Abstract
Background/Aim: Budding uninhibited by benzimidazole-related 1 (BUBR1) and endothelial nitric oxide synthase (eNOS) are related to aging and angiogenesis. This study examined the effect of low BUBR1 expression on eNOS expression in vivo, in vitro, and human gastric cancer tissues. Materials and Methods: Human umbilical vein endothelial cells (HUVECs) were passaged to investigate the effect of aging on BUBR1 and eNOS expression; expression of eNOS and phospho-eNOS protein was assessed in BUBR1 siRNA-transfected HUVECs. Additionally, guanosine 3’,5’ cyclic monophosphate (cGMP) and eNOS protein levels were measured in BUBR1-insufficient mice (Bubr1L/−). BUBR1 and eNOS expression levels were also evaluated in human gastric cancer tissues. Results: BUBR1 and eNOS, but not p-eNOS, levels were reduced significantly in aged and BUBR1 siRNA-transfected HUVECs. Additionally, cGMP production and the eNOS protein level were reduced in Bubr1L/− mice. Human gastric cancer tissues with low BUBR1 expression showed no eNOS expression. Conclusion: A decrease in BUBR1 reduced eNOS bioavailability through a pathway other than eNOS phosphorylation.
Both aging and angiogenesis are critical risk factors for the development of cancer and cardiovascular diseases (1). Aging, i.e. the inevitable time-dependent decline in organ function and tissue viability, advances cancer initiation and progression in many ways (1, 2). A variety of age-dependent changes, such as aneuploidy and DNA damage, occur at the cellular level in tissues (3). Aging also has a remarkable effect on endothelial function, leading to an increase in cardiovascular diseases including atherosclerosis, myocardial infarction, and stroke. Notably, endothelial dysfunction is one of the major phenotypic changes in aged arteries (4, 5).
Angiogenesis triggered by chemical signals from tumour cells makes tumours and metastasis progress rapidly (6). In contrast, therapeutic angiogenesis using fibroblast growth factor 2 (FGF2) (7), hepatocyte growth factor (HGF) (8), or vascular endothelial growth factor (VEGF) (9) has been widely regarded as an attractive approach for treating ischaemic limbs and for improving endothelial function (10).
Budding uninhibited by benzimidazole-related 1 (BUBR1) is an essential molecule in the spindle assembly checkpoint. The anaphase-promoting complex/cyclosome (APC/C) is inactivated by checkpoint protein complexes containing BUBR1 [mitotic arrest-deficient 3 (MAD3) in yeast], budding uninhibited by benzimidazole 3 (BUB3), MAD2 and cell-division cycle protein 20 (CDC20) until all kinetochores are attached to microtubules in an appropriate manner (11). Previous work demonstrated that BUBR1 overexpression is observed in malignant tumours derived from many organs, such as the stomach and liver (12). Notably, other studies demonstrated that mice expressing only 10% of normal BUBR1 level develop progressive aneuploidy and many aging phenotypes, such as cataracts, lordokyphosis, and loss of subcutaneous fat (13). In addition, specific early aging-associated vascular phenotypes were observed in these mice, including a reduced number of smooth muscle cells, reduced amount of elasticity, impairment of endothelial-dependent relaxation, and increased production of superoxide anions (5). BUBR1 insufficiency impairs angiogenesis and results in limb loss in ischaemic hind limbs through suppressing VEGF up-regulation (14).
Nitric oxide (NO) is produced from the conversion of L-arginine to L-citrulline by the enzyme endothelial nitric oxide synthase (eNOS) (15). eNOS is responsible for endothelium-derived NO production (16). To function in this role, eNOS must first be activated via phosphorylation (17). The activity of eNOS is regulated by many steps and acts in concert with various mediators (18).
Previous work suggested that in tumour tissues, eNOS induces angiogenesis and recruits perivascular cells, both of which promote tumour growth or metastasis, via NO production (19). In contrast, another study reported that eNOS-derived NO inhibits tumour metastasis (20).
Notably, both the level of expression and activity of eNOS are reduced by aging (21). The loss of NO activity is a central mechanism responsible for dysfunction of the vascular endothelium. An aging-induced increase in the formation of oxygen-derived free radicals appears to be the most likely explanation for the loss of NO (22).
