Invasion capacity in HCCLM3 cells treated with indicated Bo Hu was supported by National Natural Science Foundation of China adenosine receptor antagonists detected by Transwell assays. (C) Expressions (8180101167 and 31771472) and Zhongshan Hospital Science Foundation of EMT-related markers in HCCLM3 cells treated with indicated adenosine (2016ZSQN30, 2017ZSQN18, 2017ZSYQ26). receptor antagonists were detected by RT-PCR. (D) Proliferation capacity in Hep3B cells treated with indicated adenosine receptor antagonists detected Availability of data and materials by CCK-8 assays. (E) Invasion capacity in Hep3B cells treated with indicated All data generated or analyzed during this study are included in this adenosine receptor antagonists detected by Transwell assays. (F) Expressions published article and its supplementary information files. of EMT-related markers of Hep3B cells treated with indicated adenosine receptor antagonists were detected by RT-PCR. Asterisk indicated P <0.050, Authors’ contributions all in vitro experiments were performed in triplicate, and t tests were used. XLM, XRY, and BH contributed to the study design. XLM, BH, WJY, LHL, HW, (TIF 5733 kb) YZ, CYZ, ALJ, and YFS contributed to the acquisition, analysis, and interpretation Additional file 10: Figure S9. Role of Rap1 activation in AKT of the data. XLM, BH, and XRY contributed to the drafting of the manuscript. phosphorylation. (A) Expressions of P110α, P110β, pAKT, and AKT in HCCLM3 JZ, BSP, WG, XRY, and JF contributed to the critical revision of the manuscript. cells were detected by WB assays. (B) Expressions of P110α, P110β, pAKT, and All authors read and approved the final manuscript. AKT in Hep3B cells were detected by WB assays. (C) Expressions of pAKT and AKT in 293T transfected with indicated plasmids were detected by WB assays. Ethics approval and consent to participate (D) Cellular PIP2 and PIP3 levels of 293T transfected with indicated plasmids Present study was performed in accordance with the 1975 Declaration of were detected by ELISA assays. “N.S.” indicated not significant; asterisk Helsinki. Approval for the use of human subjects was obtained from the indicated P < 0.050, experiments were performed in triplicate, and t tests research ethics committee of Zhongshan Hospital, and informed consent were used. (TIF 725 kb) was obtained from each individual enrolled in this study. Additional file 11: Figure S10. miR-193b served as an upstream negative regulator for CD73 expression in HCC. (A) Prediction results of microRNAs Consent for publication that potentially regulate CD73 according to StarBase 2.0. (B) Effects of Not applicable miR-193a mimics transfection on CD73 expression in HCC cells were evaluated by RT-PCR and WB assays. (C) Representative images of Transwell Competing interests assays in indicated HCCLM3 (upper) and Hep3B (lower) cells. (D) Migration The authors declare that they have no competing interests. evaluation via wound healing assays of miR-193b mimic-transfected HCC cells with or without CD73 overexpression. All in vitro experiments were performed in triplicate, and t tests were used. (TIF 6462 kb) Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published Additional file 12: Figure S11 Clinical significance of CD73 in HCC maps and institutional affiliations. according to TCGA database. Data was collected from The Human Protein Atlas (https://www.proteinatlas.org/ENSG00000135318-NT5E/ Author details pathology/tissue/liver+cancer). (TIF 378 kb) Department of Laboratory Medicine, Zhongshan Hospital, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, People’s Republic of China. Abbreviations Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, A2AR: Adenosine A2A receptor; BCLC: Barcelona Clinic Liver Cancer; co- Fudan University, Key Laboratory of Carcinogenesis and Cancer Invasion, IP: Co-immunoprecipitation; CTC: Circulating tumor cell; DMEM: Dulbecco’s Ministry of Education, Shanghai 200032, People’s Republic of China. Liver modified Eagle’s medium; EMT: Epithelial-mesenchymal transition; Cancer Institute, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, HCC: Hepatocellular carcinoma; HR: Hazard ratio; NOG: NOD/SCID/γc(null); People’s Republic of China. OS: Overall survival; shRNA: Short hairpin RNA; TTR: Time to recurrence; WB: Western blot Received: 21 December 2018 Accepted: 27 March 2019 Acknowledgements We thank Prof Ji Yuan and Dr. Lu Shao-Hua from the Department of Pathology, References Zhongshan Hospital for their help in pathological diagnosis. Also, we thank the 1. Llovet JM, Montal R, Sia D, Finn RS. Molecular therapies and precision medicine participating patients for the source of clinical blood samples. for hepatocellular carcinoma. Nat Rev Clin Oncol. 2018;15:599–616. 2. Reig M, Da FL, Faivre S. New trials and results in systemic treatment of HCC. Funding J Hepatol. 2018;69:525–33. Wei Guo was supported by the National Natural Science Foundation of China 3. Dutta R, Mahato RI. Recent advances in hepatocellular carcinoma therapy. (81772263 and 81572064) and Key Developing Disciplines of Shanghai Pharmacol Ther. 2017;173:106–17. Municipal Commission of Health and Family Planning (2015ZB0201). 4. Galle PR, Tovoli F, Foerster F, Worns MA, Cucchetti A, Bolondi L. The Xin-Rong Yang was supported by the National Natural Science Foundation treatment of intermediate stage tumours beyond TACE: from surgery to of China (81672839, 81472676 and 81872355), National Key Research and systemic therapy. J Hepatol. 2017;67:173–83. Development Program of China (2016YFF0101405), the project from Shanghai 5. Huo X, Han S, Wu G, Latchoumanin O, Zhou G, Hebbard L, et al. Science and Technology Commission (14DZ1940302, 1411970200, 14140902301), Dysregulated long noncoding RNAs (lncRNAs) in hepatocellular carcinoma: the Strategic Priority Research Program of the Chinese Academy of Sciences implications for tumorigenesis, disease progression, and liver cancer stem (XDA12020103). cells. Mol Cancer. 2017;16:165. Jia Fan was supported by the National High Technology Research and 6. Liu FF, Liu YN, Chen Z. Tim-3 expression and its role in hepatocellular Development Program (863 Program) of China (2015AA020401), the State carcinoma. J Hematol Oncol. 2018;11:126. Key Program of National Natural Science of China (81530077, and 81772551), 7. Zhou SL, Hu ZQ, Zhou ZJ, Dai Z, Wang Z, Cao Y, et al. miR-28-5p-IL-34- Specialized Research Fund for the Doctoral Program of Higher Education macrophage feedback loop modulates hepatocellular carcinoma metastasis. and Research Grants Council Earmarked Research Grants Joint Research Hepatology. 2016;63:1560–75. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 17 of 17 8. Antonioli L, Pacher P, Vizi ES, Hasko G. CD39 and CD73 in immunity and 35. Kordass T, Osen W, Eichmuller SB. Controlling the immune suppressor: inflammation. Trends Mol Med. 2013;19:355–67. transcription factors and microRNAs regulating CD73/NT5E. Front Immunol. 9. Allard D, Allard B, Gaudreau PO, Chrobak P, Stagg J. CD73-adenosine: a next- 2018;9:813. generation target in immuno-oncology. Immunotherapy. 2016;8:145–63. 36. Jayachandran A, Dhungel B, Steel JC. Epithelial-to-mesenchymal plasticity of 10. Allard B, Longhi MS, Robson SC, Stagg J. The ectonucleotidases CD39 and cancer stem cells: therapeutic targets in hepatocellular carcinoma. J CD73: novel checkpoint inhibitor targets. Immunol Rev. 2017;276:121–44. Hematol Oncol. 2016;9:74. 37. Breitbach M, Kimura K, Luis TC, Fuegemann CJ, Woll PS, Hesse M, et al. In 11. Gourdin N, Bossennec M, Rodriguez C, Vigano S, Machon C, Jandus C, et al. vivo labeling by CD73 marks multipotent stromal cells and highlights Autocrine adenosine regulates tumor polyfunctional CD73(+)CD4(+) effector endothelial heterogeneity in the bone marrow niche. Cell Stem Cell. 2018; T cells devoid of immune checkpoints. Cancer Res. 2018;78:3604–18. 22:262–76. 12. Li L, Wang L, Li J, Fan Z, Yang L, Zhang Z, et al. Metformin-induced 38. Suto EG, Mabuchi Y, Suzuki N, Suzuki K, Ogata Y, Taguchi M, et al. Prospectively reduction of CD39 and CD73 blocks myeloid-derived suppressor cell activity isolated mesenchymal stem/stromal cells are enriched in the CD73(+) in patients with ovarian cancer. Cancer Res. 2018;78:1779–91. population and exhibit efficacy after transplantation. Sci Rep. 2017;7:4838. 13. Allard B, Pommey S, Smyth MJ, Stagg J. Targeting CD73 enhances the 39. Zhou Q, Lui VW, Yeo W. Targeting the PI3K/Akt/mTOR pathway in antitumor activity of anti-PD-1 and anti-CTLA-4 mAbs. Clin Cancer Res. 2013; hepatocellular carcinoma. Future Oncol. 2011;7:1149–67. 19:5626–35. 40. Cao MQ, You AB, Zhu XD, Zhang W, Zhang YY, Zhang SZ, et al. miR-182-5p 14. Young A, Ngiow SF, Barkauskas DS, Sult E, Hay C, Blake SJ, et al. Co- promotes hepatocellular carcinoma progression by repressing FOXO3a. J inhibition of CD73 and A2AR adenosine signaling improves anti-tumor Hematol Oncol. 2018;11:12. immune responses. Cancer Cell. 2016;30:391–403. 41. Frische EW, Zwartkruis FJ. Rap1, a mercenary among the Ras-like GTPases. 15. Gao ZW, Dong K, Zhang HZ. The roles of CD73 in cancer. Biomed Res Int. Dev Biol. 2010;340:1–9. 2014;2014:460654. 42. Onodera Y, Nam JM, Bissell MJ. Increased sugar uptake promotes oncogenesis 16. Lupia M, Angiolini F, Bertalot G, Freddi S, Sachsenmeier KF, Chisci E, et al. via EPAC/RAP1 and O-GlcNAc pathway. J Clin Invest. 2014;124:367–84. CD73 regulates stemness and epithelial-mesenchymal transition in ovarian cancer-initiating cells. Stem Cell Rep. 2018;10:1412–25. 17. Yu J, Wang X, Lu Q, Wang J, Li L, Liao X, et al. Extracellular 5′-nucleotidase (CD73) promotes human breast cancer cells growth through AKT/GSK-3beta/ beta-catenin/cyclinD1 signaling pathway. Int J Cancer. 2018;142:959–67. 18. Allard B, Turcotte M, Spring K, Pommey S, Royal I, Stagg J. Anti-CD73 therapy impairs tumor angiogenesis. Int J Cancer. 2014;134:1466–73. 19. Beavis PA, Divisekera U, Paget C, Chow MT, John LB, Devaud C, et al. Blockade of A2A receptors potently suppresses the metastasis of CD73+ tumors. Proc Natl Acad Sci U S A. 2013;110:14711–6. 