Adrenomedullin Increases the Expression of Calcitonin-Like Receptor and Receptor Activity Modifying Protein 2 Mrna in Human Microvascular Endothelial Cells

Home , Somatostatin receptor

505 Adrenomedullin increases the expression of calcitonin-like receptor and receptor activity modifying protein 2 mRNA in human microvascular endothelial cells

Nele Schwarz, Derek Renshaw1, Supriya Kapas2 and Joy P Hinson Centre of Molecular Endocrinology, St Bartholomew’s and the Royal London School of Medicine and Dentistry, Queen Mary University, London, UK 1Centre for Biochemical Pharmacology, St Bartholomew’s and the Royal London School of Medicine and Dentistry, Queen Mary University, London, UK 2Research Centre for Clinical and Diagnostic Oral Sciences, St Bartholomew’s and the Royal London School of Medicine and Dentistry, Queen Mary University, London, UK (Requests for offprints should be addressed to N Schwarz; Email: [email protected])

Abstract Adrenomedullin (AM) is a multifunctional peptide hormone, specific, since we were able to block the AM-induced effect which plays a significant role in vasodilation and angiogenesis, with 1 mM U0126, a specific mitogen-activated protein/ex- implicating it in hypertension as well as in carcinogenesis. AM tracellular signal-regulated kinase kinase inhibitor. Using real- exerts its effects via the calcitonin receptor-like receptor time PCR, we were able to show for the first time that AM (CRLR, now known as CL) complexed with either receptor upregulates peptide and mRNA expression of vascular activity modifying protein (RAMP) 2 or 3. We have endothelial growth factor (VEGF). However, AM treatment investigated the effect of AM on immortalized human of cells did not result in increased cell proliferation. Instead, microvascular endothelial cells 1, since endothelial cells are we observed that AM and VEGF induced cell migration, a major source as well as a target of AM actions in vivo. Cells which could be inhibited by the AM22–52 and anti-VEGF treated with AM showed elevated cAMP in a time (5– K6 K14 antibody respectively. AM also significantly elevated mRNA 45 min)-dependent and dose (10 –10 M)-dependent levels of CL (after 2 and 24 h treatment) and RAMP2 (after 1 manner. Pre-treatment with the AM receptor antagonist and 24 h treatment). The upregulation of the AM receptor at AM22–52 partially suppressed the AM-induced increase in two time points reflects possibly different cellular responses to cAMP levels. An increase in extracellular signal-regulated short- and long-term exposure to AM. kinase 1/2 phosphorylation was observed after 5 min of K Journal of Endocrinology (2006) 190, 505–514 treatment with 10 8 M AM. This phosphorylation was

Introduction receptor) (McLatchie et al. 1998, Poyner et al. 2002). Although the sequence identity between both RAMPs is Adrenomedullin (AM) is a 52-amino acid peptide originally only 30%, AM1 and AM2 receptors are pharmacologically isolated by Kitamura et al. (1993) from a human pheochro- indistinguishable and are usually co-expressed within the mocytoma. It is a multifunctional peptide produced by many same tissue (Kuwasako et al. 2002). cells and tissue systems (Kitamura et al. 1993), reviewed by Even though the adrenal glands show the highest tissue Hinson et al. (2000). AM was initially characterized by its AM mRNA concentration, endothelial cells show levels ability to stimulate cAMP production in human platelets and that are 20-fold higher than those found in the adrenal, exerted a potent and long-lasting vasodilative effect in the rat lending evidence towards making AM an important (Kitamura et al. 1993). AM has been proposed early as an hormone in the regulation of vascular tone. AM is important hormone in circulation control and maintenance considered to be most important in the paracrine control of vascular tone, since it regulates endothelial permeability of vascular, particularly microvascular, function where it acts (Hippenstiel et al. 2002) and it contributes to the as a potent vasodilator (Smith et al. 2002). AM increases differentiation of bone marrow-derived mononuclear cells coronary blood ﬂow and heart rate, as well as heart into endothelial progenitor cells (Iwase et al. 2005). contractability (Nagaya et al. 2002). The importance of AM AM exerts its effects via the G-protein coupled calcitonin- in vascular physiology was demonstrated by AM gene like receptor (CL) complexed with a receptor accessory knockout mice, leading to death in utero due to severe modifying protein known as receptor activity modifying malformations in vascular morphogenesis (Caron & protein (RAMP) 2 (AM1 receptor) or RAMP 3 (AM2 Smithies 2000, Shindo et al. 2001).

Journal of Endocrinology (2006) 190, 505–514 DOI: 10.1677/joe.1.06806 0022–0795/06/0190–505 q 2006 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access 506 N SCHWARZ and others $ AM elevates mRNA levels of CL and RAMP2

