Identification of HIV-1 Tat peptidesfor future therapeutic angiogenesis

Angiogenesis, the formation of new blood vesselsfrom pre-existing vessels, is a multi-step processinvolving a large number of extracellular andintracellular factors including c-Jun (1, 2) and theplasminogen activator inhibitor-1 (PAI-1). ThePAI-1 is produced by endothelial cells, inhibitsthe tissue-type and urokinase-type of plasminogenactivators (uPA) and is required for differentsteps of angiogenesis (3–5). The transcriptionfactors c-Jun and SP-1 are responsible for theinduction of PAI-1 (6, 7) and of the vascularendothelial growth factor (VEGF) (1), which isalso an important inductor of PAI-1 in endothel-ial cells (8). Two potential mechanisms explainwhy PAI-1 is essential for angiogenesis. First, byprotecting the extracellular matrix against exces-sive degradation, PAI-1 may serve to maintainthe matrix scaffold required for endothelial cellmigration and tube formation (8). Second, acomplex series of interactions have been describedbetween PAI-1, uPA, integrins, and the extra-cellular matrix component vitronectin, which

affects adhesive, migratory and growth propertiesof endothelial cells (9–11), a prerequisite for tubeformation.

A novel strategy for tumor therapy is theinhibition of tumor angiogenesis by appropriateagents [for rec. rev. see (12)]. On the other hand, theneed of pro-angiogenic drugs for therapeutic angi-ogenesis is obvious. For example, the chronicperipheral arterial occlusive disease (PAOD) ischaracterized by symptom progression from inter-mittent claudication to pain at rest and loss oftissue integrity with the development of ischemiculceration and gangrene (13). In more than one-third of these patients major amputation is required(14). Although the techniques of percutaneousinterventional and surgical revascularization arecontinuously being improved (15), there is still asignificant number of patients with end-stagePAOD that is not amenable to any of thesetherapies. Consequently, the need of alternativepharmacological treatment strategies for thesepatients is compelling.

Ismail M, Henklein P, Huang X, Braumann C, Ruckert RI, Dubiel W.Identification of HIV-1 Tat peptides for future therapeutic angiogenesis.

Abstract: Therapeutic angiogenesis represents a novel approach to treatcritical limb ischemia when revascularization is no more an option. Theclinical use of the vascular endothelial growth factor is questioned,because of its side effects. This study was designed to identify andcharacterize human immunodeficiency virus type 1 (HIV-1) Tat-derivedpeptides based on their pro-angiogenic properties. A series of Tat-derived peptides were synthesized containing mutations in the basicdomain. To minimize side effects Tat peptides were selected exerting noeffects on the proteasome and on the viability of human umbilical veinendothelial cells (HUVEC). Tatpep5, 15, and 16 increased theendogenous levels of the pro-angiogenic transcription factors c-Jun andSP-1 as well as the production of the plasminogen activator inhibitor-1(PAI-1) by HUVEC. A significant induction of endothelial cell invasionwas observed upon treatment of HUVEC with Tat peptides. In addition,selected Tat peptides induced tube formation by HUVEC as visualizedand quantified in a Matrigel matrix. Our data demonstrate that theselected Tat peptides fulfill essential criteria for pro-angiogenicsubstances. They represent the basis for the development of novel pro-angiogenic drugs for future therapeutic angiogenesis, which might beapplied for treatment of unreconstructible critical limb ischemia.

Mahmoud Ismail1,2, PeterHenklein3, Xiaohua Huang2, ChrisBraumann1, Ralph I. R�ckert4,Wolfgang Dubiel21Department of Surgery, 2Division of Molecular Biology,3Institute of Biochemistry, Charit�, Universit�tsmedizinBerlin, Berlin, Germany; 4Department of Surgery,Franziskus Hospital Berlin, Berlin, Germany

Key words: human immunodeficiency virus type 1 Tatpeptides; angiogenesis; plasminogen activator inhibitor-1; cell invasion; tubulogenesis; peripheral vasculardisease

Correspondence: Prof. Dr. Wolfgang Dubiel, Departmentof Surgery, Division of Molecular Biology, Charit�,Universit�tsmedizin Berlin, Monbijoustr. 2, 10117 Berlin,GermanyTel: +49-30-450522305Fax: +49-30-450522928e-mail: wolfgang.dubiel@charite.de

Accepted for publication 16 March 2006

Eur J Haematol 2006: 77: 157–165doi:10.1111/j.1600-0609.2006.00682.xAll rights reserved

� 2006 The AuthorsJournal compilation � 2006 Blackwell Munksgaard

EUROPEAN JOURNAL OF HAEMATOLOGY

157

A very interesting pro-angiogenic polypeptide isthe Tat protein of the human immunodeficiencyvirus type 1 (HIV-1). Tat is a transactivatingfactor for viral and some host cell genes (16). Itspro-angiogenic properties (17) are relevant for thepathogenesis of Kaposi’s sarcoma (18), an angio-proliferative disease, which is very frequent andaggressive, if associated with HIV-1 infection. Tatcan be released by infected cells in the microen-vironment modulating HIV-1 infection (19) andmay affect the function of immunocompetent andmesenchymal cells (16). On endothelial cells, Tatmimics the effect of VEGF by activating theangiogenic program after interaction with the Flk-1/KDR receptor (20, 21). Tat binds VEGF recep-tors at new sites that might be useful targets forpharmacological intervention and has also beenshown to synergize with basic fibroblast growthfactor (bFGF) in inducing endothelial cell growthand promoting angiogenesis. The amino acids 49–57 constitute the basic domain of Tat, which ismostly responsible for the pro-angiogenic effect(18).

