Hypoxia Modulates the Expression of Adenosine Receptors in Human

Expanded Methods MS ID# HYPERTENSION/2004/032854-R1

Hypoxia Modulates the Expression of Adenosine Receptors in Human

Endothelial and Smooth Muscle Cells Towards an A2B Angiogenic Phenotype

Running title: Hypoxia and Adenosine Receptors

Igor Feoktistov1, Sergey Ryzhov2, Hongyan Zhong3, Anna E. Goldstein3Goldstein4,

Anton Matafonov4Matafonov5,

Hongyan Zhong5, Dewan Zeng6, Italo Biaggioni7

Divisions of Cardiovascular Medicine1,3 4 and Clinical Pharmacology2,45,7,

Departments of Medicine1,34,7 and Pharmacology2,45,7,

Vanderbilt University, Nashville, TN

Department of Drug Research and Pharmacological Sciences5Sciences3,6, CV Therapeutics, Inc., Palo Alto, CA

Word count:

Figures number: 6

Corresponding author: Igor Feoktistov, Ph.D.

360 PRB

Vanderbilt University

Nashville, TN 37232-6300

615-936-1732

615-936-1733

ABSTRACT

We previously reported that adenosine A2B receptors stimulate angiogenesis. Because hypoxia is a potent stimulus for the release of both adenosine and angiogenic factors, we tested the hypothesis that hypoxia alters the expression of adenosine receptor towards an “angiogenic” phenotype, using human umbilical vein endothelial cells (HUVEC) and bronchial smooth muscle cells (BSMC). These cells were selected because aAdenosine does not release vascular endothelial growth factor (VEGF) under normoxic conditions in these cells. In HUVEC, three hours of hypoxia (4.6% pO2) decreased A2A adenosine receptor mRNA from 1.56±0.3% of β-actin expression to 0.16±0.01%, whereas A2B mRNA increased from 0.08±0.01 to 0.27±0.05%. Normoxic HUVEC express a characteristic A2A phenotype (the selective A2A agonist CGS21680 was as potent as the non-selective agonist NECA in activating adenylate cyclase), whereas CGS21680 became ineffective in hypoxic HUVEC (NECA>CGS21680, A2B phenotype). NECA did not change VEGF protein levels in normoxic HUVEC, but increased them from 9.5±1.0 to 14.2±1.2 pg/mL (p<0.05) after three hours of hypoxia, indicating that increased A2B receptors were functionally coupled to upregulation of VEGF. Hypoxia had similar effects on BSMC, increasing A2B mRNA 2.4±0.3-fold from 0.425±0.0439% to 1.001±0.1326% β-actin. NECA upregulated VEGF in hypoxic but not normoxic BSMC, from 74.6±9.6 to 188.3±16.7 pg/mL (p<0.01), an effect inhibited by the selective A2B antagonist CVT-6694. A2B receptors activated a mutant VEFG reporter made unresponsive to hypoxia by mutating its hypoxia-inducible factor 1 binding elementunresponsive to hypoxia, indicating independent mechanisms independent from hypoxia. In conclusion, hypoxia modulates the expression of adenosine receptors in human endothelial and smooth muscle cells towards an A2B “angiogenic” phenotype.

Word count: 2490

Keywords: Adenosine; Receptors, Purinergic P1; Endothelium, Vascular; Hypoxia; Vascular Endothelial Growth Factor A, hypoxia-inducible factor 1.

The purine nucleoside adenosine is an intermediate product of adenine nucleotides metabolism. Adenosine plays an important physiological role and, in many organs, it serves as a “retaliatory metabolite” in situations when oxygen supply is decreased or energy consumption is increased. Under these conditions, adenosine is released into the extracellular space and signals to restore the balance between energy supply and demandlocal energy requirements. Four extracellular G protein-coupled receptors, namely A1, A2A, A2B and A3, mediate adenosine actions. A2B receptors have a lower affinity compared to other receptor subtypes and require micromolar concentrations of adenosine for their stimulation1. Such high levels of extracellular adenosine can be reached during hypoxia, ischemia, inflammation, and injury2. The low affinity of A2B receptors suggests that they are primarily engaged under these pathophysiological conditions.