Additionally, AM has also been implicated in physiological and for different time points (time-course). To prevent cAMP and pathological angiogenesis using different types of knockout degradation, cell medium contained 1 mM 3-isobutyl-1- mice, xenografted tumors, and in vitro models (Hague et al. methylxanthine. To stop the reaction, 95% ethanol was added 2000, Nikitenko et al. 2000, Oehler et al. 2001, Martinez et al. and plates were stored at K20 8C overnight. After defrosting, 2002, Kim et al. 2003); AM has also been reported to be the following day, liquid from the wells was completely expressed as an angiogenic factor in tumors. For example, in transferred to clean 1$5 ml Eppendorff tubes and the samples some cancers, AM expression is associated with vascular density were dried in a vacuum centrifuge. The dried samples were and endothelial cell proliferation (Hague et al. 2000). In vitro reconstituted in 500 ml assay buffer ED2, provided in the kit, studies have shown the AM receptor antagonist AM22–52 was and stored at K20 8C until further use. able to abrogate tumor formation of a pancreatic tumor cell line suggesting a role for AM in this cancer (Ishikawa et al. 2003). In the endothelial cells, AM and vascular endothelial growth MTT assay factor (VEGF) act in a conjoined manner to induce angiogenic To determine the effects of AM on cell proliferation, effects in vitro, but the angiogenic actions of AM appear to be approximately 60 000 cells were plated into each well of independent of VEGF secretion (Fernandez-Sauze et al. 2004). 24-well plates. Instead, AM upregulates VEGF mRNA levels in the ischemic Cells were deprived of serum overnight prior to incubation hind limb of mice after 1 day of treatment, thereby possibly with 0 (serum free, SF), different concentrations of AM or enhancing the angiogenic effect of VEGF (Iimuro et al. 2004). growth factor-containing medium (serum rich, SR). After 24, The main aim of the present study was to determine the 48, or 72 h treatment, cells were washed with PBS and biological actions, functions, and possible target genes of AM subsequently incubated for 2 h in the dark with 0$5 mg/ml using human microvascular endothelial cells (HMEC) 1. Most MTTreagent (thiazolyl blue tetrazolium bromide). The MTT pathological events involving endothelial cells occur at the solution was removed and cells were lysed by adding 250 ml level of the microvasculature, which is also thought to be the 10% DMSO/90% isopropanol mixture. Plates were centri- main target tissue of AM in vivo. fuged for 5 min at 13 000 r.p.m. and 200 ml of each sample transferred into wells of a 96-well plate. Plates were read using Materials and Methods an optical density of 570 nm in a microtiter plate reader. Cell culture RNA extraction and cDNA synthesis HMECs were purchased from the Center for Disease Control in Atlanta, GA, USA. This cell line was obtained by Total RNA was extracted from cultured cells using the transfecting human dermal microvascular endothelial cells RNeasy Mini Kit (Qiagen). The extractions were carried out with a PBR-322-based plasmid containing the coding region according to the manufacturer’s instructions. cDNA was for the Simian virus 40T gene product, and large T antigen synthesized using the First-strand cDNA Synthesis kit from (Ades et al. 1992). Cells were cultured in MCDB-131 GE Healthcare (Amersham) following the manufacturer’s medium with 11$6 g/l L-glutamine, 5% fetal bovine serum, instructions. 10 mg/l epidermal growth factor and 1% penicillin/strepto- 2 mycin mixture (all obtained from Sigma–Aldrich) in T75 cm Western blot analysis tissue culture flasks (Triple Red, Thame, Oxon, UK) or six- well plates (Nunc, VWR, Leicestershire, UK). Culture Total protein was extracted using 60% confluent cells from medium was changed every 2–3 days and cells split when T75 flasks. Prior to extraction, cells were washed with PBS $ they reached 90% confluency. For the experiments, cells with (137 mM NaCl, 27 mM KCl, 43 mM Na2HPO4 7H2O, $ passage numbers between 10 and 20 were used. 14 mM KH2PO4,pH7 4). After adding 300 ml CytoBuster Protein Extraction Buffer (Novagen, Windsor, Berks, UK), Materials flasks were incubated at room temperature for 5 min. The resulting solution was transferred to a sterile 1$5ml Peptides and cell culture medium including necessary Eppendorff tube and centrifuged for 5 min at 16 000 g at supplements were purchased from Sigma–Aldrich 4 8C. The supernatant was removed to a fresh tube and sonicated for 10 s to achieve a homogenous solution. Equal cAMP assay amounts of protein were loaded on 10% Tris–HCl gels and cAMP levels were determined by ELISA following the subsequently transferred to a Hybond-P polyvinylidene manufacturer’s instructions (Cyclic AMP Immunoassay; diflouride membrane (GE Healthcare, Amersham). After R&D System, Abingdon, Oxfordshire, UK). Approximately, blocking with 5% non-fat dry milk solution, the membranes 250 000 cells were plated into each well of a six-well plates were incubated with primary antibody at 1:1000 dilution at and deprived of serum overnight. The next day, cells were 4 8C overnight. Phospho-p44/42 mitogen-activated protein treated with AM at various concentrations (dose response) kinase (MAPK) (Thr202/Tyr204) (E10) monoclonal