In addition, Tat has impact on the ubiquitin/proteasome system, the major proteolytic pathwayin eukaryotic cells. It blocks the activity of the 20Sproteasome and competes with the 11S regulator(22), which reduces antigen presentation after virusinfection (23, 24).

Therefore, to alter Tat into an ideal pro-angiogenic factor it must keep its ability toinitiate the complete program of angiogenesis,but it should have no impact on antigen presen-tation. For that aim we studied the effects of anumber of Tat peptides with pro-angiogenicproperties. We modified the basic domain ofTat in a way that the resulting peptides did notinhibit the proteasome anymore. We identifiedTat-derived peptides that stimulated the produc-tion of PAI-1, enhanced the migration of endot-

helial cells and induced tubulogenesis in aMatrigel matrix.

Materials and methods

Synthesis of Tat protein and derived peptides

The HIV-1 Tat protein and Tat derived peptidesshown in Table 1 were synthesized using the FMOCstrategy on a 433A peptide synthesizer (ABI). Full-lengthTat proteinwas obtained from theNIHAIDSReagent Program (contributor Dr. J. Brady).

Kinetic studies with the 20S proteasome

The 20S proteasome is the core particle of the 26Sproteasome involved in the production of antigenicpeptides presented byMHCclass-Imolecules (23). Itwas isolated from red blood cells as described (25).To test 20S proteasome inhibition, 1 lg/mL ofisolated 20S proteasome was incubated at 37�C with100 lm Suc-Leu-Leu-Val-Tyr-AMC (Bachem, Ger-many) as substrate in a final volume of 100 lL. Thefluorescence was measured at 37�C with a microtiterplate reader (Fluoroscan II, Labsystems, Germany)at 355 nm excitation and 460 nm emissions over a60-min period in 5-min intervals. During this period,the reaction was linear. Tat peptides were added atfinal concentrations indicated.

Cell culture

A human umbilical vein endothelial cell (HUVEC)line was established (Cell-Lining, Germany) andcultured in an endothelial cell culture medium(PromoCell, Germany) at 37�C in a humidified 5%CO2 atmosphere. Cells were passaged at confluencefollowing dissociation with 0.05% trypsin (Invitro-gen, Germany). Cultured HUVEC were usedbetween passages four and nine.

Table 1. Sequences of synthetic HIV-1 Tat-derivedpeptides. The full-length Tat protein consists of 87amino acids. Residues 49–57 form the basic domain ofthe Tat protein (24). Tatpep1 and 5 consist of aminoacids 37–72. In Tatpep5 Lys51, Arg52, and Asp67 weremutated to Ala. Tatpep6 represents a small fragmentof the basic domain, whereas Tatpep7 contains thecomplete basic domain and few flanking amino acids.Tatpep8–18 possess two mutations to Ala in differentpositions. Plus and minus indicates the effect on the20S proteasome peptidase activity (20S)

37 72 Inhibition of 20S

Tatpep1 CFITKGLGIS YGRKKRRQRR RPSQGGQTHQ DPIPKQ +Tatpep5 CFITKGLGIS YGRKAARQRR RPSQGGQTHQ APIPKQ )Tatpep6 RR RPSQ +Tatpep7 GRKKRRQRR RPSQG +Tatpep8 GAAKRRQRR RPSQG +Tatpep9 GRAARRQRR RPSQG +Tatpep10 GRKAARQRR RPSQG +Tatpep11 GRKKAAQRR RPSQG +Tatpep12 GRKKRAARR RPSQG +Tatpep13 GRKKRRAAR RPSQG +Tatpep14 GRKKRRQAA RPSQG +Tatpep15 GRKKRRQRA APSQG )Tatpep16 GRKKRRQRR AASQG )Tatpep17 GRKKRRQRR RAAQG +Tatpep18 GRKKRRQRR RPAAG +

Ismail et al.

158

Western blots

The HUVEC were harvested and lysed in buffercontaining 20 mm Tris, pH 8.0, 150 mm NaCl,0.5% Na-desoxycholat, 1.0% nonidet P-40, 0.1%SDS, 5.0% glycerol, 1 mm EGTA, 100 lm phenyl-methylsulfonyl fluoride, and 4 lm aprotinin. Mem-branes were disrupted by passing the cellsuspension through a 26-G needle. After centrifu-gation aliquots of the supernatants were separatedby 12.5% SDS–PAGE. Proteins were blotted tonitrocellulose and probed with anti-c-Jun (Onco-gene, Germany), anti-SP-1 (Geneka, Canada), andanti-b-tubulin (Covance, USA). Western blots weredeveloped using the ECL method (AmershamBiosciences, Germany).