A2B receptors regulate various pathological processes, including mast cell activation3, vasodilation4, inhibition of cardiac fibroblast5 and vascular smooth muscle growth6, stimulation of endothelial cell growth7, and angiogenesis8-10. Stimulation of angiogenesis appears to be an important function for A2B receptors. We have previously shown that A2B receptors upregulate the production of angiogenic factors in human mast cells, retinal endothelial cells and human microvascular endothelial cells (HMEC-1) under normoxic conditions8;9.

Tissue hypoxia is a powerful stimulus for the expression of genes associated with angiogenesis and, as mentioned previously, it is during hypoxia that adenosine levels increase to concentrations that engage A2B receptors. Therefore, in this study we tested the hypothesis that hypoxia would also modulate expression of adenosine receptors subtypes towards an “angiogenic” A2B phenotype.

Materials and Methods

Cell culture and treatment conditions

HUVEC were purchased from Cambrex Bio Science (Walkersville, MD), HMEC-1 were were kindly provided by Dr. D.E. Vaughan (Vanderbilt University, Nashville, TN) and normal human bronchial smooth muscle cells (BSMC) were obtained from Clonetics (San Diego, CA). Cells were maintained as described previously9;11.

Before each experiment, the growing medium was replaced with a fresh one containing 1 U/mL adenosine deaminase. Confluent monolayer cultures were exposed to hypoxia by placing them in an modular incubation chamber (Billups-Rothenberg, Del Mar, Ca) with the atmosphere continuously monitored by an oxygen analyzer. Hypoxic conditions were created by flusheding with a 5% CO2, 95% N2 gas mixture until oxygen concentration inside chamber reached 4.6%. The hypoxic chamber was then sealed and placed in a 370C cell culture incubator for the indicated time.

Chemicals

5'-N-ethylcarboxamidoadenosine (NECA) and 2-p-(2-carboxyethyl)phenethylamino-NECA (CGS21680) were purchased from Research Biochemicals, Inc. (Natick, MA). CVT-6694 was made at CV Therapeutics (Palo Alto, CA).

Gene expression assay and real-time reverse transcription-polymerase chain reaction (RT-PCR)

Human adenosine receptors gene expression array was custom designed by Super Array (Bethesda, MA). The assay and real-time RT-PCR were performed as previously described9.

Measurement of cAMP

Cyclic AMP concentrations were determined as previously described12.

Determination of vascular endothelial growth factor A (VEGF) levels

HUVEC were lysed with PBS containing 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, and a 0.01% protease inhibitor cocktail. After passing through a 23G needle, the cell debris was pelleted by centrifugation and VEGF was measured in supernatant using an ELISA kit (PeproTech, Rocky Hill, NJ). VEGF secretion from BSMC was determined using an ELISA kit (Biosource, Camarillo, CA).

Transfections and luciferase reporter assay

HMEC-1 cells were transfected as we previously described9. P11w, a firefly luciferase reporter plasmid, comprising 5' flanking -985 to -939 base pairs of the human VEGF gene that that include a HIF-1 binding site, and P11m, the mutated version of P11w containing non-functional HIF-1 binding site13 were obtained from ATCC (Manassas, VA). Eighteen hours after transfection, cells were incubated under hypoxic or normoxic conditions for the next 12 hours and reporter activity was measured as previously described9.

Results

Effect of hypoxia on mRNA expression of adenosine receptor subtypes in HUVEC

Gene expression array showed that HUVEC, incubated in normoxic conditions, express preferentially A2A receptor mRNA (Figure 1). As a percentage of β-actin expression, HUVEC expressed levels of A2A and A2Breceptors of 1.56±0.3 and 0.08±0.01%, respectively. No mRNA encoding A1 or A3 receptors was detected in agreement with previously published results9. A2A receptor mRNA expression decreased in HUVEC cells exposed to hypoxia for one and three hours, from 1.56±0.3 to 0.51±0.12 and 0.16±0.01% of β-actin expression, respectively. In contrast, A2B receptor mRNA increased from 0.08±0.01 to 0.30±0.01 and 0.27±0.05%, respectively. Importantly, tThe ratio of expression of A2B:A2A mRNA changed from a predominance of A2A (0.05:1) to a predominance of A2B (1.6:1) after 3 hours of hypoxia.