Journal of Endocrinology (2006) 190, 505–514 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access AM elevates mRNA levels of CL and RAMP2 $ N SCHWARZ and others 507 antibody and p44/42 MAPK antibody were purchased from 4!106 cells/ml in SF medium and either plated directly into Cell Signaling Technology (Hitchin, Herts, UK). the wells of the polycarbonate filter or pre-incubated for A peroxidase-conjugated secondary goat anti-rabbit 45 min with an anti-VEGF-antibody (Sigma–Aldrich) or K7 antibody was used at 1:1000 dilution. The blots were AM22–52 (10 M). The bottom wells of the neuroprobe developed using the solutions of the ECL plus Western plate were loaded with 300 ml of different concentrations of Blotting Detection System (GE Healthcare, Amersham). AM or VEGF (Sigma–Aldrich). In the bottom wells, 300 ml serum-containing and SF media were also loaded and served Gene quantitation as positive and negative controls for migration respectively. The polycarbonate filter was placed on top of the lower Quantitative real-time reverse transcriptase (RT)-PCR was chamber and 25 ml HMECs were placed on top of the filter. performed using dual-labeled fluorescent taqman probes to To determine migrated cell numbers, a serial dilution of cells determine CL, RAMP2, RAMP3, and VEGF-A mRNA ranging from 4!106 to 12$5!104 cells was plated directly levels as listed in Table 1. RT-PCR amplifications were into the wells of the lower plate. As an additional control to carried out for all genes studied and the resulting products determine whether the migratory response is genuinely sequenced to ensure primer specificity. Taqman reagents were chemotactic, cells were incubated with either AM or VEGF obtained from Stratagene (Amsterdam, Netherlands) and and placed on the ploycarbonate filters with either SF primers and probes were supplied by Sigma–Genosys. medium in the lower chamber or the corresponding Analysis was carried out using the MX4000 (Stratagene) concentration of AM and VEGF. starting cycle with an initial 95 8C for 10 min followed by 40 Plates were incubated for 5 h in an incubator at 37 8C with cycles consisting of 95 8C for 30 s and 55 8C for 1 min. 5% CO2. Cells remaining on top of the filter were absorbed off and the filter tops were washed with SF medium to ensure VEGF peptide assay removal of all non-migrated cells. The filters were incubated with 2 mM EDTA for 30 min at 4 8C to loosen cells from the VEGF levels were determined by ELISA following the inner filter membrane. The EDTA solution was removed and manufacturer’s instructions (Quantikine, Human VEGF plates were subsequently spun at 400 g for 10 min. The filter Immunoassay kit, R&D System). Approximately, 250 000 was removed and 150 ml solution were taken out of the bottom cells were plated into each well of a six-well plate and wells and replaced with 150 ml2! MTT dye. Cell deprived of serum overnight. Cells were incubated for 4 h quantification was carried out as described for the MTT K7 K9 with 10 and 10 M AM. Liquid from the wells was assay above. completely transferred to clean 1$5 ml Eppendorff tubes and centrifuged. Supernatants were stored at K20 8C until further use. Statistical analysis The statistical analysis was carried out using GraphPad Migration assay PRISM software (version 3.0; Graph pad, San Diego, CA, USA), one-way ANOVA and Turkey–Kramer post hoc test. Chemotaxis assays were performed using a Neuroprobe Data were expressed as meansGS.E.M. ChemoTx plate with 8 mm pore size (Receptor Technologies Ltd, Adderbury, Oxfordshire, UK). HMECs were diluted to

Table 1 Sequences of the used primers and dual-labeled fluorescent Results probes for CL, RAMP2, RAMP3, and VEGF AM stimulates cAMP in HMECs Primer/probe sequences Product size (bp) Treatment of HMECs with varying concentrations of AM K K Gene (10 6–10 14 M) for 5 min resulted in a significant increase in K K CL S: aaagcacaatcaacttttctgagc 118 cAMP above basal (Fig. 1). In particular, AM at 10 6–10 As: aataagggtagaatcatgcccaac 8 M increased cAMP levels by approximately twofold. Probe: agccagttccagcacaccattgca RAMP2 S: tctccctaggacccgagtcg 112 Forskolin (FSK) was used as a positive stimulation control AS: tgtgcctgtggtgggaagag of AM in these cells and we found that it enhanced the cAMP Probe: agccgtccgcctcctccttctgc response 3$5-fold above basal. Co-treatment of AM and FSK RAMP3 S: caggtttctatgctgtttcttagc 83 showed that AM can potentiate the FSK-induced increase in AS: gcaaggagtcctagtccaagc cAMP. Probe: ccagcctagccttagccgcagtct K8 VEGF S: aaccagcagaaagaggaaagagg 121 In time-course experiments, AM (10 M) significantly AS: cactcactttgcccctgtcg increased cAMP levels after 2$5 min of treatment (Fig. 2). Probe: ttcgctgctcgcacgcccgc The AM response peaked after 5 min of treatment resulting in an approximately 2$7-fold increase in cAMP above the bp, base pairs; S, sense strand; As, antisense strand corresponding SF control. www.endocrinology-journals.org Journal of Endocrinology (2006) 190, 505–514

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access 508 N SCHWARZ and others $ AM elevates mRNA levels of CL and RAMP2

Figure 1 cAMP levels in HMECs following incubation with AM. HMECs were incubated K K K with increasing concentrations of AM (10 6–10 14 M), 10 4 M forskolin (FSK) or serum- free (SF) medium. Values are meansGS.E.M., nZ9. *P!0$05, **P!0$01, ***P!0$001 compared with serum-free control.

K Todetermine whether the observed response is AM-speciﬁc, for 72 h or with 10 8 M AM for 24, 28, and 72 h. In our we pre-treated the cells for 30 min with varying concentrations hands, AM did not cause an increase in cell proliferation K6 K10 of AM receptor antagonist AM22–52 (10 –10 M; Fig. 3). compared with the serum deplete, negative control in either The AM-induced increase in cAMP levels was reduced in the experimental set-up (Fig. 6A and B). K6 K7 presence of AM22–52 at high concentrations (10 –10 M), Quantitative real-time PCR revealed that treatment of cells K8 but not completely abolished. Treatment with AM22–52 with AM (10 M) signiﬁcantly increased the mRNA levels K (10 6 M) had no effect on the cAMP levels of the positive or of CL, RAMP2 as well as VEGF. After 4 h, CL mRNA was negative controls, excluding short-term toxic effects of the elevated 4$8-fold with a further increase after 24 h to 15-fold inhibitor on the cells.