20. Buisseret L, Pommey S, Allard B, Garaud S, Bergeron M, Cousineau I, et al. Clinical significance of CD73 in triple-negative breast cancer: multiplex analysis of a phase III clinical trial. Ann Oncol. 2018;29:1056–62. 21. Turcotte M, Spring K, Pommey S, Chouinard G, Cousineau I, George J, et al. CD73 is associated with poor prognosis in high-grade serous ovarian cancer. Cancer Res. 2015;75:4494–503. 22. Shali S, Yu J, Zhang X, Wang X, Jin Y, Su M, et al. Ecto-5′-nucleotidase (CD73) is a potential target of hepatocellular carcinoma. J Cell Physiol. 2018. https://doi.org/10.1002/jcp.27694. 23. Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis. 2010;30:61–74. 24. Guo W, Sun YF, Shen MN, Ma XL, Wu J, Zhang CY, et al. Circulating tumor cells with stem-like phenotypes for diagnosis, prognosis, and therapeutic response evaluation in hepatocellular carcinoma. Clin Cancer Res. 2018;24:2203–13. 25. Cha I, Lee SH, Jeon TJ. Chemoattractant-mediated Rap1 activation requires GPCR/G proteins. Mol Cells. 2010;30:563–7. 26. Kortholt A, Bolourani P, Rehmann H, Keizer-Gunnink I, Weeks G, Wittinghofer A, et al. A Rap/phosphatidylinositol 3-kinase pathway controls pseudopod formation [corrected]. Mol Biol Cell. 2010;21:936–45. 27. Shah S, Brock EJ, Ji K, Mattingly RR. Ras and Rap1: a tale of two GTPases. Semin Cancer Biol. 2018. https://doi.org/10.1016/j.semcancer.2018.03.005. 28. Xiong Y, Ye C, Yang N, Li M, Liu H. Ubc9 binds to ADAP and is required for Rap1 membrane recruitment, Rac1 activation, and integrin-mediated T cell adhesion. J Immunol. 2017;199:4142–54. 29. Thorpe LM, Yuzugullu H, Zhao JJ. PI3K in cancer: divergent roles of isoforms, modes of activation and therapeutic targeting. Nat Rev Cancer. 2015;15:7–24. 30. Zhu J, Zeng Y, Li W, Qin H, Lei Z, Shen D, et al. CD73/NT5E is a target of miR-30a-5p and plays an important role in the pathogenesis of non-small cell lung cancer. Mol Cancer. 2017;16:34. 31. Xie M, Qin H, Luo Q, Huang Q, He X, Yang Z, et al. MicroRNA-30a regulates cell proliferation and tumor growth of colorectal cancer by targeting CD73. BMC Cancer. 2017;17:305. 32. Li JH, Liu S, Zhou H, Qu LH, Yang JH. starBase v2.0: decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP- Seq data. Nucleic Acids Res. 2014;42:D92–7. 33. Li S, Yao J, Xie M, LiuY ZM. Exosomal miRNAs in hepatocellular carcinoma development and clinical responses. J Hematol Oncol. 2018;11:54. 34. Wei R, Huang GL, Zhang MY, Li BK, Zhang HZ, Shi M, et al. Clinical significance and prognostic value of microRNA expression signature in hepatocellular carcinoma. Clin Cancer Res. 2013;19:4780–91. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Hematology & Oncology Springer Journals

CD73 promotes hepatocellular carcinoma progression and metastasis via activating PI3K/AKT signaling by inducing Rap1-mediated membrane localization of P110β and predicts poor prognosis

Loading next page...
 
/lp/springer-journals/cd73-promotes-hepatocellular-carcinoma-progression-and-metastasis-via-5DMF0KIyRs

References (50)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s).
Subject
Medicine & Public Health; Oncology; Hematology; Cancer Research
eISSN
1756-8722
DOI
10.1186/s13045-019-0724-7
Publisher site
See Article on Publisher Site

Abstract

Background: Hepatocellular carcinoma (HCC) is one of the most prevalent malignancies worldwide because of rapid progression and high incidence of metastasis or recurrence. Accumulating evidence shows that CD73-expressing tumor cell is implicated in development of several types of cancer. However, the role of CD73 in HCC cell has not been systematically investigated and its underlying mechanism remains elusive. Methods: CD73 expression in HCC cell was determined by RT-PCR, Western blot, and immunohistochemistry staining. Clinical significance of CD73 was evaluated by Cox regression analysis. Cell counting kit-8 and colony formation assays were used for proliferation evaluation. Transwell assays were used for motility evaluations. Co-immunoprecipitation, cytosolic and plasma membrane fractionation separation, and ELISA were applied for evaluating membrane localization of P110β and its catalytic activity. NOD/SCID/γc(null) (NOG) mice model was used to investigate the in vivo functions of CD73. Results: In the present study, we demonstrate that CD73 was crucial for epithelial-mesenchymal transition (EMT), progression and metastasis in HCC. CD73 expression is increased in HCC cells and correlated with aggressive clinicopathological characteristics. Clinically, CD73 is identified as an independent poor prognostic indicator for both time to recurrence and overall survival. CD73 knockdown dramatically inhibits HCC cells proliferation, migration, invasion, and EMT in vitro and hinders tumor growth and metastasis in vivo. Opposite results could be observed when CD73 is overexpressed. Mechanistically, adenosine produced by CD73 binds to adenosine A2A receptor (A2AR) and activates Rap1, which recruits P110β to the plasma membrane and triggers PIP3 production, thereby promoting AKT phosphorylation in HCC cells. Notably, a combination of anti-CD73 and anti-A2AR achieves synergistic depression effects on HCC growth and metastasis than single agent alone. (Continued on next page) * Correspondence: yang.xinrong@zs-hospital.sh.cn; guo.wei@zs-hospital.sh.cn Xiao-Lu Ma, Min-Na Shen and Bo Hu contributed equally to this work. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai 200032, People’s Republic of China Department of Laboratory Medicine, Zhongshan Hospital, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, People’s Republic of China Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 2 of 17 (Continued from previous page) Conclusions: CD73 promotes progression and metastasis through activating PI3K/AKT signaling, indicating a novel prognostic biomarker for HCC. Our data demonstrate the importance of CD73 in HCC in addition to its immunosuppressive functions and revealed that co-targeting CD73 and A2AR strategy may be a promising novel therapeutic strategy for future HCC management. Keywords: Hepatocellular carcinoma, CD73, Epithelial-mesenchymal-transition,Prognosis,PI3K/AKT Introduction vivo. Adenosine produced by CD73 binds to adenosine Hepatocellular carcinoma (HCC) is one of the most A2A receptor (A2AR) and activates Rap1, which recruits prevalent malignancies worldwide and the second most P110β to the plasma membrane and triggers PIP3 pro- common cause of cancer-related death [1, 2]. Although duction, thereby promoting AKT phosphorylation. Im- great improvements in treatment have been made, the portantly, the combination of CD73 and A2AR prognosis of HCC remains unfavorable, with an approxi- inhibitors provided more synergistic tumor inhibition mate 30% overall 5-year survival rate even after curative than either regimen alone in HCC. resection [3–5]. Recurrence or metastasis is the major reason for the poor survival of HCC patients [6, 7], but the precise regulatory mechanisms of invasion and me- Methods and materials tastasis remain elusive. Therefore, identification of crit- Clinical specimens and follow-up ical molecules that contribute to the invasive phenotype Four groups of patients were recruited in the present of HCC and clarification of the underlying mechanism study. In group I, frozen tumor tissues with paired para- are urgently needed to improve HCC prognosis. tumor normal tissues from 25 HCC patients receiving CD73, an AMP hydrolyzing enzyme that regulates the curative resection in October 2010 in Zhongshan conversion of extracellular ATP into adenosine, functions Hospital were collected for comparisons of CD73 ex- as a powerful immunosuppressor for maintaining tissue pression between matched tumor and paratumor tissues homeostasis and preventing immune responses during in- by RT-PCR and Western blot (WB) assays. In group II, flammation [8]. Accumulating evidence demonstrates the eight recurrent HCC patients receiving curative resec- vital role of CD73 in tumor [9, 10]. Hematopoietic CD73 tion from April to November 2014 in Zhongshan Hos- overexpression within the tumor microenvironment is ob- pital were enrolled, and matched frozen paratumor, served in a variety of cancers, and these CD73-expressing primary tumor, and recurrent tumor tissues were col- cells hamper the immune reaction towards cancer cells lected for further WB assays. In group III, 10 HCC pa- and disable the cytotoxic antitumor immune response by tients suffered lung metastasis after curative resection producing high levels of adenosine [11–14]. CD73 from 2011 to 2015 in Zhongshan hospital were enrolled, expressed on hematopoietic cells can also function as a and matched paraffin sections of paratumor, primary co-stimulatory molecule in human T cells to block CD8 tumor, and metastatic tumor tissues were collected for T cell activation [15]. Recent reports also demonstrated further immunohistochemical staining. In group IV, 189 that CD73 overexpression is involved in migration, inva- HCC patients receiving curative resection in Zhongshan sion, and angiogenesis [15–18]. Moreover, clinical data Hospital from March 2012 to September 2013 were en- demonstrated that CD73 is a biomarker of poor prognosis rolled, and paratumor and tumor specimens were col- in solid cancers [19–21]. These investigations suggested lected for tissue microarrays (TMA) establishment. that CD73 could be a critical regulator that promotes Enrollment criteria were according to a previous study tumor progression in an immune-independent manner. A [7]. HCC diagnosis was based on histopathology accord- preliminary study reported that CD73 was overexpressed ing to the American Association for Study of Liver in HCC cells and was positively correlated with EGFR ex- Disease guidelines. The Barcelona Clinic Liver Cancer pression [22]. However, whether CD73 could promote (BCLC) staging system was used to assess tumor stage HCC progression and metastasis and the underlying regu- [23]. Approval for the use of human subjects was ob- latory mechanism still needs to be elucidated. tained from the research ethics committee of Zhongshan Here, we report that CD73 expression is positively cor- Hospital, and informed consent was obtained from each related with metastasis in HCC and is an