Effect of hypoxia on functional expression of adenosine receptors in HUVEC

Both A2A and A2B adenosine receptors are known to stimulate adenylate cyclase with specific agonist profiles. Therefore, we measured accumulation of cAMP as a way to determine if hypoxia-induced changes in expression of adenosine receptors mRNA translate into an A2B-type phenotype. Our results confirmed the functional predominance of A2A receptors in HUVEC maintained under normoxic conditions (Figure 2A). The A2A-selective agonist CGS21680 activated adenylate cyclase with an EC50 of 849 nmol/L. The non-selective agonist NECA had similar potency (EC50 of 948 nmol/L), and efficacy as CGS21680. This pharmacological profile is consistent with predominant expression of A2A adenosine receptors14.

In contrast, we observed loss of functional coupling of A2A adenosine receptors to adenylate cyclase in HUVEC incubated under hypoxic conditions for 3 hours. NECA stimulated accumulation of cAMP with an EC50of 1.4 µmol/L, whereas CGS21680 was virtually ineffective (Figure 2B). This pharmacological profile is consistent with the functional predominance of A2B receptors12;15.

Effect of hypoxia on regulation of VEGF by adenosine in HUVEC

Figure 3 shows the effects of 100 µmol/L NECA, 100 µmol/L, had no effect on the production of VEGF protein in HUVEC incubated for 3 hours under normoxic or hypoxic conditions (Figure 2), . Under normoxic conditions NECA had no effect on VEGF protein levels in HUVEC, in agreement with our previous studies9. Under hypoxic conditions, however, NECA increased VEGF protein levels in HUVEC from 9.5±1.0 to 14.2±1.2 pg/mL (p<0.05).

Effect of hypoxia on A2B receptor mRNA expression and regulation of VEGF by adenosine in smooth muscle cells

We used the human smooth muscle cell line BSMC as a model of a then examined whether hypoxia-induced gain of angiogenic function by adenosine is seen in nonvascular cell types expressing higher levels of VEGF . We selected the human smooth muscle cell line BSMC as a model because we our previously studies demonstrated the presence of A2B receptors in these cells11, and the lack of effect of adenosine on VEGF secretion under normoxic conditions. Real-time RT-PCR analysis showed that uUnder normoxic conditions the these cells preferentially express A2B receptors had the highest transcript levels of (0.425±0.039% β-actin, (fFigure 5). Lower levels of A1 and A2A receptors were also detected (0.0023±0.0003 and 0.0097±0.002% β-actin, respectively), whereas transcripts for A3 receptors were below detection levels. Incubation of BSMC under hypoxic conditions for 1 hour increased A2B receptor transcript level 2.4±0.3-fold (to 1.001±0.126% β-actin, p<0.05, n=3), but had no significant effect on expression of A1 and A2A receptors (0.004±0.0006 and 0.0087±0.0003% β-actin, respectively).

As shown in figure 6, incubation of BSMC with 10 μM NECA under normoxic conditions for 24 hours had no effect on VEGF secretion compared to the basal levels. When cells were iIncubationed under hypoxia for 24 hours increased , the VEGF release of VEGF was increased by 2.6±0.3-fold compared to normoxia; VEGF release was further enhanced by NECA by 2.5±0.2-fold compared to basal levels. The selective A2B antagonist CVT-669411 (1μM) attenuated the effect of NECA by 64±21% indicating the involvement of A2B receptors.