AM stimulates ERK 1/2 phosphorylation in HMECs To further examine the AM-induced signaling cascade in HMECs, we studied phosphorylation of MAPK/ERK 1/2 using western-blotting analysis. Treatment of HMECs with K AM (10 8 M) resulted in a signiﬁcant increase in ERK 1/2 phosphorylation after 5 min exposure (Fig. 4). The maximum mean ERK 1/2 phosphorylation increase observed was 2$9-fold above basal and this effect returned to basal levels after 10 min of treatment. This effect of AM on ERK 1/2 was speciﬁc since use of the MEK inhibitor, U0126, completely abolished the AM-induced phosphorylation of ERK 1/2 (Fig. 5).

Activation of AM target genes Since MAPK participates in a protein kinase cascade that plays a critical role in the regulation of cell growth, especially by the Figure 2 cAMP levels in HMECs after exposure to AM or forskolin K8 : K5 activation of ERK 1/2, we examined the effect of AM on with time. HMECs were incubated with 10 MAM( )or10 M forskolin (&) for increasing amounts of time and intracellular endothelial cell proliferation by MTT assay. Cells were either cAMP was measured by ELISA. Values are meansGS.E.M., nZ9. K6 K11 treated with different AM concentrations (10 –10 M) ***P!0$001 compared with control (time point 0).

Journal of Endocrinology (2006) 190, 505–514 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access AM elevates mRNA levels of CL and RAMP2 $ N SCHWARZ and others 509

Figure 3 Effect of AM receptor antagonist on HMEC cAMP levels. HMECs were incubated with decreasing concentrations of the AM antagonist AM22–52 for 30 min prior to being K8 exposed to 10 M AM for a further 5 min. Values are meansGS.E.M., nZ9. *P!0$05, **P!0$01, ***P!0$001 compared with serum-free control. above basal (Fig. 7). RAMP2 mRNA levels increased for 45 min with 20 mg/ml anti-VEGF antibody, prior to 1$5-fold after 1 h and 2$9-fold after 24 h treatment with loading into the upper chamber. Serum-containing medium AM (Fig. 7). The expression of RAMP3 mRNA levels did caused 78% of all cells to migrate into the lower chamber, not change significantly with the treatment and time points compared with 15% random migration in the serum-deplete studied (data not shown). negative control. Random cell migration using additional K The biggest change in mRNA levels after AM treatment controls with cell suspensions containing AM (10 10 M) or was seen with VEGF (Fig. 8). After 2 h exposure to AM, the VEGF (40, 60, or 100 ng/ml) with the equivalent initially very low VEGF levels increased 70-fold. This was followed by a decline to basal and a second mRNA elevation after 48 h (5$8-fold above basal). Treatment of cells for 4 h with AM also increased VEGF peptide levels (Fig. 9). AM K K 10 7 and 10 9 M increased VEGF peptide levels 2$2- and 1$9-fold respectively. Since neither AM nor the increased VEGF levels following AM treatment caused cell proliferation, we investigated whether these two peptides have a migratory effect on HMECs (Fig. 10). Using a modified Boyden chamber method, cell migration in response to AM, VEGF, and SR positive and SF negative controls was K determined. Placement of 10 10 MAMinthelower chamber induced a 1$9-fold increase in HMEC migration, equivalent to 29% of all cells. This effect was AM-specific, K7 since it could be blocked using AM22–52 (10 M). VEGF also caused a dose-dependent increase in cell migration. Figure 4 Western blot of ERK 1/2 phosphorylation in HMECs. Treatment with 100 ng/ml VEGF increased cell migration HMECs were serum starved overnight and subsequently treated K twofold (32% of cells), while 60 and 40 ng/ml VEGF with 10 8 M AM for the indicated time points. Cell homogenates increased cell migration 1$8- and 1$7-fold (27 and 25% of (15 mg protein) were tested for phosphorylated ERK 1/2 (A), stripped and re-probed for total ERK 1/2 (B). Scanning densitometry of three cells respectively) above SF control respectively. VEGF- different experiments is shown as a bar graph. Values are meansG induced cell migration was blocked by pre-incubating cells S.E.M.*P!0$05 compared with time 0. www.endocrinology-journals.org Journal of Endocrinology (2006) 190, 505–514

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access 510 N SCHWARZ and others $ AM elevates mRNA levels of CL and RAMP2

Figure 7 Expression of CL and RAMP2 mRNA expression in K response to AM. HMECs were treated with 10 8 M AM for increasing lengths of time and CL (triangles) and RAMP2 (open squares) mRNA was quantiﬁed using real-time RT-PCR and normalized to the housekeeping gene glyseraldehyde-3-phosphate Figure 5 Effect of an MEK inhibitor on ERK 1/2 phosphorylation in dehydrogenase. Values are expressed as meansGS.E.M., nZ3. HMECs. Western-blot analysis of 15 mg protein of cell homogenates *P!0$05, **P!0$01 when compared with time point 0. for phosphorylated (p) ERK 1/2 expression (upper panel) or total ERK K K 1/2 (lower panel) in HMECs exposed to 10 8 M AM, 10 5 M forskolin (FSK) or serum-free medium (SF) for 5 min in the absence concentration of peptide or SF medium in the bottom (K) or presence (C)of1mM U0126. Scanning densitometry of three chamber was determined. In all cases, migration in the different experiments is shown as a bar graph. Values are meansG additional controls was not signiﬁcantly different from the Z ! $ S.E.M., n 3. ***P 0 001 compared with presence of U0126. observed random unstimulated migration with the SF control (15%), ranging from approximately 12 to 18% of HMECs. All observed migratory effects with AM and VEGF are, therefore, genuine chemotactic responses of the cells and only 15% of total cell migration in the experiments can be attributed to chemokinetic cell movement.

Discussion

In this study, we have shown that AM induces cAMP elevation in HMECs, as has previously been reported for other endothelial cells (Isumi et al. 1998, Hippentiel et al. 2002). It was observed that AM was a very potent stimulant of cAMP in HMECs, since already very low concentrations of

Figure 6 MTT assay analysis of HMEC cell proliferation. HMECs K were treated for 72 h with increasing concentrations of AM (10 6– K Figure 8 Expression of VEGF mRNA expression in response to AM. 11 K 10 M) as well as serum-rich (SR) and serum-free (SF) medium (A). HMECs were treated with 10 8 M AM for increasing lengths of time For the time-course experiment, HMECs were treated for 24, 48, and K8 and VEGF mRNA was quantiﬁed using real-time RT-PCR and 72 h with either 10 M AM (white bars), SF (grey bars) or SR (black normalized to the housekeeping gene glyseraldehyde-3-phosphate bars) (B). Values are meansGS.E.M., nZ9. *P!0$05, **P!0$01, G Z ! ! $ dehydrogenase. Values are expressed as means S.E.M., n 3. *P ***P 0 001 compared with serum-free control. 0$05, **P!0$01 when compared with time point 0.

Journal of Endocrinology (2006) 190, 505–514 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access AM elevates mRNA levels of CL and RAMP2 $ N SCHWARZ and others 511

to be inefficient at blocking the AM receptor, therefore, the result observed in our study is not surprising (Champion et al. 1997, Nishikimi et al. 1998, Ziolkowska et al. 2003). It is clear that HMECs express at least one fully functional AM receptor sub-type, which when activated leads to an increased level of cAMP. Further, downstream of the cellular-signaling cascade, AM has been reported to induce as well as inhibit phosphorylation of ERK 1/2 in various different cell types. AM-induced decreases of phosphorylated ERK 1/2 were observed in rat mesangial cells (Parameswaran et al. 2000), while Jiang et al. (2004) report that AM inhibited the aldosterone-stimulated ERK activity in rat cardiac fibroblasts. In both reports, treatment of cells with AM leads to increased cAMP levels, Figure 9 Expression of VEGF peptide in response to AM in HMECs. K K Following treatment with 10 7 and 10 9 M AM for 4 h, decreases in ERK 1/2, and subsequently to inhibition of cell supernatants were collected and analyzed. Values are expressed as proliferation and protein synthesis. Other studies have shown picogram VEGF per 25!105 cells, nZ6. ***P!0$001 when that AM increases ERK 1/2 phosphorylation, thereby compared with serum-free control. exerting angiogenic and proliferative effects in rat vascular K K AM (10 10–10 14 M) were already able to elicit a significant smooth muscle cells and human umbilical vein endothelial cAMP response. The AM receptor antagonist AM22–52 did cells (HUVECs) (Iwasaki et al. 2001, Kim et al. 2003). K not completely block this response, even at 10 6 M In our study, treatment of cells not only resulted in elevated concentration. This antagonist has been shown previously cAMP levels, but also in an increase in ERK 1/2

Figure 10 Adrenomedullin and VEGF treatment of HMECs induce cell migration. The lower chambers of the neuroprobe chamber were loaded with 300 ml of varying K K concentrations (10 9–10 12 M) AM or 40, 60, and 100 ng/ml VEGF, while serum-free (SF) and serum-rich (SR) media served as negative and positive migration controls. After loading cells into the wells of the upper chamber, plates were incubated for 5 h, prior to staining with MTT. The migratory effect of both peptides was blocked by incubating the cell K7 suspension for 45 min with the AM receptor antagonist AM22–52 (10 M) or anti-VEGF antibody (Vab) (20 mg/ml) prior to loading on the upper chamber. Migration is expressed as fold change in cell number in the bottom chamber above serum-free (basal) control. Values are expressed as meansGS.E.M., nZ6. *P!0$05, **P!0$01 when compared with serum-free control. www.endocrinology-journals.org Journal of Endocrinology (2006) 190, 505–514

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access 512 N SCHWARZ and others $ AM elevates mRNA levels of CL and RAMP2