Cell viability assay

Cell viability was assessed by the 3-[4,5-dimethylthi-azol-2-yl]-2,5-diphenyltetrazolium bromide (MTT)method as recommended by the manufacturer (Sig-ma-Aldrich, Germany). Cells were incubated withTat-derived peptides (2 lg/mL each) or curcumin(10 lm) (Sigma-Aldrich) for 24 h. Then a solution ofMTT in phosphate-buffered saline (PBS) was addedto each well to a final concentration of 0.5 lg/lL.After 4 h incubation purple formazan was solubi-lized with 100 lL DMSO. Absorbance was meas-ured at 570 nm with background subtraction at620 nm using an ELISA reader.

Quantification of PAI-1 production

TheHUVECwere incubated with Tat protein (2 lg/mL) or Tat peptides (2 lg/mL each) for 24 h. ThePAI-1 concentrations in cell culture supernatantswere measured using a PAI-1 ELISA kit (AmericanDiagnostic, Germany). In short, 1 mL of cell culturesupernatant was centrifuged at 15 000 g and 4�C for5 min and the supernatant was stored at )70�C forup to 72 h. The next steps were performed accordingto the manufacturer’s instructions. The basic PAI-1production amounted to 993.8 ± 11.6 ng/mL andwas subtracted from all data.

Matrigel invasion assay

Invasion of HUVEC was determined using modi-fied Boyden chamber technology. Transwells withpolycarbonate membranes (8 lm pore size) insix-well tissue culture plates (Corning Costar,Germany) were precoated with Growth FactorReduced Matrigel Basement Membrane Matrix(Becton Dickinson, Germany) diluted in serum freemedium (1 mg/mL; 675 lL per transwell) and wereincubated for 2 h at 37�C as recommended by the

producer (Becton Dickinson). Thereafter, tran-swells were rinsed with serum free medium andHUVEC were seeded at a concentration of 300 000cells per mL (1.5 mL total) into the transwell. Tatpeptides (2 lg/mL) or bFGF (2 ng/mL) were addedto the upper compartment. Curcumin (10 lm) wasadded to the upper and lower compartment. After24 h, cultures were incubated with 0.5 mg/mL ofMTT for 4 h as described elsewhere (26). Invasivecells attached to the bottom of the polycarbonatemembrane were scraped off and rinsed into thelower tissue culture well with DMSO. After that,non-invasive cells in the transwell were dissolvedwith DMSO. Formazan solutions were transferredin a 96-well microplate and absorbance was meas-ured at 570 nm with background subtraction at620 nm using an ELISA reader. Invasion wasexpressed as percentage of cell invasion comparedwith DMSO treated cells as control.

Matrigel assay of tubulogenesis

Tubulogenesis was assayed as described by Moraleset al. with few modifications (27). In brief, GrowthFactor Reduced Matrigel Basement MembraneMatrix (Becton Dickinson) was thawed on iceovernight and then spread evenly over each well ofa 24-well plate. The plates were incubated at 37�C.After 30 min HUVEC were seeded (25 · 103 cells/cm2) and cultured in basal medium (PromoCell) inthe absence or in the presence of Tat-peptides (2 lg/mL each) or bFGF (2 ng/mL) for 24 h. Incubationwith curcumin (10 lm), served as negative control.The formation of tubes by HUVEC suspended inmatrigel was assessed under an Olympus CK40Inverted Laboratory System Microscope (at magni-tude ·10) and photographed with a C2020Z Olym-pus Digital Camera. Triplicates were used in allexperiments and each experiment was repeated atleast three times. For quantification of tube forma-tion the number of branch points of the formed tubeswas counted as in (28).

Statistical analysis

All data shown are from at least three independenttriplet experiments and are expressed asmean ± SE. The statistical significance was deter-mined using Student’s t-test. P-values smaller than0.05 were considered as significant.