Suppressor of sensor kinase 1 (SSK1) is related to numerous serine-threonine kinases that have high homology with BUBR1. The identification of SSK1 as a cDNA fragment overexpressed following laminar shear stress (23) suggests that BUBR1 could have a relationship with eNOS. Furthermore, the connection between BUBR1 and eNOS may be shared by both aging and angiogenesis. However, the relationship or interaction between BUBR1 and eNOS has not yet been investigated as far as we are aware. Here, we examined the influence of BUBR1 reduction on eNOS expression and activity by performing experiments in vitro, in vivo, and on human gastric cancer tissues.
Materials and Methods
Cell culture. Human umbilical vein endothelial cells (HUVECs) were obtained from a commercial provider (CC-2519; Lonza, Walkersville, MD, USA). HUVECs were cultured according to the supplier's instructions (CC-2535; Lonza). Cells at passages 4-9 were used for experiments and were plated in triplicate for each experimental condition.
Experimental animals. All animal experiments were conducted in accordance with experimental protocols approved by the University Committee of the Kyushu University Health Science Center (approval no. A24-157-01).
Low-BUBR1-expressing mutant mice (Bubr1L/− mice) and wild-type (WT) littermates (Bubr1+/+) with a mixed genetic background of C57BL/6J and 129/SvJ were generated in our laboratory, as previously described (14). The thymuses of experimental mice were frozen for use in western blotting and 3’,5’ cyclic monophosphate (cGMP) measurement. Mice were sacrificed at 7 weeks old for western blotting analysis and at 18-20 weeks old for cGMP measurement.
HUVEC transfection with small interfering RNA (siRNA). BUBR1 siRNA was purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). As control groups, a group transfected with scrambled siRNA (Invitrogen™, Carlsbad, CA, USA), and a group without siRNA transfection were included.
Protein extraction and western blot analysis. Total cellular proteins and murine thymus proteins were prepared as previously described (14). Membranes were incubated overnight at 4°C with primary: mouse monoclonal antibodies directed against BUBR1 (diluted 1:1000; BD Biosciences, Franklin Lakes, NJ, USA); eNOS (diluted 1:1000; BD Biosciences); or β-actin (diluted 1:2000; BioVision, Milpitas, CA, USA); with rabbit monoclonal antibodies directed against phospho(p)-eNOS (Ser 1177; diluted 1:1000; Cell Signaling, Danvers, MA, USA); or with rabbit polyclonal antibodies directed against BUBR1 (diluted 1:1000; Littleton, Novus Biologicals, CO, USA); or α-tubulin (diluted 1:1000; Abcam, Cambridge, UK).
RNA extraction and real-time polymerase chain reaction (RT-PCR). Total RNA of HUVECs was extracted using ISOGEN reagent (NIPPON GENE, Tokyo, Japan) according to the manufacturer's protocol. Total RNA was reverse transcribed to cDNA using the SuperScript III First-strand SuperMix (Invitrogen) according to the manufacturer's protocol.
For quantitative RT-PCR, 25 μl of amplification mixtures (Applied Biosystems StepOnePlus™ SYBR Green; Thermo Fisher Scientific, Waltham, MA, USA) were prepared containing cDNA (equivalent to 12.5-50.0 ng of reverse-transcribed RNA) and 0.4-1.0 μM primers according to the manufacturer's instructions. Reactions were run on an Applied Biosystems StepOnePlus™. Each sample was analysed in triplicate and normalized to the level of β-actin mRNA. The PCR protocol was an initial denaturation at 95°C for 5 min followed by 40 cycles of 95°C for 10 s, 60-65°C for 20 s, and 72°C for 10 s. Specific primer pairs used for the reactions were previously described (2, 24, 25).
VEGF treatment. After HUVECs were cultured under each experimental condition, 50 ng/ml recombinant human VEGF 165 (R&D systems, Minneapolis, MN, USA) was added to the medium, and the cells were then cultured for 10 min (26). Cell lysates were then prepared as above, and each protein was measured by western blotting.
cGMP assay. The function of endothelial cells of mice was assessed based on the level of cGMP, which was measured in lysates of the murine thymuses using an EIA kit from Cayman Chemical Inc. (Ann Arbor, MI, USA), as described in the manufacturer's instructions.