phosphorylation after exposure to AM for 5 min. This effect AM-induced increase in mRNA is also translated to the has been shown to be specific by the complete inhibition of protein level interactions of RAMP2 and CL with other response with 1 mM U0126. Elevated ERK 1/2 levels are proteins, additional signaling complexes may be formed. commonly associated with cell proliferation (Cobb et al. Since AM did not alter the expression of RAMP3, we 1994, Roux & Blenis 2004). However, in our hands, AM did propose that transcriptional regulation of this gene occurs via not increase HMEC proliferation during treatment at any of a yet unknown mechanism. the time points or AM concentrations used in this study. Another novel finding in HMECs is that AM upregulates Therefore, we propose that proliferation might be regulated VEGF mRNA and peptide levels. It has been shown in HMECs independent of cAMP and ERK1/2 phosphoryl- previously that AM induces VEGF mRNA levels in ation by a yet unknown pathway. HUVECs as well as in the hind limb of AM-treated mice We have shown for the first time that AM upregulates the (Iimuro et al. 2004). Therefore, AM might be an influential mRNA levels of CL and RAMP2 in HMECs. Quantitative factor in angiogenesis in HMECs, but stimulation of VEGF analysis of CL mRNA levels showed upregulation of the gene expression also gives AM an important role in carcinogenesis. in a bi-phasic manner, 4 and 24 h after exposure. The increase VEGF is implicated in endothelial barrier dysfunction, which in CL mRNA was higher after 24 h compared with the allows cancer cells to migrate across the vascular lining of elevation after 4 h. RAMP2 mRNA levels also appear to be vessels, one of the key events in cancer metastasis. regulated in a bi-phasic manner with an increase after 1 and Dysfunction of the vascular endothelial barrier might 24 h treatment. This could suggest that the initial increase in facilitate the widespread dissemination of cancer cells mRNA levels are a direct response to the AM stimulation, (Zachary & Gliki 2001). whereas the later CL and RAMP2 induction might be It seems surprising that AM has no indirect effect on cell mediated by distinct second messenger system components. proliferation either directly or via the elevation of VEGF In rat ileum, platelet-activating factor (PAF) increases the mRNA and peptide levels. AM has been reported previously mRNA levels of PAF receptor (PAF-R) in a bi-phasic manner to promote cell growth as well as inhibit proliferation. AM after 30 min and 6 h post-stimulation (Wang et al. 1997). PAF stimulates cell proliferation in zona glomerulosa cells, skin has been known to stimulate tumor necrosis factor (TNF) fibroblasts and keratinocytes (Albertin et al. 2003), gliobas- secretion in rat ileum cells, which in turn increases toma, lung cancer cells, and endometrial tumors (Oehler et al. endogenous PAF levels. The authors hypothesize that PAF 2002). Cell growth is inhibited by AM in myocytes, cardiac induces production of endogenous TNF in the cells, which in fibroblasts, vascular smooth muscle cells, prostate cancer cells turn increases the levels of PAF. This feedback loop then leads (Abasolo et al. 2004), and mesangial cells (Segawa et al. 1996). to the observed second peak in transcriptional activation of This discrepancy in biological function might be related to PAF-R mRNA. A similar yet unknown mechanism might be varying cell-type-specific differences in signal transduction involved in the regulation of CL and RAMP2 expression. pathways. In HMECs, AM and VEGF have been shown to Alternatively, the second peak of RAMP2/CL mRNA have significant effects on cell migration. However, the may reflect failure of HMECs to induce a repressor, which migratory response of HMECs to SR control is significantly downregulates or destabilizes these mRNA species. higher than to either AM or VEGF. It seems likely that other It seems apparent that AM can partially regulate the factors in addition to AM and VEGF might play a role in the expression of its own receptor. Transcriptional regulation of complex process of migration and angiogenesis. Angiopoetins receptors by their ligands has been shown in various different or cytokines such as pleiotrophin might be additional studies. Exposure of agonists to G protein-coupled receptors requirements for migration of microvascular endothelial frequently results in downregulation of the receptor transcript cells, as well as signals from other adjacent cell types, such as levels, but receptor upregulation has also been demonstrated fibroblasts and pericytes within their microenvironment. (Siegrist et al. 1994, Schanstra et al. 1998, Froidevaux & Nevertheless, AM seems to be an important contributing Eberle 2002, Anko¨ & Panula 2006). Inoue et al. (1999) found factor in the complex system of angiogenesis and vascular that calcitonin downregulates calcitonin receptor mRNA in morphogenesis, especially, since AM gene knockout mice mouse bone marrow cells and Dupre et al. (2003) report the display severe cardiovascular abnormalities and die early ligand-induced downregulation of the PAF-R. On the other during embryonic development. However, further studies are hand, gene expression of the human somatostatin receptor required to fully understand the role of AM in physiological type I is actively upregulated by somatostatin (Hukovic et al. and pathological events concerning the vasculature. HMECs 1999). The regulation of CL/RAMP2 mRNA levels by AM carry many traits of primary endothelial cells (Unger et al. in HMECs might reflect a physiological adaptation, which 2002), reviewed in Bouı¨s et al. (2001), but they are enables the cells to adjust the sensitivity of receptor-mediated immortalized cell lines and in vitro findings might not always processes by changes in receptor number, and, therefore, reflect those of in vivo studies. according to the level of receptor activation. In conclusion, we have demonstrated that the treatment of The fact that CL and RAMP2 mRNA levels are also HMECs with AM increases cAMP levels and phosphoryl- upregulated at different time points, 4 and 1 h respectively, ation of the ERK 1/2 pathway. In our cell line, this does not might indicate yet unknown functions of the proteins. If the lead to an increase in cell proliferation. Instead, we have