independent individual. Follow-up ended on December 2016. Time to indicator for predicting prognosis. CD73 overexpression recurrence (TTR) was defined as the interval between enhanced HCC progression and metastasis in the ab- treatment and intrahepatic recurrence or extrahepatic sence of an immunological environment in vitro and in metastasis. Overall survival (OS) was defined as the Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 3 of 17 interval between treatment and death of any cause or multivariate analysis demonstrated high CD73 expres- the last observation date. sion on HCC cell and was an independent indicator for predicting both TTR [HR 2.82 (1.87–4.26), P <0.001] and OS [HR 3.35 (2.01–5.52), P < 0.001, Tables 2 and 3]. Statistical analysis Intriguingly, patients with high CD73 had significantly Statistical analyses were performed using SPSS 20.0 soft- higher pre-treatment circulating tumor cell (CTC) load ware (IBM, Chicago, IL, USA). Experimental values for (P < 0.001, Fig. 1h). Using 0.80 as the cutoff value for continuous variables were expressed as the mean ± stand- CTC load [24], a higher proportion of patients in the ard error of the mean. The chi-squared test, Fisher’sexact high CD73 group had high CTC loads (Fig. 1h). Notably, probability tests, and the Student’s t test were used as ap- CD73 showed elevated expression in MHCC97H and propriate to evaluate the significance of differences in data HCCLM3 cells, which have higher invasive and meta- between groups. If variances within groups were not static potentials (Fig. 1i). In comparison, cell lines with homogeneous, a nonparametric Mann–Whitney test or a lower invasive and metastatic capacities such as Wilcoxon signed-rank test was used. Prognostic value was SMMC7721 and MHCC97L showed relatively low CD73 evaluated by Kaplan–Meier survival curves, log-rank tests, levels. Collectively, our data demonstrated the clinical and Cox proportional hazards models. A P value less than significance of CD73 and suggested it as a potential pro- 0.05 was considered significant. moter for HCC metastasis/recurrence. Further details of materials and methods are described in Additional file 1. CD73 promotes HCC progression in vitro Results CD73 expression in HCC cell lines was manipulated by CD73 is overexpressed in HCC tissues and correlates with ectopic expression or short hairpin RNA (shRNA) poor prognosis knockdown. Two distinct shRNAs were designed to high CD73 expression was markedly higher in 55% of HCC knock down CD73 expression in two CD73 HCC cell tissues than paired adjacent normal liver tissues lines, HCCLM3 and Hep3B cells. Knockdown effects (Additional file 2: Figure S1A). WB assays confirmed the were validated by RT-PCR and WB assays (Fig. 2a, RT-PCR findings (Fig. 1b). Moreover, immunohisto- Additional file 3: Figure S2A). Stable ectopic expression low chemistry (IHC) staining was conducted with 189 HCC of CD73 in a CD73 HCC cell line, SMMC7721, was patients receiving curative resection, and results showed also confirmed by RT-PCR and WB assays. We found KD that HCC tissues expressed higher level of CD73 than CD73 HCCLM3 cells exhibited significantly decreased adjacent non-cancerous liver tissues, according to the proliferation potential compared with controls according criteria of CD73 expression levels (Fig. 1c). These data to CCK-8 and colony formation assays (Fig. 2b and c). revealed the potential oncogenic role of CD73 in HCC. Similar reduced proliferation was observed in Hep3B Next, the correlation between CD73 and HCC progno- cells (Additional file 3: Figure S2B). In contrast, overex- sis was evaluated. Samples from eight recurrent patients pression of CD73 in SMMC7721 cells increased their were collected. WB assays showed that expression levels proliferation in CCK-8 assays and clonogenicity capacity of CD73 in matched non-tumor, primary tumor, and re- (Fig. 2b and c). Cell cycle assays demonstrated that currence lesions exhibited an escalating pattern in most CD73 knockdown resulted in G0/G1 arrest in HCC cells patients (Fig. 1d). For patients with metastasis, the CD73 (Fig. 2d, Additional file 3: Figure S2C), while CD73 over- expression level of metastasis foci was also significantly expression accelerated the cell cycle in SMMC7721 cells elevated according to IHC results (n = 10, Fig. 1e). More- (Fig. 2d). Apoptosis assays further showed that CD73 over, we found patients with high CD73 were prone to knockdown greatly induced apoptosis in HCC cells, show incomplete tumor encapsulation, microvascular in- while CD73 overexpression protected SMMC7721 cells vasion, and poor differentiation (all P < 0.050, Table 1). from a serum-free environment (Fig. 2e, Additional file 3: Median TTR was significantly shorter in patients with Figure S2D). Transwell assays showed that CD73 knock- high CD73 expression (median 25.73 months vs. not down HCC cells had significantly fewer migrating and OE reached, P < 0.001, Fig. 1f), and the recurrence rate was invading cells than parental cells, while CD73 also higher in these patients (70.45% vs. 45.54%, Fig. 1g). SMMC7721 significantly higher migratory and invasive Similarly, patients with high CD73 expression had sig- capacities (Fig. 2f, Additional file 4: Figure S3A and B). nificantly shorter OS (median 38.88 months vs. not Moreover, wound healing assays confirmed the findings reached, P < 0.001) and a higher death rate (57.95% vs. of Transwell assays that knockdown of CD73 greatly 21.78%). Furthermore, CD73 retained its prognostic hindered HCC migration, while overexpression of CD73 prediction value in early-stage and low-AFP subgroups promoted migratory potentials (Fig. 2g, Additional file 4: (all P < 0.050, Additional file 2: Figure S1A–D). Notably, Figure S3C). Importantly, we also conducted CD73 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 4 of 17 Fig. 1 CD73 is overexpressed in HCC and indicates poor prognosis. a RT-PCR analysis of CD73 expression levels in 25 HCC tissues and paired non-tumor liver tissues. b Western blot (WB) analysis of CD73 expression levels in 25 HCC tissues and paired non-tumor liver tissues. n = 25; paired t test was used. c Representative immunohistochemistry (IHC) staining of CD73 expression, and comparison of the IHC score between cancerous and non-cancerous tissues (asterisk indicated P < 0.050, n = 189, non-parametric test was used). d WB analysis of CD73 expression levels in matched paratumor, primary-tumor, and recurrent tumor tissues from the same case (n = 8). e Representative IHC staining of CD73 expression in matched paratumor, primary-tumor, and metastatic tissues from the same case (left); comparisons of the IHC score were shown at the right panel (n = 10; non-parametric test was used). f Kaplan–Meier analysis of TTR (left) and overall survival (right) of HCC patients after curative resection according to CD73 expression level. g Recurrence (left) and death (right) incidences of HCC patients according to CD73 expression level. h Circulating tumor cell (CTC) load level of HCC patients according to CD73 expression level (left), and incidence of high CTC load in patients according to CD73 expression levels (right). (i) Expression level of CD73 in different HCC cell lines overexpression in HCCLM3 cells to confirm that the dif- migration, and invasion capacities in HCCLM3 cells ference observed above was not due to the variability (Additional file 5:Figure S4B–D). However, no significant between cell lines. CD73 expression was greatly in- alteration of apoptosis and cell cycle was observed after creased after plasmid transfection in HCCLM3 cells CD73 overexpression (Additional file 5: Figure S4E and F). (Additional file 5: Figure S4A). Similar to the effects of Taken together, our data indicated that CD73 promotes CD73 overexpression on SMMC7721 cells, forced expres- HCC cell proliferation, migration, and invasion capacities sion of CD73 also significantly promoted proliferation, and prevents apoptosis in vitro. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 5 of 17 Table 1 Correlation between clinicopathological parameters of Table 1 Correlation between clinicopathological parameters of patients enrolled patients enrolled (Continued) Clinical characteristics No. of patients CD73 low CD73 high P Clinical characteristics No. of patients CD73 low CD73 high P (n = 189) (n = 189) Age, years BCLC stage ≤ 50 79 35 44 0.033 0 + A 158 86 72 0.537 > 50 110 66 44 B+C 31 15 16 ALT alanine aminotransferase, AST aspartate transaminase, AFP α-fetoprotein, Sex BCLC Barcelona Clinic Liver Cancer Female 38 16 22 0.117 Bold indicated statistical significance Male 151 85 66 Child–Pugh score CD73 promotes HCC growth and metastasis in vivo A 179 96 83 0.823 Since shCD73-1 exhibited more efficiency in knocking down CD73 expression and resulting in more significant biological B10 5 5 alterations, further in vivo experiments and mechanism in- Liver cirrhosis vestigation were conducted using this shRNA. To minimize No 42 20 22 0.391 the impact of the immune system, NOG mice were used. Yes 147 81 66 Model mice were sacrificed after 6 weeks, and analyses of ALT, U/L liver orthotopic xenograft tumors showed that CD73 knock- ≤ 40 134 71 63 0.845 down greatly inhibited tumor growth, whereas CD73 over- expression promoted tumor growth in vivo (Fig. 2h). >40 55 30 25 Moreover, tumor tissues with high CD73 expression also ex- AST, U/L hibited high proliferating cell nuclear antigen levels (Add- ≤ 40 135 73 62 0.782 itional file 6: Figure S5A). Metastatic foci were examined >40 54 28 26 based on their tissue structure and cell morphology by two AFP, ng/ml senior histopathologists to evaluate the effects of CD73 on ≤ 400 140 74 66 0.786 HCC metastasis. Both the intrahepatic and pulmonary me- KD tastasis rates in mice with tumors generated from CD73 > 400 49 27 22 HCCLM3 cells were lower than in mice with tumors de- No. of tumors rived from parental HCCLM3 cells (Fig. 2i, Additional file 6: Single 169 92 77 0.424 Figure S5B and C). In contrast, both intrahepatic and pul- Multiple 20 9 11 monary metastasis rates were increased in tumors derived OE Tumor size, cm from CD73 SMMC7721 cells compared with parental ≤ 5 119 62 57 0.631 SMMC7721 cells. Collectively, our data demonstrated that CD73 promotes HCC proliferation and metastasis in vivo. > 5 70 39 31 Tumor encapsulation CD73 triggers epithelial-mesenchymal transition in HCC Complete 123 59 64 0.040 We further examined whether CD73 contributes to EMT. None 66 42 24 KD Through phalloidin staining, we found CD73 HCCLM3 Satellite lesion cells showed an epithelial cobblestone phenotype with less OE No 171 88 83 0.093 pseudopod, whereas CD73 SMMC7721 cells trans- formed into a spindle-like shape with more pseudopods Yes 18 13 5 (Fig. 3a). Moreover, CD73 knockdown resulted in an Macrovascular invasion epithelial-like molecular phenotype, while CD73 overex- No 175 94 81 0.784 pression induced a mesenchymal-like molecular phenotype Yes 14 7 7 according to RT-PCR and WB results (Fig. 3bandc).Con- Microvascular invasion sistently, immunofluorescence analysis confirmed these No 107 64 43 0.045 findings (Fig. 3d). Moreover, xenograft tumors generated KD from CD73 HCCLM3 cells exhibited an epithelial-like Yes 82 37 45 OE phenotype, while tumors derived from CD73 Edmondson stage SMMC7721 cells showed a mesenchymal-like phenotype I-II 126 74 52 0.039 according to IHC staining (Fig. 3e). We next examined the III-IV 63 27 36 correlation between CD73 and EMT in a clinical cohort. IHC staining in HCC consecutive sections indicated that Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 6 of 17 Table 2 Univariate Cox proportional regression analysis of factors associated with recurrence and overall survival Variables Recurrence Overall survival HR (95% CI) P HR (95% CI) P Age (> 50 years versus ≤ 50 years) 0.73 (0.50–1.07) 0.104 0.64 (0.41–1.02) 0.060 Sex (male versus female) 0.76 (0.48–1.21) 0.249 1.16 (0.69–1.95) 0.563 Liver cirrhosis (yes versus no) 1.52 (0.92–2.49) 0.099 1.11 (0.63–1.96) 0.712 ALT (> 40 U/L versus ≤ 40 U/L) 1.57 (1.06–2.34) 0.025 1.8 (0.85–2.24) 0.190 AST (> 40 U/L versus ≤40 U/L) 1.78 (1.20–2.64) 0.004 1.49 (0.92–2.41) 0.101 AFP (> 400 ng/ml versus ≤ 400 ng/ml) 1.86 (1.25–2.78) 0.002 1.49 (0.91–2.43) 0.117 No. of tumors (multi versus single) 1.81 (1.06–3.08) 0.030 1.57 (0.84–3.02) 0.157 Tumor size (> 5 cm versus ≤ 5 cm) 2.35 (1.60–3.43) < 0.001 1.59 (0.99–2.48) 0.056 Tumor encapsulation (none versus complete) 1.15 (0.78–1.69) 0.452 1.03 (0.64–1.67) 0.903 Satellite lesions (yes versus no) 1.67 (0.96–2.88) 0.069 0.78 (0.33–1.76) 0.528 Macrovascular invasion (yes versus no) 2.30 (1.28–4.12) 0.005 2.19 (1.13–4.28) 0.021 Microvascular invasion (yes versus no) 2.08 (1.42–3.04) < 0.001 1.91 (1.21–3.04) 0.006 Edmondson stage (III–IV versus I–II) 1.78 (1.21–2.61) 0.003 1.50 (0.94–2.39) 0.087 BCLC stage (B + C versus 0 + A) 1.98 (1.27–3.09) 0.003 1.72 (1.01–2.94) 0.045 ALBI grade (II versus I) 1.23 (0.80–1.89) 0.344 1.86 (1.12–3.09) 0.017 CD73 (high versus low) 2.33 (1.59–3.42) < 0.001 3.64 (2.20–6.01) < 0.001 Bold indicated statistical significance high CD73 expression was correlated with low level of was used and results showed that PI3K/AKT signaling ex- E-Cadherin but high level of N-Cadherin, and vice versa in hibited the greatest fold changes due to CD73 expression HCC tissue with low CD73 expression (Fig. 3f). Notably, manipulation (Fig. 4a). We therefore examined the PI3K/ correlation analysis revealed that high CD73 expression AKT pathway as a candidate critical signaling pathway. was positively correlated with N-Cadherin but negatively WB assays in HCC cells with CD73 knockdown con- correlated with E-Cadherin expression in clinical HCC firmed that phosphorylation levels of AKT and GSK3β,a samples (Fig. 3g). Collectively, our results suggested that direct substrate of pAKT, were reduced (Fig. 4b). In CD73 is a critical regulator of EMT in HCC. addition, expression of FOXO3a, which would undergo degradation due to AKT activation, was also increased CD73 promotes HCC progression and EMT by activating after CD73 knockdown. In contrast, pAKT and pGSK3β the PI3K-AKT signaling pathway levels were and increased, while FOXO3a was decreased To identify the underlying signaling of CD73 in HCC, with ectopic expression of CD73 in SMMC7721 cells Cignal Finder RTK signaling 10-Pathway Reporter Array (Fig. 4b). We next examined whether the effects of CD73 Table 3 Multivariate cox proportional regression analysis of factors associated with recurrence and overall survival Variables Recurrence Overall survival HR (95% CI) P HR (95% CI) P ALT (> 40 U/L versus ≤ 40 U/L) 1.24 (0.78–1.96) 0.369 N.A. N.A. AST (> 40 U/L versus ≤ 40 U/L) 1.46 (0.93–2.27) 0.099 N.A. N.A. AFP (> 400 ng/ml versus ≤ 400 ng/ml) 1.94 (1.29–2.94) 0.002 N.A. N.A. No. of tumors (multi versus single) 4.71 (1.28–17.42) 0.020 N.A. N.A. Tumor size (> 5 cm versus ≤ 5 cm) 2.01 (1.29–3.13) 0.002 N.A. N.A. Macrovascular invasion (yes versus no) 3.19 (0.83–12.29) 0.092 1.59 (0.95–2.67) 0.078 Microvascular invasion (yes versus no) 1.18 (0.75–1.85) 0.471 1.86 (1.16–2.96) 0.009 Edmondson stage (III–IV versus I–II) 1.33 (0.88–2.02) 0.182 0.98 (0.60–1.61) 0.932 ALBI grade (II versus I) 1.51 (0.63–3.60) 0.355 1.89 (1.09–3.18) 0.021 CD73 (high versus low) 2.82 (1.87–4.26) < 0.001 3.35 (2.01–5.52) < 0.001 Bold indicated statistical significance Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 7 of 17 Fig. 2 Biological functions of CD73 in HCC. a Efficiencies of CD73 knockdown and overexpression were validated by RT-PCR (left) and WB (right) assays. b Evaluation of the influence of CD73 on HCC proliferation by CCK-8. c Evaluations of the influence of CD73 on HCC proliferation by colony formation assays. d Evaluations of the influence of CD73 on cell cycle in HCC cells by flow cytometry. e Evaluations of the influence of CD73 on apoptosis in HCC cells by flow cytometry. f Evaluations of the influence of CD73 on migration and invasion activities of HCC cells by Transwell assays. g Evaluations of the influence of CD73 on migration activities of HCC cells by wound healing assays. h Establishment of the orthotopic xenograft model with NOG mice. Tumor volumes are shown in the right panel. i Incidences of intrahepatic (left) or lung (right) metastasis in indicated orthotopic xenograft model groups. All in vitro experiments were performed in triplicate; asterisk indicated P <0.050, and t tests were used Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 8 of 17 Fig. 3 CD73 triggers EMT in HCC. a Representative images of HCC cells with CD73 downregulation or overexpression. Cells were stained by FITC- phalloidin; scale bar 20 μm. b Expressions of EMT-related markers in HCC cells with CD73 downregulation or overexpression by RT-PCR assays. c Expressions of EMT-related markers in HCC cells with CD73 downregulation or overexpression by WB assays. d Expressions of CD73, E-Cadherin, N-Cadherin, and Vimentin in HCC cell lines with CD73 downregulation or overexpression by immunofluorescence staining; scale bar 200 μm. e Representative IHC images of CD73, E-Cadherin, and N-Cadherin expressions in tumor tissues derived from HCC cell lines as indicated. f Representative IHC images of CD73, E-Cadherin, and N-Cadherin expressions in consecutive tissue sections of clinical HCC samples. g Correlations between CD73 and E-Cadherin or N-Cadherin in clinical HCC samples were analyzed by Spearman’s rank correlation test. PCR assays were performed in triplicate. “N.S.” indicated not significant; asterisk indicated P <0.050, t tests were used on HCC cell activities were dependent on PI3K/AKT sig- HCCLM3 cells effectively attenuated the epithelial-induc- OE naling. Treatment of HCCLM3 cells with MK-2206, an tion effect, while inhibition of AKT signaling in CD73 AKT inhibitor, resulted in an epithelial-like molecular SMMC7721 cells prevented the mesenchymal-induction KD phenotype, which resembled the phenotype of CD73 effect (Fig. 4c). Also, biological experiments showed that HCCLM3 cells (Fig. 4c, Additional file 7: Figure S6A). AKT inhibition hindered proliferation and invasion cap- high Treatment of SMMC7721 cells with SC-79, an AKT acti- acities and promoted apoptosis in CD73 HCCLM3 vator, resulted in a mesenchymal-like phenotype that cells to levels observed in CD73 knockdown cells, while OE resembled the phenotype of CD73 SMMC7721 cells. AKT activation promoted proliferation and invasion cap- KD low Importantly, reactivation of AKT signaling in CD73 acities and prevented apoptosis in CD73 SMMC7721 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 9 of 17 Fig. 4 CD73 require its enzymatic activity and occur through adenosine receptor A2A to activate the PI3K/AKT pathway. a Cignal Finder RTK signaling 10-Pathway Reporter Array results demonstrate signaling changes in the indicated HCC cells. b Phosphorylation levels of AKT, GSK3β, and FOXO3a expression levels in the indicated HCC cells as determined by WB assays. c Expressions of EMT-related markers in the indicated HCC cells were detected by WB assays. d Proliferation in the indicated HCC cell lines was evaluated by CCK-8 and colony formation assays. “N.S.” indicated not significant; asterisk indicated P <0.050, and t tests were used. e Apoptosis in the indicated HCC cell lines was evaluated by flow cytometry assays; asterisk indicated P < 0.050 when compared with control group, and t tests were used. f Invasion in the indicated HCC cell lines was evaluated by Transwell assays; asterisk indicated P < 0.050 when compared with control group, and t tests were used. g Phosphorylation levels of AKT in HCCLM3 cells under different concentrations of APCP (left) or in SMMC7721 cells under different concentrations of adenosine treatments (right) as detectedby WB assays. h Phosphorylation levels of AKT in the indicated HCCLM3 (left) or SMMC7721 cells (right) as detected by WB assays. i Phosphorylation levels of AKT in HCCLM3 (left) or Hep3B cells (right) treated with antagonists targeting specific adenosine receptors (A1R, DPCPX; A2AR, KW6002; A2BR, CVT6883; A3R, Reversine). j Phosphorylation levels of AKT in indicated HCCLM3 (left) or SMMC7721 cells (right) as detected by WB assays. All in vitro experiments were performed in triplicate cells to levels observed in cells with CD73 overexpression CD73 function in HCC mainly