Human gastric cancer tissue samples. This study included samples from 106 unselected Japanese patients with primary gastric cancer, all of whom underwent a gastrectomy between 1994 and 2006 at the Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University Hospital. The patients were 72 men and 34 women, ranging in age from 29 to 86 years (mean=63.6 years). Disclosure of the study information on our homepage was substituted for written informed consent (approval no. 29-396). Thorough histological examination was made using hematoxylin-eosin-stained tissue preparations. No patient treated preoperatively with cytotoxic drugs was included in this study.
Immunohistochemical staining of BUBR1 and eNOS. Formalin-fixed, paraffin-embedded tissue specimens were used for immunohistochemical staining. A paraffin block that contained both cancerous tissue, which had invaded the deepest area of the stomach wall, and adjacent noncancerous tissue, was used in each case. Immunohistochemical staining was performed as previously described (12, 27) and used mouse monoclonal antibodies against BUBR1 (1:100; BD Biosciences) or eNOS (1:100 BD Biosciences). The intensity of cytoplasmic staining of BUBR1 was scored on a three-point scale in comparison with the staining of lymph follicles that were equally stained in all 106 specimens. Staining that was weaker than or similar to follicular staining was scored as 0 or 1 and defined as ‘low’ expression, while stronger staining was scored as 2 and defined as ‘high’ expression. The assessment of eNOS for all samples was performed as previously described (27).
Statistical analysis. All data are reported as means±SE. An unpaired Student's t-test was used to detect significant differences when two groups were compared. Statistical significance was accepted at a level of p<0.05.
Results
Senescence reduced the expression of both BUBR1 and eNOS in HUVECs. Western blotting was performed to determine the changes in BUBR1 and eNOS protein expression across HUVEC passages. The results show that the expression levels of both proteins were significantly reduced as HUVECs were cultured from passage 5 to 9 (Figure 1A-C. BUBR1: p<0.05, eNOS: p<0.01).
Next, we performed quantitative RT-PCR on sequential HUVEC passages. As the cells aged from passage 5 to 9, the mRNA expression levels of both BUBR1 and eNOS in HUVECs decreased, following the same trend as their protein expression levels (Figure 1D and E).
BUBR1 repression by siRNA transfection in HUVECs reduced eNOS expression without affecting eNOS activity. When the expression of BUBR1 was inhibited by siRNA transfection in HUVECs, the eNOS expression level was significantly reduced (passage 4, n=5) (Figure 2A-C). In contrast, the expression of p-eNOS by HUVECs transfected with BUBR1 siRNA was not significantly different from that of normal HUVECs (Figure 2D).
eNOS expression was reduced in the thymuses of low-BUBR1-expressing mice. Western blots were performed to measure eNOS protein expression in the thymuses of Bubr1+/+ and Bubr1L/− mice. In Bubr1L/− mice, the eNOS expression was significantly reduced compared with Bubr1+/+ mice (Figure 3).
Aged Bubr1L/− mice produce less cGMP than aged WT mice. To determine the NO production ability of mice, the cGMP quantity in mouse thymuses was measured by performing cGMP EIA assays. In aged mice (18-20 weeks old), the cGMP production in Bubr1L/− mice was significantly lower than that of Bubr1+/+ mice (n=3, p<0.05) (Figure 4).
VEGF-induced acceleration of eNOS phosphorylation was not significantly altered by BUBR1 knockdown. HUVECs with normal or reduced (siRNA-transfected) BUBR1 levels were treated with VEGF to stimulate the expression of p-eNOS. The ratio of p-eNOS expression following VEGF treatment was not significantly different between control HUVECs and BUBR1 knockdown HUVECs (Figure 5).
Human gastric cancer tissues with low BUBR1 expression lacked eNOS expression. Among the 106 human gastric cancer tissues examined via immunohistochemical staining in this study, 98 samples had high levels of BUBR1 expression and eight had low levels of BUBR1 expression. In the eight samples with low BubR1 expression, eNOS expression was not observed at all (Figure 6A), despite the detection of both BUBR1 and eNOS expression in positive control samples (Figure 6B).
Discussion
Previous studies reported that BUBR1 has a high homology with SSK1, a cDNA fragment of which was overexpressed following laminar shear stress in HUVECs (23). Here, we showed that BUBR1 insufficiency reduced eNOS expression both in vitro and in vivo, and we also found a lack of eNOS expression in gastric cancer tissues with low BUBR1 expression.