Journal of Endocrinology (2006) 190, 505–514 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access AM elevates mRNA levels of CL and RAMP2 $ N SCHWARZ and others 513 identified that the cAMP, ERK 1/2 pathway is involved in Hinson J, Kapas S & Smith D 2000 Adrenomedullin, a multifunctional AM-induced upregulation of CL, RAMP2, and VEGF regulatory peptide. Endocrine Reviews 21 138–167. mRNA and protein levels. AM and VEGF are able to induce Hippenstiel S, Witzenrath M, Schmeck B, Hocke A, Krisp M, Krull M, Seybold J, Seeger W, Rascher W, Schutte H & Suttorp N 2002 HMEC cell migration, suggesting a role for AM in Adrenomedullin reduces endothelial hyperpermeability. Circulation Research microvascular endothelial physiology and pathology. Further 91 618–625. studies are needed to determine whether AM influences the Hukovic N, Rocheville M, Kumar U, Sasi R, Khare S & Patel Y 1999 regulation of other genes, possibly inflammatory cytokines or Agonist-dependent up-regulation of human somatostatin receptor type 1 other growth factors. requires molecular signals in the cytoplasmic tail. Journal of Biological Chemistry 274 24550–24558. Iimuro S, Shindo T, Moriyama N, Niu P, Takeda N, Iwata H, Zhang Y, Ebihara A & Nagai R 2004 Angiogenic effects of adrenomedullin in ischemia and tumour growth. Circulation Research 95 415–423. Funding Inoue D, Shih C, Galson D, Goldring S & Horne W 1999 Calcium- dependent down-regulation of the mouse C1a calcitonin receptor in cells The work was supported by the Special Trustees of the Royal of the osteoclast lineage involves a transcriptional mechanism. Endocrinology London and St Bartholomew’s Hospitals. The authors declare 140 1060–1068. that there is no conflict of interest that would prejudice the Ishikawa T, Chen J, Wang J, Okada F, Sugiyama T, Kobayashi T, Shindo M, Higashino F, Katoh H, Asaka M et al. 2003 Adrenomedullin antagonist impartiality of this scientific work. suppresses in vivo growth of human pancreatic cancer cells in SCID mice by suppressing angiogenesis. Oncogene 22 1238–1242. Isumi Y, Shoji H, Sugo S, Tochimoto T, Yoshioka M, Kangawa K, Matsuo H References & Minamino N 1998 Regulation of adrenomedullin production in rat endothelial cells. Endocrinology 139 838–846. Iwasaki H, Shichiri M & Marumo F 2001 Adrenomedullin stimulates proline- Abasolo I, Wang Z, Montuenga L & Calvo A 2004 Adrenomedullin inhibits rich tyrosine kinase 2 in vascular smooth muscle cells. Endocrinology 142 prostate cancer cell proliferation through a cAMP-independent autocrine 564–572. mechanism. Biochemical and Biophysical Research Communications 322 878–886. Iwase T, Nagaya N, Fujii T, Itoh T, Ishibashi-Ueda H, Yamagashi M, Miyatake Ades E, Candal F, Swerlick R, George V, Summers S, Bosse D & Lawley T K, Matsumoto T, Kitamura S & Kangawa K 2005 Adrenomedullin 1992 HMEC-1: establishment of an immortalized human microvascular enhances angiogenic potency of bone marrow transplantation in a rat model endothelial cell line. Journal of Investigative Dermatology 99 683–690. of hindlimb ischemia. Circulation 111 356–362. Albertin G, Carraro G, Parnigotto P, Conconi M, Ziolkowska A, Jiang W,Yang JH, Wang SH, Pan CS, Qi YF,Zhao J & Tang CS 2004 Effect of Malendowicz L & Nussdorfer G 2003 Human skin keratinocytes and adrenomedullin on aldosterone-induced cell proliferation in rat cardiac fibroblasts express adrenomedullin and its receptors, and adrenomedullin fibroblasts. Biochimica et Biophysica Acta 1690 265–275. enhances their growth in vitro by stimulating proliferation and inhibiting Kim W,Moon S-O, Sung M, Kim S, Lee S, So J-N & Park S 2003 Angiogenic apoptosis. International Journal of Molecular Medicine 11 635–639. role of adrenomedullin through activation of Akt, mitogen-activated Anko¨ ML & Panula P 2006 Regulation of endogenous human NPFF2 protein kinase, and focal adhesion kinase in endothelial cells. FASEB Journal receptor by neuropeptide FF in SK-N-MC neuroblastoma cell line. 17 1937–1939. Journal of Neurochemistry 96 573–584. Kitamura K, Kangawa K, Kawamoto M, Ichiki Y, Nakamura S, Matsuo H & Bouı¨s D, Hospers GAP, Meijer C, Molema G & Mulder NH 2001 Eto T 1993 Adrenomedullin: a novel hypotensive peptide isolated Endothelium in vitro: a review of human vascular endothelial cell lines for from human pheochromocytoma. Biochemical and Biophysical Research blood vessel-related research. Angiogenesis 4 91–102. Communications 192 553–560. Caron K & Smithies O 2000 Extreme hydrops fetalis and cardiovascular Kuwasako K, Kitamura K, Onitsuka H, Uemura T, Nagoshi Y, Kato J & abnormalities in mice lacking a functional adrenomedullin gene. PNAS Tanenao E 2002 Rat RAMP domains involved in adrenomedullin binding 98 615–619. specificity. FEBS Letters 519 113–116. Champion HC, Santiago JA, Murphy WA, Coy DH & Kadowitz PJ 1997 Martinez A, Vos M, Guedez L, Kaur G, Chen Z, Garayoa M, Pio R, Moody T, Adrenomedullin-(22–52) antagonizes vasodilator responses to CGRP but Stetler-Stevenson WG, Kleinman HK & Cuttita F 2002 The effects of not adrenomedullin in the cat. American Journal of Physiology 272 adrenomedullin overexpression in breast tumour cells. Journal of the National R234–R242. Cancer Institute 94 1226–1237. Cobb MH, Hepler JE, Cheng M & Robbins D 1994 The mitogen-activated McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J, Thompson N, Solari R, protein kinases, ERK1 and ERK2. Seminars in Cancer Biology 5 261–268. Lee MG & Foord SM 1998 RAMPs regulate the transport and ligand binding Dupre D, Chen Z, Le Gouill C, Theriault C, Parent J, Rola-Pleszczynski M & specificity of the calcitonin-receptor-like receptor. Nature 393 333–339. Stankova J 2003 Trafficking, ubiquitination, and down-regulation of the Nagaya N, Goto Y, Satoh T, Sumida H, Kojima S, Miyatake K & Kangawa K human platelet-activating factor receptor. Journal of Biological Chemistry 278 2002 Intravenous adrenomedullin in myocardial function and energy 48228–48235. metabolism in patients after myocardial infarction. Journal of Cardiovascular Fernandez-Sauze S, Delfino C, Marbrouk K, Dussert C, Chinot O, Martin P-M, Grisoli F, Ouafik L & Boudouresque F 2004 Effects of Pharmacology 39 754–760. adrenomedullin on endothelial cells in the multistep process of Nikitenko LL, MacKenzie IZ, Rees MCP & Bicknell R 2000 Adrenome- angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 dullin is an autocrine regulator of endothelial growth in human receptors. International Journal of Cancer 108 797–804. endometrium. Molecular Human Reproduction 6 811–819. Froidevaux S & Eberle AN 2002 Homologous regulation of melanocortin-1 Nishikimi T, Horio T, Yoshihara F, Nagaya N, Matsuo H & Kangawa K 1998 receptor (MC1R) expression in melanoma tumour cells in vivo. Journal of Effect of adrenomedullin on cAMP and cGMP levels in rat cardiac Receptor Signalling Transduction Research 22 111–121. myocytes and nonmyocytes. European Journal of Pharmacology 353 337–344. Hague S, Zhang L, Oehler MK, Manek S, MacKenzie IZ, Bicknell R & Rees Oehler MK, Norbury C, Hague S, Rees MC & Bicknell R 2001 MC 2000 Expression of the hypoxically regulated angiogenic factor Adrenomedullin inhibits hypoxic cell death by up-regulation of Bcl-2 in adrenomedullin correlates with uterine leiomyoma vascular density. Clinical endometrial cancer cells: a possible promotion mechanism for tumour Cancer Research 3 2808–2814. growth. Oncogene 20 2937–2945. www.endocrinology-journals.org Journal of Endocrinology (2006) 190, 505–514