depends on its enzymatic (Fig. 4d–f, Additional file 7:Figure S6B–D). Notably, AKT activity and occurs through adenosine receptor A2A inhibition abolished the effects of CD73 overexpression We next investigated whether the enzymatic activity of on HCC proliferation and invasion, whereas AKT activa- CD73 was required for its function in promoting progres- tion attenuated the inhibitory effects of CD73 knockdown sion and metastasis in HCC by using APCP, a CD73 en- high on HCC proliferation and invasion. These results indi- zyme activity inhibitor. APCP treatment of CD73 cated that CD73 promotes HCC progression and EMT by HCCLM3 cells reduced pAKT levels in a dose-dependent activating PI3K/AKT signaling. manner (Fig. 4g). Contrarily, exogenous adenosine Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 10 of 17 low treatment of CD73 SMMC7721 cells increased pAKT We next determined the role of Rap1 activation in in a dose-dependent manner. Notably, adenosine could re- AKT phosphorylation via manipulating Rap1 expression store the reduction of pAKT caused by CD73 knockdown, (Fig. 5b). In HCCLM3 cells, Rap1 knockdown dramatically while APCP treatment could partly abolish the increase of decreased pAKT levels, which could not be rescued by pAKT caused by CD73 overexpression (Fig. 4h). Function- CGS21680 treatment (Fig. 5c).However,transfectionof ally, APCP produced similar inhibitory effects on prolifer- Rap1G12V, a constitutively active mutant, attenuated the ation and invasion, induced an epithelial-like phenotype in reduction of pAKT level caused by CD73 knockdown or HCCLM3 cells as observed in CD73 knockdown cells, A2AR antagonist. Moreover, transfection of Rap1S17N, a and abolished the effects of CD73 overexpression (Add- dominant negative mutant, failed to restore the inhibitory itional file 8: Figure S7A-C), whereas adenosine promoted effects of Rap1 knockdown on pAKT levels (Fig. 5c). In proliferation and invasion, induced a mesenchymal-like SMMC7721 cells, transfection of Rap1G12V greatly in- phenotype in SMMC7721 cells as observed in CD73 over- creased pAKT levels, which could not be abolished by expression cells, and attenuated the effects of CD73 A2AR antagonist. However, CD73 overexpression or A2AR knockdown (Additional file 8:Figure S7D–F). Collectively, agonist showed no effects on pAKT level when Rap1 was our data demonstrated that CD73 functions in HCC re- knocked down. In addition, Rap1G12V expression rescued quire its enzymatic activity. the inhibition of AKT activation caused by Rap1 knock- We next searched for the adenosine receptor involved down (Fig. 5c). Importantly, through analyzing cytosolic in mediating the functions of CD73 in HCC using select- and plasma membrane fractions, we found that the majority ive inhibitors targeting adenosine A1 (DPCPX), A2A of Rap1 was recruited to the membrane after activation by (KW6002), A2B (CVT6883), and A3 receptors (Rever- the CD73-A2AR axis (Fig. 5d), which was consistent with a sine). Only KW6002 reduced AKT phosphorylation in previous study [28]. These results indicate that Rap1 serves high both CD73 HCC cell lines (Fig. 4i). Furthermore, as a key downstream regulator of the CD73-A2AR axis and KW6002 inhibited HCC proliferation and invasion to the that activation of Rap1 is essential for CD73-A2AR axis me- maximum extent and induced an epithelial-like phenotype diating AKT phosphorylation in HCC. (Additional file 9: Figure S8A–F). We thus focused on A2A receptor (A2AR) for further investigation. Inhibition The CD73-A2AR axis promotes Rap1-P110β interaction effects of A2AR on pAKT level could not be rescued by Previous studies reported that Rap1 could bind to Class I exogenous adenosine, while APCP failed to abolish the in- PI3Ks to promote PIP3 production, resulting in AKT acti- crease of pAKT resulting from using CGS21650, a select- vation [26, 27]. Thus we hypothesized that CD73-A2AR ive A2AR agonist (Fig. 4j). Importantly, A2AR agonist axis activates AKT signaling via promoting interaction be- could effectively restore the reduction of pAKT caused by tween Class I PI3K member and Rap1. First, to identify CD73 knockdown, while A2AR antagonist greatly abol- the key involved Class I PI3K member, we silenced two ished the increase of pAKT resulting from CD73 overex- major Class I PI3Ks, P110α, and P110β, in HCCLM3 cells pression (Fig. 4j). These data demonstrated that CD73 and found that P110β knockdown resulted in a dramatic exerts its function in HCC through A2AR. reduction of pAKT levels (Additional file 10:Figure S9A). Similar results were observed in Hep3B cells (Additional file 10: Figure S9B). These results suggested a The CD73-A2AR axis induces Rap1 activation and promotes critical role for P110β in CD73-mediated AKT activation its membrane recruitment in HCC. Next, Co-IP experiments in 293 T cells Adenosine A2AR belongs to the G-protein-coupled recep- co-transfected with HA-P110β and either Flag-Rap1G12V tor family, which is closely associated with the activation or Flag-Rap1S12N showed that only activated form Rap1 of Rap1 [25]. Previous studies showed that Rap1 is a crit- could interact with P110β (Fig. 5e). Notably, level of intra- ical regulator of PI3K activity [26, 27]. We therefore hy- cellular PIP3 and pAKT only elevated when Rap1 binding pothesized that the CD73-A2AR axis might activate PI3K/ to P110β, implying a requirement for this interaction in AKT signaling by inducing Rap1 activation. We first promoting catalytic activity of P110β (Additional file 10: assessed the effects of the CD73-A2AR axis on Rap1 acti- Figure S9C, D). vation. CD73 knockdown or A2AR antagonist greatly re- Above data demonstrated the significance of the inter- duced the level of Rap1-GTP, an active form of Rap1, action between Rap1 and P110β in regulating AKT acti- while A2AR agonist restored the reduction of Rap1-GTP vation. Next, the role of CD73-A2AR axis in promoting level caused by CD73 knockdown in HCCLM3 cells endogenous Rap1-P110β interaction was further vali- (Fig. 5a). Conversely, CD73 overexpression or A2AR dated by co-IP assays. CD73 knockdown or A2AR antag- agonist increased Rap1-GTP, while A2AR antagonist abol- onist greatly hindered the Rap1-P110β interaction, ished the increase of Rap1-GTP level resulting from CD73 whereas A2AR agonist successfully rescued the inhib- overexpression. ition of Rap1-P110β interaction caused by CD73 Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 11 of 17 Fig. 5 CD73-A2AR axis activates PI3K/AKT signaling by inducing Rap1-mediated membrane localization of P110β. a Rap1 activation status was detected in the indicated HCC cells. b Rap1 expression was silenced via using shRNA in HCCLM3 and SMMC7721 cells and was confirmed by WB assays. c pAKT levels in the indicated HCCLM3 (upper) or SMMC7721 cells (lower) by WB assays. d Cytosolic and plasma membrane fractions were + + isolated from the indicated HCCLM3 (left) or SMMC7721 cells (right), followed by WB assays with antibody against Rap1. Na -K ATPase was used as an internal control for membrane fractions, and β-actin was used as an internal control for cytosolic fractions. e Co-immunoprecipitation of HA-P110β and Flag-Rap1G12V or Flag-Rap1S12N in SMMC7721 cells. f Effects of the CD73-A2AR axis on promoting Rap1-P110β interaction in co-IP assays. g Cytosolic and plasma membrane fractions were isolated from the indicated HCCLM3 (left) or SMMC7721 cells (right), followed by WB assays with antibody against P110β. pAKT level and PIP3 concentrations were also detected. ELISA experiments for PIP3 determinationwereconductedintriplicate. h Immunofluorescence staining of Rap1 (green) and P110β (red) with specific antibodies in indicated HCCLM3 (left) or SMMC7721 cells (right). DAPI was used for nuclear staining; scale bar 50 μm. Asterisk indicated P < 0.050 when compared with control group, t tests were used knockdown in HCCLM3 cells (Fig. 5f). Contrarily, CD73 the effects of CD73 overexpression on promoting overexpression or A2AR agonist promoted the Rap1-P110β Rap1-P110β binding in SMMC7721 cells. Our findings in- interaction, while A2AR antagonist dramatically inhibited dicate that CD73-A2AR serves as an upstream promoter of Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 12 of 17 the Rap1-P110β interaction, resulting in PIP3 production (Fig. 6c). Moreover, miR-193b expression was signifi- and AKT phosphorylation. cantly downregulated in HCC tissues compared to paratumor tissues (Fig. 6d) and reversely correlated The CD73-A2AR axis promotes Rap1-mediated membrane with CD73 mRNA in clinical HCC samples (r = − 0.83, localization of P110β and PIP3 production P <0.001, Fig. 6e). Functional assays demonstrated that As our results showed that activated Rap1 was recruited transfection of miR-193b could significantly inhibit HCC to the membrane, and Rap1 binds to P110β, we further proliferation, migration, and invasion, and these inhibitory evaluated whether CD73-A2AR could induce P110β effects could be rescued by CD73 re-expression (Fig. 6f–h, membrane localization, a critical process for its catalytic Additional file 11: Figure S10C and D). Additionally, in- function [29], in a Rap1-dependent manner. Through duction of epithelial-like phenotype and downregulation analyzing cytosolic and plasma membrane fractions by of phosphorylation of AKT caused by miR-193b mimic WB assays, we found that, similar to the effect of Rap1 transfection could also be reversed by re-expression of knockdown, CD73 knockdown or A2AR antagonist CD73 (Fig. 6i). Taken together, our data identified greatly reduced the plasma membrane localization of miR-193b as a novel negative regulator of CD73 expres- P110β in HCCLM3 cells, and this could be rescued by sion in HCC. Rap1G12V expression (Fig. 5g). On the other hand, con- sistent with the results of Rap1G12V transfection, CD73 Co-inhibition of CD73 and A2AR shows therapeutic overexpression or A2AR agonist promoted plasma mem- potential in vitro and in immunodeficient mice brane localization of P110β in SMMC7721 cells, and this We examined whether targeting CD73 in combination was greatly abolished by Rap1 knockdown. Notably, with A2AR inhibition would inhibit HCC growth