Role of hypoxia-inducible factor 1 (HIF-1) in regulation of VEGF transcription by hypoxia and adenosine

HIF-1 is a transcription factor that mediates the effects of hypoxia on VEGF expression by binding to the hypoxia response element of the VEGF promoter13. We have previously shown that NECA, acting via A2B receptors, stimulates VEGF promoter in HMEC-19. Therefore, we examined if NECA interacts with hypoxia via the HIF-1 pathway, leading to upregulation of VEGF transcription. To examine if NECA interacts with the HIF-1 pathway to upregulate VEGF transcriptionanswer this question, we used two previously described luciferase reporters. The P11w reporter is regulated by a fragment of the VEGF promoter that includes a HIF-1 binding site. The P11m reporter is identical but for similar to P11w, but contains a three base pair mutation that prevents HIF-1 binding13. We co-expressed these reporters together with A2B receptors to minimize the variability in expression levels of A2B receptors in experimental groups of cells due to potential effects of hypoxia on native adenosine receptors17. As seen in figure 4A, 10 μM NECA increased luciferase activity of the P11w reporter by 2.3±0.03-fold in HMEC-1 incubated under normoxic conditions for 12 hours. Incubation of cells under hypoxic conditions for 12 hours resulted in 2.2±0.3-fold and 4.1±0.1-fold increase in P11w reporter activity in the absence and in the presence of NECA, respectively. NECA also stimulated the activity of the P11m reporter by 2.3±0.04-fold and 2.0±0.27-fold under normoxic or hypoxic conditions, but hypoxia itself had no effect on activity of this reporter containing non-functional HIF-1 binding site (figure 4B). These results suggest that, in contrast to hypoxia, adenosine-induced VEGF upregulation does not require the HIF-1 pathway.

Discussion

Hypoxia is a feature of many pathophysiological conditions including ischemia, inflammation, and tumor growth. Cell hypoxia is a potent stimulus for adenosine release. The physiological importance of adenosine, therefore, is probably greatest during hypoxic conditions. Mammalian cells respond to hypoxia by significant genetic reprogramming with selective gene induction and down-regulation. Our data show that hypoxia produces dramatic changes in the expression of adenosine receptor subtypes in HUVEC. These cells normally express both A2A and A2B receptors, but A2A receptor predominates in terms of both mRNA expression and functional coupling9. Hypoxia down-regulated expression of high-affinity A2A receptors, but at the same time it increased expression of low-affinity A2B receptors. Hypoxia also altered responses to adenosine agonists, indicating that changes in mRNA expression led to changes in expression of functional receptors. The non-selective A2 agonist NECA stimulated adenylate cyclase in hypoxic HUVEC, whereas the selective A2A agonist CGS21680 was no longer able to do so, indicating that A2B receptors became functionally predominant in HUVEC after hypoxic treatment.

It should be noted, however, that hypoxia does not always downregulate A2A receptors. It has been reported that hypoxia increased expression of A2A receptors in rat PC12 pheochromocytoma cells18. It is possible, therefore, that the effects of hypoxia on expression of A2A adenosine receptors can be cell- and tissue-specific.

In contrast to downregulating high-affinity A2A receptors, hypoxia increased mRNA expression of low-affinity A2B receptors in HUVEC. These results agree with our previous data, when simulation of hypoxia with cobalt ions upregulated expression of A2B receptors in U87MG cells16. Studies from other laboratories have also shown an upregulation of adenosine A2B receptors by hypoxia17;19. Of interest, hypoxia not only upregulates A2B receptors, but also increases the levels of endogenous ligand by enhancing hydrolysis of ATP to adenosine through upregulation of apirase and 5’-nucleotidase17. Therefore, the functional importance of low-affinity A2B is probably greatest under hypoxic conditions, when concentrations of extracellular adenosine are increased. The differential upregulation of adenosine receptors we report here may provide a positive feedback mechanism for the engagement of A2B receptors during hypoxia. The observation that adenosine induced upregulation of VEGF protein levels only after treatment with hypoxia, an effect that was absent in HUVEC maintained under normoxic conditions, indicate that hypoxia-induced modulation of adenosine receptors was functionally important. A switch from expression of mostly A2A receptors to predominantly A2B phenotype could be beneficial if it can promote cell and tissue survival, or may be detrimental if A2B receptors participate in inflammatory processes or pathological angiogenesis. In the cardiovascular system, an upregulation of A2B receptors in cardiac fibroblasts and smooth muscle cells may prevent cardiac remodeling associated with hypertension, myocardial infarction, and myocardial reperfusion injury after ischemia5;6 while promoting endothelial cell growth and angiogenesis7;9.