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access 514 N SCHWARZ and others $ AM elevates mRNA levels of CL and RAMP2

Oehler M, Hague S, Rees M & Bicknell R 2002 Adrenomedullin promotes Smith D, Coppock H, Withers D, Owji A, Hay D, Choski T, Chakravarty P, formation of xenografted endometrial tumors by stimulation of autocrine Legon S & Poyner D 2002 Adrenomedullin: receptor and signal growth and angiogenesis. Oncogene 21 2815–2821. transduction. Biochemical Society Transactions 30 432–437. Parameswaran N, Nambi P, Hall CS, Brooks DP & Spielman WS 2000 Unger RE, Krump-Konvalinkova V, Peters K & Kirkpatrick CJ 2002 In vitro Adrenomedullin decreases extracellular signal-regulated kinase activity expression of the endothelial phenotype: comparative study of primary through an increase in protein phosphatase-2A activity in mesangial cells. isolated cells and cell lines, including the novel cell line HPMEC-ST1.6R. European Journal of Pharmacology 388 133–138. Microvascular Research 64 384–397. Poyner D, Sexton P, Marshall I, Smith D, Quirion R, Born W, Muff R, Wang H, Tan X-d, Chang H, Gonzalez-Crussi F, Remick DG & Hsueh W Fischer J & Foord S 2002 The mammalian calcitonin gene-related peptides, 1997 Regulation of platelet-activating factor receptor gene expression adrenomedullin, amylin, and calcitonin receptors. Pharmacological Reviews in vivo by endotoxin, platelet activating factor and endogenous tumour 54 233–246. necrosis factor. Biochemical Journal 322 603–608. Roux PR & Blenis J 2004 ERK and p38 MAPK-activated protein kinases: Zachary I & Gliki G 2001 Signaling transduction mechanisms mediating a family of protein kinases with diverse biological functions. Microbiology and biological actions of the vascular endothelial growth factor family. Molecular Biology Reviews 68 320–344. Cardiovascular Research 49 568–581. Schanstra JP, Bataille E, Castano ME, Barascud Y, Hirtz C, Pesquero JB, Ziolkowska A, Budzynska K, Trejter M, Tortorella C, Belloni AS & Pecher C, Gauthier F, Girolami JP & Bascands JL 1998 The B1-agonist Malendowicz LK 2003 Effects of adrenomedullin and its fragment 22–52 [des-Arg10]-kallidin activates transcription factor NF-kappaB and induces on basal and ACTH-stimulated secretion of cultured rat adrenocortical homologous upregulation of the bradykinin B1-receptor in cultured cells. International Journal of Molecular Medicine 11 613–615. human lung ﬁbroblasts. Journal of Clinical Investigations 101 2080–2091. Segawa K, Minami K, Sata T, Kuroiwa A & Shigematsu A 1996 Inhibitory effect of adrenomedullin on rat mesangial cell mitogenesis. Nephron 74 577–579. Shindo T, Kurihara Y, Nishimatsu H, Moriyama N, Kakoki M, Wang Y, Imai Y, Ebihara A, Kuwaki T, Ju KH et al. 2001 Vascular abnormalities and Received 13 February 2006 elevated blood pressure in mice lacking adrenomedullin gene. Circulation Received in ﬁnal form 5 May 2006 104 1964–1971. Accepted 8 May 2006 Siegrist W, Stutz S & Eberle AN 1994 Homologous and heterologous regulation of alpha-melanocyte-stimulating hormone receptors in human Made available online as an Accepted Preprint and mouse melanoma cell lines. Cancer Research 54 2604–2610. 26 May 2006

Journal of Endocrinology (2006) 190, 505–514 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/26/2021 02:09:31AM via free access