inde- higher PIP3 concentrations and pAKT levels were ob- pendent of the immune system. Colony formation assays served when the majority of P110β localized on plasma showed that co-treatment with inhibitors against CD73 membrane (Fig. 5g), indicating abundant catalytic activ- and A2AR dramatically reduced cell proliferation in high ity of P110β under these circumstances. Assays were fur- CD73 HCCLM3 cells, while inhibiting CD73 or ther performed and results further confirmed the A2AR alone showed moderate inhibition effects (Fig. 7a). membranous co-localization of Rap1-P110β due to Moreover, co-targeting strategy induced more apoptosis CD73-A2AR axis regulation, which was consistent with in HCCLM3 cells than targeting CD73 or A2AR alone the results of WB assays (Fig. 5h). These data demon- (Fig. 7b). Targeting CD73 or A2AR alone in NOG mice strated that the CD73-A2AR axis activates AKT by indu- caused moderate suppression of tumor growth in vivo. cing Rap1-mediated membrane localization of P110β. However, dramatic and more durable responses were ob- served in mice co-treated with both CD73 and A2AR in- CD73 is negatively regulated by miR-193b hibitors (Fig. 7c). Similarly, Targeting CD73 or A2AR Previous studies revealed miRNA was responsible for alone caused moderate suppression of lung metastasis, the dysregulation of CD73 expression [30, 31]. We fur- while co-targeting strategy exerted a synergistic effect, ther investigated the potential upstream miRNA of resulting in maximal suppression of lung metastasis. CD73 in HCC. Seven microRNAs were predicted by two (Fig. 7d). Our findings revealed co-targeting CD73 and bioinformatics algorithms (TargetScan and miRanda) to A2AR exhibited potentials for preventing HCC progres- be potential upstream regulators of CD73 according to sion and metastasis. StarBase database [32] (Additional file 11: Figure S10A). Among these potential miRNAs, two homologous miR- Discussion NAs, miR-193a and miR-193b were selected for further CD73 plays a vital role in adenosinergic signaling by cata- study, as previous studies reported them as tumor sup- lyzing AMP into adenosine [10]. Hence, CD73 has been pressors in HCC [33, 34]. A putative binding site of mainly studied for its immunosuppression functions and miR-193a/miR-193b and 3′UTR of CD73 was identified was identified as a novel immune checkpoint target with (Fig. 6a). Transfection of miR-193b mimics significantly promising potential for suppressing tumor development high reduced CD73 expression in two CD73 cell lines [19, 35]. Meanwhile, studies also showed CD73 promotes (Fig. 6b). However, miR-193a mimics failed to exert tumor growth in an immune-independent manner [16, 17, similar effects (Additional file 11: Figure S10B), and 30, 31]. Here, functional experiments demonstrated that thus, miR-193b was identified as the potential regulator CD73 could promote HCC progression and metastasis. for CD73 expression in our study. To validate its regula- We also confirmed the clinical significance of CD73 as an tory role, dual luciferase reporter assays were firstly per- independent prognostic indicator of TTR and OS for formed and results revealed that miR-193b markedly HCC patients after curative resection. Moreover, CD73 inhibited the activity of a luciferase vector containing retained significant prognostic prediction value in the wild-type CD73 3′UTR, but not the mutant 3′UTR early-HCC and AFP-low patients, whose clinical outcomes Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 13 of 17 Fig. 6 CD73 is regulated by miR-193b. a Putative binding site of miR-193a/miR-193b and CD73. b Expression of CD73 in Hep3B and HCCLM3 cells after miR-193b mimic transfection as detected by RT-PCR and WB assays; asterisk indicated P < 0.050 when compared with parental group, and t tests were used. c Effect of miR-193b on the activity of CD73 mRNA 3′UTR was evaluated by luciferase reporter assays. d Comparison of miR-193b expression level between paratumor and tumor tissues determined by RT-PCR assays; asterisk indicated P < 0.050, n = 25, and t tests were used. e Correlation of miR-193b and CD73 mRNA in clinical HCC samples was analyzed by Spearman’s rank correlation test, n = 25. f Evaluations of proliferation capacities of indicated HCCLM3 (left) and Hep3B cells (right) by CCK-8 assays; “N.S.” indicated not significant; asterisk indicated P < 0.050, and t tests were used. g Evaluations of migration capacities of indicated HCCLM3 (left) and Hep3B cells (right) by wound healing assays; asterisk indicated P < 0.050, and t tests were used. h Evaluations of invasion capacities of indicated HCCLM3 (left) and Hep3B cells (right) by wound healing assays; asterisk indicated P < 0.050, t tests were used. i Evaluations of EMT-related genes and pAKT expression levels of indicated HCCLM3 (left) and Hep3B cells (right) by WB assays. In vitro experiments were performed in triplicate were difficult to predict by conventional indexes [24]. Our [36]. Previous studies reported CD73 as a critical regulator findings collaborated with the prognostic significance of for the maintenance of mesenchymal traits in mesenchymal CD73 in TCGA database (Additional file 12: Figure S11) stem cells and ovarian cancer cells [16, 37, 38]. In the and indicated that it might serve as an indicator to identify present study, we showed that CD73 expression was essen- HCC patients at high risk of recurrence that require fur- tial for maintaining the mesenchymal-like phenotype in ther interventions in addition to resection to improve HCC. Clinical samples confirmed our findings, showing their prognosis. that CD73 expression significantly correlated with EMT sta- EMT plays an important role in facilitating HCC metasta- tus. Therefore, we provide the first evidence demonstrating sis, despite its controversial function during carcinogenesis that CD73 serves as a trigger for EMT in HCC. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 14 of 17 Fig. 7 Co-targeting CD73 and A2AR shows therapeutic potential in HCC. a Colony formation assays in the indicated HCCLM3 cells. b Apoptosis assays in the indicated HCCLM3 cells. c Tumor growth kinetics in NOG model mice subcutaneously implanted with HCCLM3 cells treated as indicated. d Incidences of lung metastasis of different administered groups versus controls in NOG model mice intravenously injected with HCCLM3 cells. Asterisk indicated P < 0.050 when compared with controls; number sign indicated P < 0.050 when compared with APCP/KW5002 group alone; all in vitro experiments were performed in triplicate; and t tests were used Previous studies reported that CD73 exerted its functions speculated that co-targeting CD73 and A2AR might achieve by both enzymatic-dependent and enzymatic-independent satisfactory treatment effects in HCC. Indeed, CD73 inhib- ways [15]. Here, we found that CD73 exerts its tumor pro- ition in combination with A2AR blockade resulted in signifi- motion functions mainly through its enzymatic activity due cantly decreased tumor growth as well as metastasis to the significant changes in HCC cells after APCP or ad- compared with controls. Since our findings were observed enosine treatment. However, it should be noted that APCP in immunodeficient mouse models, we hypothesize that this showed a weaker effect on cell proliferation and invasion co-inhibition strategy might achieve more promising treat- than CD73, although this difference did not reach a statis- ment response by not only blocking progressive traits of tical significance. Similar results were observed when HCC but also restoring anti-tumor immune function within comparing adenosine with CD73 overexpression. These the tumor microenvironment. findings suggested the possibility of other regulatory Abnormal AKT activation is a hallmark of tumor pro- mechanisms of CD73 beyond producing adenosine. gression in various cancers, including HCC [39]. In the Thus, our future work will examine the non-enzymatic present study, Alterations of pGSK3β and pAKT levels functions of CD73 to elucidate the regulatory network were observed. More importantly, FOXO3a, a of CD73 in HCC. well-established tumor suppressor in HCC according to A2AR is a critical receptor that is responsible for the func- previous study [40], was also greatly repressed due to tion of CD73 in cancer immunity, and CD73-expressing CD73 expression, which served as solid evidence for AKT cancer cells are prone to form metastasis loci via activation activation. Our findings were consistent with previous of A2AR, which results in exhaustion of natural killer cells studies that CD73 activated AKT in breast and lung can- within the tumor microenvironment [8, 13, 14, 19]. Here, cer cells [17, 30]. However, the underlying mechanism of we provided the first evidence that A2AR serves as the key how CD73 triggered AKT activation had not been eluci- downstream mediator of CD73 to regulate the progressive dated clearly by previous studies. Here, we showed that phenotype of HCC, and inhibition of A2AR dramatically the CD73-A2AR axis served as a “switch” for Rap1 activa- abolished the effects of CD73 on HCC cells. Our data tion, thus resulting in membrane localization of P110β strongly demonstrate the non-immunosuppressive effects of and PIP3 production, which are critical steps for AKT ac- A2AR, thus expanding our understanding of the function of tivation. Rap1 is a central regulator of cell adhesion and A2AR in HCC development. Based on these findings, we motility, and aberrant activation of Rap1 can result in Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 15 of 17 carcinogenesis and tumor progression [27]. Importantly, Additional files Rap1 and Ras share similar binding partners, including PI3K [41]. Our study not only confirmed the importance Additional file 1: Supplementary methods and materials. (DOCX 28 kb) of Rap1 in HCC progression and metastasis, but also iden- Additional file 2: Figure S1. Prognostic value of CD73 in subgroups of HCC patients. (A) Kaplan–Meier analysis of TTR of early-stage (BCLC 0+A) tified Rap1 as a key downstream responder of patients after curative resection according to CD73 expression level. (B) CD73-A2AR that recruits P110β to the plasma mem- Kaplan–Meier analysis of OS of early-stage (BCLC 0+A) patients after cura- brane. It was reported that cAMP acted as a key down- tive resection according to CD73 expression level. (C) Kaplan–Meier ana- lysis of TTR of curative resection according to CD73 expression level. (D) stream molecule of GPCR and could induce Rap1 Kaplan–Meier analysis of OS of low-AFP patients after curative resection activation [42]. Based on our Cignal pathway screening re- according to CD73 expression level. (TIF 1853 kb) sults, we also found that cAMP/PKA signaling exhibited Additional file 3: Figure S2. Function of CD73 in Hep3B cell line. (A) significant changes due to CD73 alteration (Fig. 4a). Efficiencies of CD73 knockdown in Hep3B cells were evaluated by WB assays. (B) Effects of CD73 knockdown on proliferation in Hep3B cells Hence, A2AR might mediate Rap1 activation through pro- were evaluated by CCK-8 and colony formation assays. (C) Effects of moting cAMP production, and detailed mechanism is cur- CD73 knockdown on cell cycle in Hep3B cells were evaluated by flow rently ongoing in our lab. Previous reports also showed cytometry assays. (D) Effects of CD73 knockdown on apoptosis in Hep3B cells were evaluated by flow cytometry assays. (E) Effects of CD73 knock- that CD73 was positively correlated with EGFR [22], and down on migration and invasion in Hep3B cells were evaluated by future studies should also examine their potential inter- Transwell assays. (F) Effects of CD73 knockdown on migration were action in HCC. validated by wound healing assays. (G) Effects of CD73 knockdown on in vivo tumor growth. “N.S.” indicated not significant; Asterisk indicated Regulatory network of miRNA-mRNA interaction was P < 0.050, all in vitro experiments were performed in triplicate, t tests important for HCC progression. miR-193a was reported were used. (TIF 10457 kb) as a tumor suppressor and could be found in HCC Additional file 4: Figure S3. Effects of CD73 on cell motility. (A) cell-derived exosomes [33]. However, miR-193b did not Representative images of Transwell assays conducted with control or CD73 knockdown (KD) HCCLM3 cells. (B) Representative images of serve as a significant prognostic indicator from Transwell assays conducted with control or CD73 overexpression (OE) HCC-derived exosomes. The major explanation for this SMMC7721 cells. (C) Representative images of wound healing assays might be the different expression regulatory mechanism conducted with control or CD73-KD HCCLM3 cells. (D) Representative images of wound healing assays conducted with control or CD73-OE of these two miRNAs in HCC: low miR-193b expression SMMC7721 cells. “N.S.” indicated not significant; Asterisk indicated P <0.050, might due to inhibition of transcription while low all in vitro experiments were performed in triplicate, t tests were used. miR-193a level might result from selective packaging and (TIF 13202 kb) secreting through exosomes. Therefore, it was rational Additional file 5: Figure S4. Biological effects of CD73 overexpression in HCCLM3 cells. (A) Efficiencies of overexpression of CD73 in HCCLM3 that miR-193b could rarely be found in HCC-derived exo- cells were validated by RT-PCR (left) and WB assays (right). (B) Effects of somes. Meanwhile, CD73 expression showed little re- CD73 overexpression on proliferation were evaluated by CCK-8 (left) and sponse to miR-193a mimics transfection in our study. We colony formation (right) assay. (C) Effects of CD73 overexpression on migration and invasion were evaluated by Transwell assays. (D) Effects of speculated this phenomenon might result from different CD73 overexpression on migration were validated by wound healing clearance mechanism of exogenous miRNA between assays. (E) Effects of CD73 overexpression on cell survival were evaluated miR-193a and miR-193b. However, our findings demon- by flow cytometry. (F) Effects of CD73 overexpression on cell cycle were evaluated by flow cytometry. (G) Effects of CD73 overexpression on strated that miR-193b was the major negative regulator of E-cadherin, N-Cadherin, and pAKT level were evaluated by WB assays. CD73 in HCC, and our future work will aim at clarifying Asterisk indicated P < 0.050, all in vitro experiments were performed in the underlying mechanism of inhibitory difference be- triplicate, t tests were used. (TIF 5605 kb) tween miR-193a and miR-193b. Additional file 6: Figure S5. In vivo function of CD73 in HCC. (A) Representative IHC images of PCNA expression level in indicated HCCLM3 (left) or SMMC7721 cells (right). (B) Representative image of intrahepatic metastasis. (C) Representative images of lung metastasis in mice implanted Conclusions with HCCLM3 or SMMC7721 cells. (TIF 10892 kb) Our study demonstrated that CD73 was an independent Additional file 7: Figure S6. Role of PI3K/AKT signaling in CD73 mediating HCC progression. (A) Expressions of EMT-related markers in the prognostic indicator for HCC. Functional experiments indicated HCC cells were detected by RT-PCR assays; asterisk indicated established the essential role of CD73 in promoting HCC P < 0.050, experiments were performed in triplicate, and t tests were used. progression and metastasis through PI3K/AKT signaling (B) Representative images of colony formation assays of the indicated HCCLM3 (upper) and SMMC7721 cells (lower). (C) Representative results by inducing Rap1-mediated membrane localization of of apoptosis assays in the indicated HCCLM3 (upper) and SMMC7721 cells P110β. We also identified miR-193b as a potential up- (lower). (D) Representative images of Transwell assays of indicated stream regulator of CD73 in HCC. Notably, co-targeting HCCLM3 (upper) and SMMC7721 cells (lower). (TIF 9692 kb) CD73 and A2AR achieved satisfactory inhibitory effects Additional file 8: Figure S7. CD73 function in HCC depends on its enzymatic activity. (A) Proliferation capacity in the indicated HCCLM3 on HCC in vitro and vivo. Together these results demon- detected by CCK-8 assays. (B) Invasion capacity in the indicated HCCLM3 strate the importance of CD73 in HCC in addition to its detected by Transwell assays. (C) Expressions of EMT-related markers in immunosuppressive functions and revealed that a CD73 indicated HCCLM3 cells were detected by RT-PCR. (D) Proliferation capacity in the indicated SMMC7721 detected by CCK-8 assays. (E) targeting strategy may be a promising novel therapeutic Invasion capacity in the indicated SMMC7721 detected by Transwell strategy for future HCC management. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 16 of 17 Scheme (20130071140008), the Projects from Shanghai Science and assays. (F) Expressions of EMT-related markers of indicated SMMC7721 Technology Commission (14DZ1940300, 14411970200), The Strategic Priority cells were detected by RT-PCR. “N.S.” indicated not significant; asterisk Research Program of the Chinese Academy of Science (XDA12020105). indicated P < 0.050, all in vitro experiments were performed in triplicate, Bei-Li Wang was supported by the Projects from Shanghai Science and and t tests were used. (TIF 4551 kb) Technology Commission (16411952100). Additional file 9: Figure S8. CD73 function in HCC occurs through Yun-Fan Sun was supported by National Natural Science Foundation of adenosine receptor A2A. (A) Proliferation capacity in HCCLM3 cells China (81602543) and the Sailing Program from the Shanghai and Technology treated with indicated adenosine receptor antagonists detected by CCK-8 Commission (16YF1401400). assays. (B) Invasion capacity in HCCLM3 cells treated with indicated Bo Hu was supported by National Natural Science Foundation of China adenosine receptor antagonists detected by Transwell assays. (C) Expressions (8180101167 and 31771472) and Zhongshan Hospital Science Foundation of EMT-related markers in HCCLM3 cells treated with indicated adenosine (2016ZSQN30, 2017ZSQN18, 2017ZSYQ26). receptor antagonists were detected by RT-PCR. (D) Proliferation capacity in Hep3B cells treated with indicated adenosine receptor antagonists detected Availability of data and materials by CCK-8 assays. (E) Invasion capacity in Hep3B cells treated with indicated All data generated or analyzed during this study are included in this adenosine receptor antagonists detected by Transwell assays. (F) Expressions published article and its supplementary information files. of EMT-related markers of Hep3B cells treated with indicated adenosine receptor antagonists were detected by RT-PCR. Asterisk indicated P <0.050, Authors’ contributions all in vitro experiments were performed in triplicate, and t tests were used. XLM, XRY, and BH contributed to the study design. XLM, BH, WJY, LHL, HW, (TIF 5733 kb) YZ, CYZ, ALJ, and YFS contributed to the acquisition, analysis, and interpretation Additional file 10: Figure S9. Role of Rap1 activation in AKT of the data. XLM, BH, and XRY contributed to the drafting of the manuscript. phosphorylation. (A) Expressions of P110α, P110β, pAKT, and AKT in HCCLM3 JZ, BSP, WG, XRY, and JF contributed to the critical revision of the manuscript. cells were detected by WB assays. (B) Expressions of P110α, P110β, pAKT, and All authors read and approved the final manuscript. AKT in Hep3B cells were detected by WB assays. (C) Expressions of pAKT and AKT in 293T transfected with indicated plasmids were detected by WB assays. Ethics approval and consent to participate (D) Cellular PIP2 and PIP3 levels of 293T transfected with indicated plasmids Present study was performed in accordance with the 1975 Declaration of were detected by ELISA assays. “N.S.” indicated not significant; asterisk Helsinki. Approval for the use of human subjects was obtained from the indicated P < 0.050, experiments were performed in triplicate, and t tests research ethics committee of Zhongshan Hospital, and informed consent were used. (TIF 725 kb) was obtained from each individual enrolled in this study. Additional file 11: Figure S10. miR-193b served as an upstream negative regulator for CD73 expression in HCC. (A) Prediction results of microRNAs Consent for publication that potentially regulate CD73 according to StarBase 2.0. (B) Effects of Not applicable miR-193a mimics transfection on CD73 expression in HCC cells were evaluated by RT-PCR and WB assays. (C) Representative images of Transwell Competing interests assays in indicated HCCLM3 (upper) and Hep3B (lower) cells. (D) Migration The authors declare that they have no competing interests. evaluation via wound healing assays of miR-193b mimic-transfected HCC cells with or without CD73 overexpression. All in vitro experiments were performed in triplicate, and t tests were used. (TIF 6462 kb) Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published Additional file 12: Figure S11 Clinical significance of CD73 in HCC maps and institutional affiliations. according to TCGA database. Data was collected from The Human Protein Atlas (https://www.proteinatlas.org/ENSG00000135318-NT5E/ Author details pathology/tissue/liver+cancer). (TIF 378 kb) Department of Laboratory Medicine, Zhongshan Hospital, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, People’s Republic of China. Abbreviations Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, A2AR: Adenosine A2A receptor; BCLC: Barcelona Clinic Liver Cancer; co- Fudan University, Key Laboratory of Carcinogenesis and Cancer Invasion, IP: Co-immunoprecipitation; CTC: Circulating tumor cell; DMEM: Dulbecco’s Ministry of Education, Shanghai 200032, People’s Republic of China. Liver modified Eagle’s medium; EMT: Epithelial-mesenchymal transition; Cancer Institute, Fudan University, 136 Yi Xue Yuan Road, Shanghai 200032, HCC: Hepatocellular carcinoma; HR: Hazard ratio; NOG: NOD/SCID/γc(null); People’s Republic of China. OS: Overall survival; shRNA: Short hairpin RNA; TTR: Time to recurrence; WB: Western blot Received: 21 December 2018 Accepted: 27 March 2019 Acknowledgements We thank Prof Ji Yuan and Dr. Lu Shao-Hua from the Department of Pathology, References Zhongshan Hospital for their help in pathological diagnosis. Also, we thank the 1. Llovet JM, Montal R, Sia D, Finn RS. Molecular therapies and precision medicine participating patients for the source of clinical blood samples. for hepatocellular carcinoma. Nat Rev Clin Oncol. 2018;15:599–616. 2. Reig M, Da FL, Faivre S. New trials and results in systemic treatment of HCC. Funding J Hepatol. 2018;69:525–33. Wei Guo was supported by the National Natural Science Foundation of China 3. Dutta R, Mahato RI. Recent advances in hepatocellular carcinoma therapy. (81772263 and 81572064) and Key Developing Disciplines of Shanghai Pharmacol Ther. 2017;173:106–17. Municipal Commission of Health and Family Planning (2015ZB0201). 4. Galle PR, Tovoli F, Foerster F, Worns MA, Cucchetti A, Bolondi L. The Xin-Rong Yang was supported by the National Natural Science Foundation treatment of intermediate stage tumours beyond TACE: from surgery to of China (81672839, 81472676 and 81872355), National Key Research and systemic therapy. J Hepatol. 2017;67:173–83. Development Program of China (2016YFF0101405), the project from Shanghai 5. Huo X, Han S, Wu G, Latchoumanin O, Zhou G, Hebbard L, et al. Science and Technology Commission (14DZ1940302, 1411970200, 14140902301), Dysregulated long noncoding RNAs (lncRNAs) in hepatocellular carcinoma: the Strategic Priority Research Program of the Chinese Academy of Sciences implications for tumorigenesis, disease progression, and liver cancer stem (XDA12020103). cells. Mol Cancer. 2017;16:165. Jia Fan was supported by the National High Technology Research and 6. Liu FF, Liu YN, Chen Z. Tim-3 expression and its role in hepatocellular Development Program (863 Program) of China (2015AA020401), the State carcinoma. J Hematol Oncol. 2018;11:126. Key Program of National Natural Science of China (81530077, and 81772551), 7. Zhou SL, Hu ZQ, Zhou ZJ, Dai Z, Wang Z, Cao Y, et al. miR-28-5p-IL-34- Specialized Research Fund for the Doctoral Program of Higher Education macrophage feedback loop modulates hepatocellular carcinoma metastasis. and Research Grants Council Earmarked Research Grants Joint Research Hepatology. 2016;63:1560–75. Ma et al. Journal of Hematology & Oncology (2019) 12:37 Page 17 of 17 8. Antonioli L, Pacher P, Vizi ES, Hasko G. CD39 and CD73 in immunity and 35. Kordass T, Osen W, Eichmuller SB. Controlling the immune suppressor: inflammation. Trends Mol Med. 2013;19:355–67. transcription factors and microRNAs regulating CD73/NT5E. Front Immunol. 9. Allard D, Allard B, Gaudreau PO, Chrobak P, Stagg J. CD73-adenosine: a next- 2018;9:813. generation target in immuno-oncology. Immunotherapy. 2016;8:145–63. 36. Jayachandran A, Dhungel B, Steel JC. Epithelial-to-mesenchymal plasticity of 10. Allard B, Longhi MS, Robson SC, Stagg J. The ectonucleotidases CD39 and cancer stem cells: therapeutic targets in hepatocellular carcinoma. J CD73: novel checkpoint inhibitor targets. Immunol Rev. 2017;276:121–44. Hematol Oncol. 2016;9:74. 37. Breitbach M, Kimura K, Luis TC, Fuegemann CJ, Woll PS, Hesse M, et al. In 11. Gourdin N, Bossennec M, Rodriguez C, Vigano S, Machon C, Jandus C, et al. vivo labeling by CD73 marks multipotent stromal cells and highlights Autocrine adenosine regulates tumor polyfunctional CD73(+)CD4(+) effector endothelial heterogeneity in the bone marrow niche. Cell Stem Cell. 2018; T cells devoid of immune checkpoints. Cancer Res. 2018;78:3604–18. 22:262–76. 12. Li L, Wang L, Li J, Fan Z, Yang L, Zhang Z, et al. Metformin-induced 38. Suto EG, Mabuchi Y, Suzuki N, Suzuki K, Ogata Y, Taguchi M, et al. Prospectively reduction of CD39 and CD73 blocks myeloid-derived suppressor cell activity isolated mesenchymal stem/stromal cells are enriched in the CD73(+) in patients with ovarian cancer. Cancer Res. 2018;78:1779–91. population and exhibit efficacy after transplantation. Sci Rep. 2017;7:4838. 13. Allard B, Pommey S, Smyth MJ, Stagg J. Targeting CD73 enhances the 39. Zhou Q, Lui VW, Yeo W. Targeting the PI3K/Akt/mTOR pathway in antitumor activity of anti-PD-1 and anti-CTLA-4 mAbs. Clin Cancer Res. 2013; hepatocellular carcinoma. Future Oncol. 2011;7:1149–67. 19:5626–35. 40. Cao MQ, You AB, Zhu XD, Zhang W, Zhang YY, Zhang SZ, et al. miR-182-5p 14. Young A, Ngiow SF, Barkauskas DS, Sult E, Hay C, Blake SJ, et al. Co- promotes hepatocellular carcinoma progression by repressing FOXO3a. J inhibition of CD73 and A2AR adenosine signaling improves anti-tumor Hematol Oncol. 2018;11:12. immune responses. Cancer Cell. 2016;30:391–403. 41. Frische EW, Zwartkruis FJ. Rap1, a mercenary among the Ras-like GTPases. 15. Gao ZW, Dong K, Zhang HZ. The roles of CD73 in cancer. Biomed Res Int. Dev Biol. 2010;340:1–9. 2014;2014:460654. 42. Onodera Y, Nam JM, Bissell MJ. Increased sugar uptake promotes oncogenesis 16. Lupia M, Angiolini F, Bertalot G, Freddi S, Sachsenmeier KF, Chisci E, et al. via EPAC/RAP1 and O-GlcNAc pathway. J Clin Invest. 2014;124:367–84. CD73 regulates stemness and epithelial-mesenchymal transition in ovarian cancer-initiating cells. Stem Cell Rep. 2018;10:1412–25. 17. Yu J, Wang X, Lu Q, Wang J, Li L, Liao X, et al. Extracellular 5′-nucleotidase (CD73) promotes human breast cancer cells growth through AKT/GSK-3beta/ beta-catenin/cyclinD1 signaling pathway. Int J Cancer. 2018;142:959–67. 18. Allard B, Turcotte M, Spring K, Pommey S, Royal I, Stagg J. Anti-CD73 therapy impairs tumor angiogenesis. Int J Cancer. 2014;134:1466–73. 19. Beavis PA, Divisekera U, Paget C, Chow MT, John LB, Devaud C, et al. Blockade of A2A receptors potently suppresses the metastasis of CD73+ tumors. Proc Natl Acad Sci U S A. 2013;110:14711–6. 20. Buisseret L, Pommey S, Allard B, Garaud S, Bergeron M, Cousineau I, et al. Clinical significance of CD73 in triple-negative breast cancer: multiplex analysis of a phase III clinical trial. Ann Oncol. 2018;29:1056–62. 21. Turcotte M, Spring K, Pommey S, Chouinard G, Cousineau I, George J, et al. CD73 is associated with poor prognosis in high-grade serous ovarian cancer. Cancer Res. 2015;75:4494–503. 22. Shali S, Yu J, Zhang X, Wang X, Jin Y, Su M, et al. Ecto-5′-nucleotidase (CD73) is a potential target of hepatocellular carcinoma. J Cell Physiol. 2018. https://doi.org/10.1002/jcp.27694. 23. Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis. 2010;30:61–74. 24. Guo W, Sun YF, Shen MN, Ma XL, Wu J, Zhang CY, et al. Circulating tumor cells with stem-like phenotypes for diagnosis, prognosis, and therapeutic response evaluation in hepatocellular carcinoma. Clin Cancer Res. 2018;24:2203–13. 25. Cha I, Lee SH, Jeon TJ. Chemoattractant-mediated Rap1 activation requires GPCR/G proteins. Mol Cells. 2010;30:563–7. 26. Kortholt A, Bolourani P, Rehmann H, Keizer-Gunnink I, Weeks G, Wittinghofer A, et al. A Rap/phosphatidylinositol 3-kinase pathway controls pseudopod formation [corrected]. Mol Biol Cell. 2010;21:936–45. 27. Shah S, Brock EJ, Ji K, Mattingly RR. Ras and Rap1: a tale of two GTPases. Semin Cancer Biol. 2018. https://doi.org/10.1016/j.semcancer.2018.03.005. 28. Xiong Y, Ye C, Yang N, Li M, Liu H. Ubc9 binds to ADAP and is required for Rap1 membrane recruitment, Rac1 activation, and integrin-mediated T cell adhesion. J Immunol. 2017;199:4142–54. 29. Thorpe LM, Yuzugullu H, Zhao JJ. PI3K in cancer: divergent roles of isoforms, modes of activation and therapeutic targeting. Nat Rev Cancer. 2015;15:7–24. 30. Zhu J, Zeng Y, Li W, Qin H, Lei Z, Shen D, et al. CD73/NT5E is a target of miR-30a-5p and plays an important role in the pathogenesis of non-small cell lung cancer. Mol Cancer. 2017;16:34. 31. Xie M, Qin H, Luo Q, Huang Q, He X, Yang Z, et al. MicroRNA-30a regulates cell proliferation and tumor growth of colorectal cancer by targeting CD73. BMC Cancer. 2017;17:305. 32. Li JH, Liu S, Zhou H, Qu LH, Yang JH. starBase v2.0: decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP- Seq data. Nucleic Acids Res. 2014;42:D92–7. 33. Li S, Yao J, Xie M, LiuY ZM. Exosomal miRNAs in hepatocellular carcinoma development and clinical responses. J Hematol Oncol. 2018;11:54. 34. Wei R, Huang GL, Zhang MY, Li BK, Zhang HZ, Shi M, et al. Clinical significance and prognostic value of microRNA expression signature in hepatocellular carcinoma. Clin Cancer Res. 2013;19:4780–91.

Journal

Journal of Hematology & OncologySpringer Journals

Published: Apr 11, 2019

There are no references for this article.