Results

Selection of Tat peptides

Our aim was to identify Tat peptides, which areeffective in stimulating angiogenesis but have a

HIV-1 Tat for therapeutic angiogenesis

159

minimum of possible side effects. We have shownthat the full-length Tat protein and a number ofTat-derived peptides containing the basic domaininhibit the 20S proteasome (22). Moreover, thesepeptides were able to compete with the 11Sregulator of the 20S proteasome and, therefore, tosuppress the production of antigens presented byMHC class-I molecules (24). These features of Tatpeptides might cause severe side effects and shouldbe eliminated before use in therapy. On the otherhand, the basic domain of Tat is important forstimulation of angiogenesis (20). Therefore, a seriesof Tat-derived peptides was synthesized (seeTable 1), which contain a mutated version of thebasic domain. Tat-derived peptides were modifiedby introducing two alanine residues into differentpositions. First, the impact of the Tat-derivedpeptides on the peptidase activity of the 20Sproteasome was evaluated. Selected data are shownin Fig. 1. As demonstrated before, Tatpep1 reducedthe cleavage of Suc-Leu-Leu-Val-Tyr-AMC by theproteasome, whereas Tatpep5 did not (24). Inaddition, Tatpep15 and 16 did not inhibit theproteasome activity (Fig. 1). All other peptidesshown in Table 1 (Tatpep6–14 and 17, 18) more orless effectively blocked the 20S proteasome activity(data not shown) and, therefore, were eliminatedfrom further experiments.

Impact of selected Tat peptides on cell viability

To test the impact on cell viability, endothelial cells(HUVEC) were incubated for 24 h with Tatpep5,15, and 16 as well as with curcumin (10 lm).Curcumin is a kinase inhibitor (29) and inducesapoptosis under the conditions used (30, 31). It is

also an inhibitor of angiogenesis (32). Therefore, inour experiments curcumin was used as a negativecontrol. As shown in Fig. 2, Tatpep5, 15, and 16did not show inhibitory effects on the growth ofHUVEC. In contrast, curcumin significantly re-duced cell growth as compared with the untreatedcontrol cells. Therefore, we expected that theselected Tat peptides have neither toxic nor apop-totic effects under our conditions.

Selected Tat peptides exert stimulatory effectson the pro-angiogenic factors c-Jun, SP-1, and PAI-1

The expression of the pro-angiogenic transcriptionfactors c-Jun and SP-1 was estimated by Westernblotting (Fig. 3A). The HUVEC were incubatedwith different concentrations of Tat peptides. After24 h cells were harvested, lysed and lysates wereanalyzed by immunoblotting. It is shown inFig. 3A, upper panel, that increasing amounts ofTatpep15 significantly elevated the amount ofendogenous c-Jun. The same effect is less pro-nounced with Tatpep5 or 16. In addition, Tatpep15clearly stimulated the expression of SP-1, whereasTatpep5 and 16 were less effective (Fig. 3A, middlepanel). As a loading control b-tubulin was tested(Fig. 3A, lower panel).

A very important regulator of angiogenesis in theextracellular matrix is PAI-1 (3–5). Therefore, wedetermined the PAI-1 concentration in HUVECsupernatants. PAI-1 protein was produced andreleased by these cells. HUVEC were incubatedwith full-length Tat protein, Tatpep5, 15, and 16.Tatpep5 as well as Tatpep15 were approximatelytwo times more effective in stimulating PAI-1

Fig. 1. Effects of selected Tat peptides on peptidase activityof the 20S proteasome. Peptidase activity was measured asDF (F: fluorescence) per minute and is expressed as relativepeptidase activity (100% ¼ 20 DF/min per lg of 20S pro-teasome) using Suc-Leu-Leu-Val-Tyr-AMC as substrate.Curves were calculated using the model described earlier(24). The data are representative of four independentexperiments.

Fig. 2. Effect of Tat peptides on HUVEC viability. A totalof 200 000 cells were plated in six-well plates and grown for24 h in the presence of Tatpep5, 15, and 16 (2 lg/mL).Untreated cells were used as control and curcumin (10 lm),which is known to stimulate apoptosis, was also tested un-der our conditions. The bFGF (2 ng/mL) was used as apositive control. After cell treatment the MTT test wasperformed as described in Materials and methods. Cellviability is expressed in % (untreated control cells: 100%).

Ismail et al.

160

production by HUVEC as compared with Tat-pep16 and to full-length Tat (Fig. 3B). We conclu-ded that Tatpep5 and 15 are presumably goodcandidates for use in pro-angiogenic therapy.

Tatpep5, Tatpep15, and Tatpep16 induce HUVEC invasion

Endothelial cell invasion and migration are vitalsteps in angiogenesis. To form new blood vesselsinvasive endothelial cells must degrade and traversebasement membrane and connective tissue (33).This event was studied using the Boyden chamber.Filters were coated with growth factor reducedMatrigel, which prevents migration of non-invasivecells. As a positive control, HUVEC were treatedwith the well-known pro-angiogenic factor bFGF.Curcumin, the kinase inhibitor that blocks angio-genesis (32), was used as a negative control. TheHUVEC were treated with 2 lg/mL of Tatpep5, 15

or 16. As demonstrated in Fig. 4, the selected Tatpeptides increased the number of invasive cellssignificantly. Tatpep5 and 15 were almost aseffective as bFGF. As expected, curcumin is aninhibitor of HUVEC invasion.

Stimulation of tubulogenesis by Tat-derived peptides

A very important question was whether Tat-derivedpeptides are able to stimulate tube formation. Toexamine the role of our synthesized Tat peptides inthe process of tubulogenesis HUVEC were culturedin 24-well plates and tube formation was visualizedin a Matrigel matrix (see Materials and methods).Tatpep5, 15 and 16 were found to enhance forma-tion of capillary-like tubes on Matrigel (seeFig. 5C–E, respectively). Tube formation with ourselected Tat peptides was stimulated to a similarextend as that with bFGF (Fig. 5A) in comparisonto the control (Fig. 5F). As expected, curcumin wasan inhibitor of tube formation (Fig. 5B). Countingof branch points in A–F as described before (28),led to the data shown in Fig. 5G. AccordinglybFGF stimulated basic tube formation by HUVECtwo times, whereas Tatpep15 increased it even threetimes.

Discussion

Effective and safe therapeutic angiogenesis wouldbe hope for patients with critical ischemia due to

Fig. 3. Induction of the pro-angiogenic factors c-Jun, SP-1,and PAI-1 by Tat peptides. (A) After treatment of HUVECfor 24 h with indicated amounts of Tat peptides westernblot analyses were performed. Cells were lysed and proteinswere separated by SDS–PAGE, blotted to nitrocellulose andprobed with an anti-c-Jun or an anti-SP-1 antibody. As aloading control b-tubulin was tested (lower panel). (B) PAI-1 concentration was determined in HUVEC supernatantsafter 24-h treatment with Tat protein (2 lg/mL) or Tatpeptides (2 lg/mL) as described in Materials and methods.The basal PAI-1 production amounted to 993.8 ± 11.6 ng/mL and was subtracted from all data. *P < 0.001 vs. Tatprotein (n ¼ 4).

Fig. 4. Enhancement of invasion and migration of HUVECin the Matrigel invasion assay by Tatpep5, 15, and 16. Cellswere cultured on transwell cell culture inserts coated withMatrigel. The HUVEC were treated for 24 h with bFGF(2 ng/mL) as a positive control. In addition, curcumin wasused as a negative control. Tatpep5, 15, and 16 were em-ployed at a final concentration of 2 lg/mL. After 24 h cellswere incubated with MTT and optical density was measuredafter dissolving cells with DMSO. Optical density data ofinvasive cells were converted into cell numbers on the basisof a standard curve and then transferred into percentage(number of cells at the beginning: 100%). The controls areuntreated cells, which were incubated as treated cells. Dataof three (Tatpep16) or seven experiments are expressed asmean ± SE. *P < 0.05 vs. control.

HIV-1 Tat for therapeutic angiogenesis

161

end-stage vascular diseases that are not amenableto any traditional therapies (15, 33). Unfortunately,although first clinical trials with the growth factorVEGF were promising (34, 35), there are certaindrawbacks that limit the clinical use of VEGF forsafety concerns (15, 36). The VEGF stimulated theformation of atherosclerotic plaques in animalexperiments (37, 38) and propagation of tumorangiogenesis is likely. Therefore, new approachesare required.

Synthesis of HIV-1 Tat-derived peptides for therapeuticangiogenesis

Our approach is to employ pro-angiogenic featuresof viral proteins for therapeutic application. Theangiogenic activity of the HIV-1 Tat protein is wellknown (16, 18, 21). The effect might be, in part,explained by Tat interaction with integrins (39)and/or with Tat binding to the VEGF receptor-2/KDR (20). The Tat protein can penetrate cellmembranes (40), which has been used in Tat fusionproteins that deliver large proteins into cells (41,42). All these effects can be due to the basic domain

of the HIV-1 Tat protein (18, 40). Therefore, oneshould be cautious with Tat fusion proteins, sincethe Tat domain used for protein shuttling containsthe pro-angiogenic basic domain (43). In addition,the basic domain is also involved in the immuno-suppressive activity of the viral protein. Tatpep1(see Table 1) inhibits the proteasome and causes areduced antigen presentation by MHC class-Imolecules (24).

Our aim was to identify Tat peptides with pro-angiogenic activities, which have no impact onthe proteasome. Inhibition of the proteasomedoes not only influence antigen processing, it alsoblocks angiogenesis (44). It impairs cell prolifer-ation and signal transduction (45), processesnecessary for the angiogenic program. Tatpep7consisting of amino acids 48–61 containing thebasic domain plus few flanking amino acidsinhibited the peptidase activity of the 20S pro-teasome to a similar extend as that shown forTatpep1 (see Fig. 1). Tatpep6 consisting of only 6amino acids (Tat amino acids 55–60) also reducedproteasome activity (data not shown). Because oftheir different sizes we speculate that Tatpep1

(A) (B) (C)

(D)

(G)

(E) (F)

Fig. 5. Selected Tat-derived peptidesinduce tube formation by HUVEC.The HUVEC (25 · 103 cells/cm2) wereincubated overnight at 37�C on aMatrigel matrix in the presence of: (A)bFGF (2 ng/mL) as positive control,(B) curcumin (10 lm) as negative con-trol, (C) Tatpep5, (D) Tatpep15, (E)Tatpep16, and (F) as an additionalcontrol, with basal medium alone.Tube formation by HUVEC was sig-nificantly induced by bFGF (A) and ina similar manner by the Tat peptides(C–E) compared with the spontaneoustube formation by HUVEC (F). Incontrast, curcumin blocked tube for-mation almost completely (B). (G) Forquantification after 24 h treatment thenumber of branch points was counted(28) in A–F. The results are expressedas mean ± SE.

Ismail et al.

162

blocks the 20S proteasome by a different mech-anism as Tatpep6 or 7. To avoid inhibition of theproteasome 11 versions of Tatpep7 were synthes-ized with two alanine residues in different posi-tions (see Table 1, Tatpep8–18). The Tat-derivedpeptides 15 and 16 did not show any influence onproteasome peptidase activity (Fig. 1). In thesepeptides amino acids 56 and 57 (Tatpep15) or 57and 58 (Tatpep16) were changed to alanine. Incontrast, in Tatpep5 amino acids 51, 52, and 67are important for proteasome inhibition. Thus,we have identified three Tat-derived peptides,Tatpep5, 15, and 16, which do not inhibit the20S proteasome presumably because of differentreasons.The selected Tat peptides had no significant

impact on the viability of HUVEC as demonstratedby the MTT test. In contrast, curcumin (10 lm),which is known to induce cell death already after12 h (30, 31, 46), reduced the number of cells byapproximately 60% (see Fig. 2).

Selected Tat-derived peptides induce tubulogenesis

The transcription factors c-Jun and SP-1 areimportant regulators in the angiogenic program.Both induce the production of the most potentpro-angiogenic factor, the VEGF (1). Moreover,the production of PAI-1 is also accelerated in thepresence of these transcription factors (6, 7).After treatment of HUVEC with Tatpep15 thereis a significant increase of cellular c-Jun as well asSP-1. The effect is less obvious, but well detect-able, with Tatpep5 and 16 (see Fig. 3A). Anessential regulator during the angiogenic programis PAI-1 (3–5). For the first time we show thatthe PAI-1 production by HUVEC is stimulatedby the Tat protein and by Tat-derived peptides(see Fig. 3B). Tatpep5 and 15 were two timesmore effective in stimulating PAI-1 as comparedwith the Tat protein. This might be due todifferent effects the two peptides exert on theangiogenic program as compared with the full-length protein.Cell migration and invasion are complex events

involving factors like PAI-1, matrix metallopro-teases, VEGF, and others. To form new bloodvessels invasive endothelial cells must degrade andmigrate through connective tissue (33, 47). Pro-angiogenic drugs should have the ability totransfer endothelial cells into invasive cells.Therefore, our Tat peptides were tested withHUVEC in a Matrigel invasion assay. Thegrowth factor bFGF, known to stimulate cellmigration, activated HUVEC by two times ascompared with control. The selected Tat peptides,Tatpep5, 15, and 16, were almost as effective as

bFGF, whereas curcumin, the inhibitor of angi-ogenesis (32), reduced the number of invasivecells significantly.

An even more important evidence for the abilityof our selected Tat peptides to induce the angio-genic program is their potency to enhance tubulo-genesis. Our qualitative and quantitative analyseswith the Matrigel matrix revealed that Tatpep5, 15,and 16 are able to stimulate tube formation just likebFGF. Tatpep15 is even more effective than bFGF(see Fig. 5G). This is a very strong support of ourhypothesis that selected Tat-derived peptides arepro-angiogenic factors.

Although the impact of the selected Tat peptideson angiogenesis has to be tested in an appropriateanimal model, our data demonstrate that Tatpep5,15, and 16 are candidate drugs for use in therapeu-tic angiogenesis. Because of their size and abilitiesTatpep15 and 16 seem to be most suitable forfuture clinical application. The development ofdrugs for future therapeutic angiogenesis on thebasis of viral Tat peptides will open new perspec-tives for the treatment of ischemic vasculardiseases.

Acknowledgements

We thank Dr. J. M. Muller, the Director of the Department ofSurgery, Charite, Universitatsmedizin Berlin, for continuoussupport.

References

1. Pollmann C, Huang X, Mall J, Bech-Otschir D,Naumann M, Dubiel W. The constitutive photomorph-ogenesis 9 signalosome directs vascular endothelialgrowth factor production in tumor cells. Cancer Res2001;61:8416–8421.

2. Zhang G, Dass CR, Sumithran E, Di Girolamo N, SunLQ, Khachigian LM. Effect of deoxyribozymes targetingc-Jun on solid tumor growth and angiogenesis in rodents.J Natl Cancer Inst 2004;96:683–696.

3. McMahon GA, Petitclerc E, Stefansson S, Smith E,Wong MK, Westrick RJ, Ginsburg D, Brooks PC,Lawrence DA. Plasminogen activator inhibitor-1 regulatestumor growth and angiogenesis. J Biol Chem2001;276:33964–33968.

4. Vogten JM, Reijerkerk A, Meijers JC, Voest EE, BorelRinkes IH, Gebbink MF. The role of the fibrinolytic sys-tem in corneal angiogenesis. Angiogenesis 2003;6:311–316.

5. Bajou K, Maillard C, Jost M, Lijnen RH, Gils A,Declerck P, Carmeliet P, Foidart JM, Noel A. Host-derived plasminogen activator inhibitor-1 (PAI-1) concen-tration is critical for in vivo tumoral angiogenesis andgrowth. Oncogene 2004;23:6986–6990.

6. Chen YQ, Su M, Walia RR, Hao Q, Covington JW,Vaughan DE. Sp1 sites mediate activation of the plasmi-nogen activator inhibitor-1 promoter by glucose in vascularsmooth muscle cells. J Biol Chem 1998;273:8225–8231.

7. Arts J, Grimbergen J, Toet K, Kooistra T. On the roleof c-Jun in the induction of PAI-1 gene expression byphorbol ester, serum, and IL-1alpha in HepG2 cells.Arterioscler Thromb Vasc Biol 1999;19:39–46.

HIV-1 Tat for therapeutic angiogenesis

163

8. Pepper MS, Ferrara N, Orci L, Montesano R. Vascularendothelial growth factor (VEGF) induces plasminogenactivators and plasminogen activator inhibitor-1 in micro-vascular endothelial cells. Biochem Biophys Res Commun1991;181:902–906.

9. Loskutoff DJ, Curriden SA, Hu G, Deng G. Regulationof cell adhesion by PAI-1. APMIS 1999;107:54–61.

10. Ossowski L, Aguirre-Ghiso JA. Urokinase receptor andintegrin partnership: co-ordination of signaling for celladhesion, migration, and growth. Curr Opin Cell Biol2000;12:613–620.

11. Preissner KT,Kanse SM,May AE. Urokinase receptor: amolecular organizer in cellular communication. Curr OpinCell Biol 2000;12:621–628.

12. Gasparini G, Longo R, Toi M, Ferrara N. Angiogenicinhibitors: a new therapeutic strategy in oncology. Nat ClinPract Oncol 2005;2:562–577.

13. Ouriel K. Peripheral arterial disease. Lancet2001;358:1257–1264.

14. Luther M, Lepantalo M, Alback A, Matzke S. Ampu-tation rates as a measure of vascular surgical results. Br JSurg 1996;83:241–244.

15. Baumgartner I, Schainfeld R, Graziani L. Manage-ment of peripheral vascular disease. Annu Rev Med2005;56:249–272.

16. Rubartelli A, Poggi A, Sitia R, Zocchi MR. HIV-I Tat:a polypeptide for all seasons. Immunol Today 1998;19:543–545.

17. Albini A, Barillari G, Benelli R, Gallo RC, Ensoli B.Angiogenicpropertiesofhumanimmunodeficiencyvirustype1 Tat protein. Proc Natl Acad Sci USA 1995;92:4838–4842.

18. Barillari G, Ensoli B. Angiogenic effects of extracellularhuman immunodeficiency virus type 1 Tat protein and itsrole in the pathogenesis of AIDS-associated Kaposi’s sar-coma. Clin Microbiol Rev 2002;15:310–326.

19. Marchio S, Alfano M, Primo L, et al. Cell surface-asso-ciated Tat modulates HIV-1 infection and spreadingthrough a specific interaction with gp120 viral envelopeprotein. Blood 2005;105:2802–2811.

20. Albini A, Soldi R, Giunciuglio D, et al. The angiogenesisinduced by HIV-1 tat protein is mediated by the Flk-1/KDR receptor on vascular endothelial cells. Nat Med1996;2:1371–1375.

21. Rusnati M, Presta M. HIV-1 Tat protein and endo-thelium: from protein/cell interaction to AIDS-associatedpathologies. Angiogenesis 2002;5:141–151.

22. Seeger M, Ferrell K, Frank R, Dubiel W. HIV-1 tatinhibits the 20S proteasome and its 11S regulator-mediatedactivation. J Biol Chem 1997;272:8145–8148.

23. Kloetzel PM. Generation of major histocompatibilitycomplex class I antigens: functional interplay betweenproteasomes and TPPII. Nat Immunol 2004;5:661–669.

24. Huang X, Seifert U, Salzmann U, Henklein P, Preiss-ner R, Henke W, Sijts AJ, Kloetzel PM, Dubiel W. TheRTP site shared by the HIV-1 Tat protein and the 11Sregulator subunit alpha is crucial for their effects on pro-teasome function including antigen processing. J Mol Biol2002;323:771–782.

25. Seeger M, Kraft R, Ferrell K, Bech-Otschir D,Dumdey R, Schade R, Gordon C, Naumann M, Dubiel

W. A novel protein complex involved in signal transductionpossessing similarities to 26S proteasome subunits. FASEBJ 1998;12:469–478.

26. Imamura H, Takao S, Aikou T. A modified invasion-3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bro-mide assay for quantitating tumor cell invasion. Cancer Res1994;54:3620–3624.

27. Morales DE,McGowan KA,Grant DS,Maheshwari S,Bhartiya D, Cid MC, Kleinman HK, Schnaper HW.Estrogen promotes angiogenic activity in human umbilical

vein endothelial cells in vitro and in a murine model. Cir-culation 1995;91:755–763.

28. Kok TW, Yue PY, Mak NK, Fan TP, Liu L, Wong RN.The anti-angiogenic effect of sinomenine. Angiogenesis2005;8:3–12.

29. Bech-Otschir D, Seeger M, Dubiel W. The COP9 sig-nalosome: at the interface between signal transduction andubiquitin-dependent proteolysis. J Cell Sci 2002;115:467–473.

30. Liontas A, Yeger H. Curcumin and resveratrol induceapoptosis and nuclear translocation and activation of p53in human neuroblastoma. Anticancer Res 2004;24:987–998.

31. Fullbeck M, Huang X, Dumdey R, Frommel C, Dubiel

W, Preissner R. Novel curcumin- and emodin-relatedcompounds identified by in silico 2D/3D conformerscreening induce apoptosis in tumor cells. BMC Cancer2005;5:97.

32. Arbiser JL, Klauber N, Rohan R, van Leeuwen R,Huang MT, Fisher C, Flynn E, Byers HR. Curcumin isan in vivo inhibitor of angiogenesis. Mol Med 1998;4:376–383.

33. Baumgartner I, Isner JM. Somatic gene therapy in thecardiovascular system. Annu Rev Physiol 2001;63:427–450.

34. Baumgartner I, Pieczek A, Manor O, Blair R, Kear-

ney M, Walsh K, Isner JM. Constitutive expression ofphVEGF165 after intramuscular gene transfer promotescollateral vessel development in patients with critical limbischemia. Circulation 1998;97:1114–1123.

35. Vale PR, Isner JM, Rosenfield K. Therapeutic angio-genesis in critical limb and myocardial ischemia. J IntervCardiol 2001;14:511–528.

36. Isner JM, Vale PR, Symes JF, Losordo DW. Assessmentof risks associated with cardiovascular gene therapy inhuman subjects. Circ Res 2001;89:389–400.

37. Celletti FL, Hilfiker PR, Ghafouri P, Dake MD. Ef-fect of human recombinant vascular endothelial growthfactor165 on progression of atherosclerotic plaque. J AmColl Cardiol 2001;37:2126–2130.

38. Celletti FL, Waugh JM, Amabile PG, Brendolan A,Hilfiker PR, Dake MD. Vascular endothelial growthfactor enhances atherosclerotic plaque progression. NatMed 2001;7:425–429.

39. Barillari G, Sgadari C, Fiorelli V, et al. The Tat pro-tein of human immunodeficiency virus type-1 promotesvascular cell growth and locomotion by engaging the al-pha5beta1 and alphavbeta3 integrins and by mobilizingsequestered basic fibroblast growth factor. Blood1999;94:663–672.

40. Vives E, Brodin P, Lebleu B. A truncated HIV-1 Tatprotein basic domain rapidly translocates through theplasma membrane and accumulates in the cell nucleus. JBiol Chem 1997;272:16010–16017.

41. Diem R, Taheri N, Dietz GP, Kuhnert A, Maier K,Sattler MB, Gadjanski I, Merkler D, Bahr M. HIV-Tat-mediated Bcl-XL delivery protects retinal ganglion cellsduring experimental autoimmune optic neuritis. NeurobiolDis 2005;20:218–226.

42. Kilic E, Kilic U, Hermann DM. TAT fusion proteinsagainst ischemic stroke: current status and future perspec-tives. Front Biosci 2006;11:1716–1721.

43. Fawell S, Seery J, Daikh Y, Moore C, Chen LL, Pe-pinsky B, Barsoum J. Tat-mediated delivery of heterolo-gous proteins into cells. Proc Natl Acad Sci USA1994;91:664–668.

44. Daniel KG, Kuhn DJ, Kazi A, Dou QP. Anti-angiogenicand anti-tumor properties of proteasome inhibitors. CurrCancer Drug Targets 2005;5:529–541.

45. Glickman MH, Ciechanover A. The ubiquitin-protea-some proteolytic pathway: destruction for the sake ofconstruction. Physiol Rev 2002;82:373–428.

Ismail et al.

164

46. Bech-Otschir D, Kraft R, Huang X, Henklein P,Kapelari B, Pollmann C, Dubiel W. COP9 signalosome-specific phosphorylation targets p53 to degradation by theubiquitin system. EMBO J 2001;20:1630–1639.

47. Bello L, Giussani C, Carrabba G, Pluderi M, Costa F,Bikfalvi A. Angiogenesis and invasion in gliomas. CancerTreat Res 2004;117:263–284.

HIV-1 Tat for therapeutic angiogenesis

165