To the best of our knowledge, there have been no previous reports on the relationship between BUBR1 and eNOS. Our novel findings in this study are summarized as follows: i) A reduction of BUBR1 gene expression in HUVECs reduced eNOS expression; ii) Levels of both eNOS protein and cGMP in the thymus were significantly lower in aged Bubr1L/− mice compared with aged WT mice; iii) VEGF-induced phosphorylation of eNOS was not influenced by BUBR1 inhibition; and iv) eNOS expression was not detected in BUBR1-negative cells within human gastric cancer tissues.
In HUVECs, steady laminar shear stress for 6 h induces an up-regulation of the SSK1 gene, which encodes for a kinase protein related to BUBR1 (23). Shear stress also increases eNOS transcription via a pathway involving activation of the tyrosine kinase c-SRC and extracellular signal-related kinase 1/2 (ERK1/2). After several hours of stimulation by shear stress, the expression levels of both eNOS mRNA and protein were increased (28).
Here, we first evaluated the expression of BUBR1 and eNOS in HUVECs to investigate the effect of aging. Aged HUVECs exhibited significantly lower mRNA and protein levels of both eNOS and BUBR1 than young HUVECs (Figure 1). Additionally, BUBR1 siRNA-transfected endothelial cells showed a significantly lower eNOS expression compared with control HUVECs, suggesting that the down-regulation of eNOS observed in aged HUVECs may be caused by an age-dependent reduction in BUBR1 level (Figure 2A-C).
Next, we investigated the effect of BUBR1 insufficiency on eNOS expression in vivo. We previously generated the Bubr1L/− mouse strain in which BUBR1 expression in the testis was reduced to 16.8% of the normal level (14). BUBR1 and eNOS were reported to be expressed abundantly in the murine thymus (29), therefore we performed western blots to assess the eNOS expression in the thymus of Bubr1L/− mice. The results revealed that BUBR1 insufficiency in vivo also reduced eNOS expression.
Biochemical analysis of Bubr1L/− mouse thymuses demonstrated significantly reduced the level of cGMP, a second messenger for NO. Because eNOS activity is regulated via phosphorylation and previous work showed that p-eNOS is reduced in aged HUVECs (21), we examined the effect of BUBR1 siRNA on eNOS phosphorylated at Ser1177. Despite the lower level of total eNOS in BUBR1 siRNA-transfected HUVECs compared with control HUVECs, we found no significant difference between their p-eNOS expressions. Together, these findings suggest that the observed age-dependent decrease of p-eNOS in HUVECs is not mediated through an age-dependent decrease in BUBR1.
The VEGF–phosphoinositide 3-kinase/protein kinase B pathway has been implicated in regulating eNOS enzyme activity through phosphorylation (17). Based on this connection, we examined the effect of BUBR1 insufficiency on VEGF-stimulated eNOS phosphorylation in vitro. The results indicate that BUBR1 insufficiency did not affect the level of VEGF-stimulated eNOS phosphorylation. These findings suggest that another pathway likely exists between BUBR1 and eNOS. Abnormal BUBR1 expression may be an important factor for DNA aneuploidy and may contribute to chromosomal instability (19-23). Additionally, abnormal BUBR1 expression was found in gastric cancer. Previous work on gastric cancer tissues found conflicting results concerning the relative level of eNOS expression in tumour tissue compared with that in normal tissue; one study reported that eNOS expression was lower in tumour tissue (30), while another found that it was higher (31). Here, we examined the relationship between BUBR1 expression and eNOS expression in gastric cancer tissues. The immunohistochemical staining results revealed that there was no eNOS expression in the gastric cancer tissues with low BUBR1 expression, which may support our in vitro and in vivo data.
The pathway by which BUBR1 reduction is connected to eNOS repression remains unclear. The molecular mechanisms responsible for the down-regulation of NO production and its biological activity in BUBR1 insufficiency are also unknown and remain to be studied. This study is a first step into analysing the function of BUBR1 in aging and angiogenesis in view of its relationship with eNOS.
Acknowledgements
The Authors thank Natsumi Maeda for assistance with the animal experiments and thank Eriko Iwata for technical assistance. We also thank Katie Oakley, Ph. D., from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
Footnotes
Sources of Funding
This work was supported in part by a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (Grant No. 25462164 to TM) and Uehara Memorial Foundation.
- Received September 18, 2018.
- Revision received October 11, 2018.
- Accepted October 15, 2018.
- Copyright© 2018, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved