HIV1 tat protein induces the production of interleukin-8 by human brain-derived endothelial cells

Ž .Journal of Neuroimmunology 94 1999 28–39

HIV-1 tat protein induces the production of interleukin-8 by humanbrain-derived endothelial cells

Florence M. Hofman a,), Peijia Chen a, Francesca Incardona a, Raphael Zidovetzki b,David R. Hinton a

a Department of Pathology, HMR 312, UniÕersity of Southern California, School of Medicine, 2011 Zonal AÕenue, Los Angeles, CA 90033, USAb Department of Cell Biology and Neuroscience, UniÕersity of California, RiÕerside, CA 92521, USA

Received 9 June 1998; revised 4 August 1998; accepted 11 August 1998

Abstract

Ž .This study focused on the role of the HIV-derived viral protein, tat, in activating central nervous system CNS -derived endothelialŽ . Ž .cells EC to produce interleukin-8 IL-8 , a stimulator and chemoattractant for neutrophils and lymphocytes. Human CNS-EC treated

Ž . Ž .with tat 100 ngrml demonstrated a 2 to 3 fold upregulation in IL-8 mRNA and protein. Tumor necrosis factor-a TNF and tat wereŽ .found to act additively in upregulating IL-8 production. In contrast, transforming growth factor b TGF b , appeared to down modulate

tat-induced IL-8 production. These data suggest that extracellular tat, especially in the presence of TNF, may be responsible for the localproduction of IL-8. q 1999 Elsevier Science B.V. All rights reserved.

Keywords: AIDSrHIV; Tat; IL-8; Endothelial cells; Brain

1. Introduction

HIV-infected patients have been shown to express highlevels of inflammatory cytokines in their serum and cere-

Ž . Žbral spinal fluid CSF Breen et al., 1990; Kobayshi et al.,.1990 . HIV-infected cells, particularly macrophages, se-

crete an array of cytokines, as well as virally derivedŽ .factors Fauci, 1988 . One of these HIV-derived viral

proteins, tat, has been shown to be secreted from virallyinfected cells and to function as a potent regulator of cell

Ž .activity Ensoli et al., 1990, 1993 . This viral protein, 86amino acids with a molecular mass of 15.5 kD, can betaken up by neighboring non-infected cells, and subse-

Žquently transported to the nucleus Frankel and Pabo,.1988 ; tat can also act as an extracellular agent, and has

been reported to alter cell function in a variety of cell

Abbreviations: CSF, cerebral spinal fluid; CNS, central nervous sys-tem; EC, endothelial cells; RPE, retinal pigmented endothelial cell; RPA,RNase protection assay; GAPDH, glyceraldehyde-3-phosphate dehydro-genase; PMN, polymorphonuclear cell; MEC, microvessel endothelialcell; BBB, blood-brain-barrier

) Corresponding author. Tel.: q1-323-4421153; Fax: q1-323-4423049.

types, including monocytermacrophages, endothelial cells,Žand T cells Hofman et al., 1993, 1994; La Frenie et al.,

.1997 . This HIV viral protein has been shown to inhibitŽ .antigen-induced T cell activity Viscidi et al., 1989 , de-

crease the proliferation of CD34 positive cells and increaseŽthe production of TGFb in bone marrow cells Breen et

.al., 1990 . Treatment with extracellular tat protects againstapoptosis and appears to upregulate Bcl-2, a protein corre-

Ž .lated with protection against cell death Zauli et al., 1993 .The tat protein has also been detected in the CNS in AIDSŽ .Hofman et al., 1994 , and shown to be a potent activatorof CNS cells. This viral protein functions to induce neu-ronal and astrocytic cell aggregation and migration in

Ž .cultures of fetal rat brain Kolson et al., 1993 , binds toneuroblastoma cells in vitro, and causes severe neurotoxicsymptoms when administered intra-cerebroventricularlyŽ .Sabatier et al., 1991 . Furthermore, tat has a direct effecton CNS-derived endothelial cells, causing an increase inthe expression of adhesion molecules, as well as the

Ž .upregulation of IL-6 production Hofman et al., 1994 .Thus tat appears to be a potent activator of a variety ofCNS cells.

There is extensive evidence that in AIDS dementiacomplex disease there is increased production of specificcytokines; TNF, IL-1 and IL-6 have been measured in the

0165-5728r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S0165-5728 98 00198-2

( )F.M. Hofman et al.rJournal of Neuroimmunology 94 1999 28–39 29

Ž .CSF Gallo et al., 1989; Laverda et al., 1994 and inŽautopsy material Tyor et al., 1992; Wesselingh et al.,

.1994; Sippy et al., 1995 . HIV infection of macrophagesand microglia has been shown to induce elevations in the

Žproduction of TNF, IL-1 and IL-6 Lepe-Zuniga et al.,.1987; Wright et al., 1988; Breen et al., 1990 . Along with

these proinflammatory cytokines, the chemokine IL-8 hasbeen reported to be upregulated in HIV positive individu-

Ž .als Matsumoto et al., 1993 and in HIV infected cellsŽ .Sopper et al., 1996; La Frenie et al., 1997 . This achemokine is synthesized and secreted in the 77 aminoacid form by endothelial cells, and is of the CXC structuralclass, containing one amino acid between two cysteinesŽ .Baggiolini et al., 1989 . IL-8 primarily mediates neu-

Žtrophil migration, adhesion and transmigration Karakurum.et al., 1994 , however, this chemokine has also been

reported to be chemotactic for a subpopulation of T cells,Ženhance monocyte binding to endothelial cells Op-

.penheim et al., 1991 , and influence physiological condi-tions such as body temperature and hormone productionŽ .Plata-Salaman and Borkoski, 1993 . The studies presentedhere demonstrate that the HIV-derived tat protein caninduce the production of IL-8 in endothelial cells of theCNS as well as in systemic endothelial cells; and thisproduction of IL-8 is regulated at the transcriptional level.Furthermore, tat regulation of IL-8 production is modu-lated by cytokines; the proinflammatory cytokine TNFfunctions additively with tat, whereas, TGFb downregu-lates tat-induced IL-8 production.

2. Materials and methods

2.1. Reagents

The following reagents were obtained commercially: tatŽ . Ž .Intracel, Cambridge, MA ; TNFa , interleukin-1b IL-1b

Ž .Boehringer-Manheim, San Diego, CA ; TGF b1Ž .Genzyme, Cambridge, MA ; rabbit and mouse mono-

Žclonal anti-human IL-8 antibody R&D, Minneapolis,. ŽMN ; irrelevant antibody rabbit anti-human prostate-

.specific antibody , biotinylated horse anti-mouse, avidin–Ž .biotin–peroxidase complex Dako, Glostrup, Denmark .

Rabbit anti-tat antibody was obtained from Intracel.

2.2. Cell culture

CNS-EC were derived from human brain as previouslyŽ .described in detail Hofman et al., 1994 . Baseline control

values for these cells vary, and may be due to differencesin tissue donors, i.e., age, gender and type of trauma.Tissues also vary in length of time before acquisition.

ŽCells were cultured in RPMI-1640 medium Gibco Labs,.Grand Islands, NY supplemented with 100 ngrml en-

Ždothelial cell growth factor Endogro VECTEC, Albany,

.NY , 2 mM L-glutamine, 10 mM Hepes, 24 mM sodiumbicarbonate, 300 u.s.p. units of heparin, 1%penicillinrstreptomycin, and 10% fetal calf serum. Thecells were used to passage 5 only; and Endogro-freemedium was used beginning 24 h before the initiation ofexperiments. The purity of CNS-EC cultures was con-firmed by immunocytochemical staining for factor VIII,glial fibrillary acidic protein, and the macrophage marker

Ž .CD1 l to be )98%. Microvessel endothelial cells MECderived from lung were commercially available through

Ž .Clonetics San Diego, CA . Retinal pigmented epithelialŽ .cells RPE were isolated and cultured as previously de-Ž .scribed Gabrielian et al., 1994 .

2.3. IL-8 assay

IL-8 production was evaluated using the commerciallyŽ .available ELISA kit R&D . Briefly, CNS-EC were grown

in culture to 80–90% confluence in 10% FCS; 24 h beforethe initiation of the experiment, the media was replaced by

Ž .media containing 2% FCS. Culture supernatant 100 ulwas removed after 72 h unless otherwise stated and evalu-ated for IL-8 content using the ELISA kit. The experimen-tal groups were prepared in duplicate, and the ELISAsamples were evaluated in triplicate. The data was ex-pressed in ng per 106 cells. The number of viable cellswere determined using trypan blue exclusion; viability wasroutinely greater than 95%.

2.4. Immunocytochemistry

Cells were treated as described above. At the termina-tion of the experiment, cell cultures were rinsed with PBS,

Ž 4prepared in suspension, and cytocentrifuged 5=10.cellsrslide . Air dried slides were fixed in acetone for 5

min, again allowed to dry, and then subjected to theŽ .staining procedure Hofman et al., 1994 . Briefly, cell

preparations were treated with the primary monoclonalŽ .mouse antibody 18 h , subsequently washed with PBS

Ž .twice 10 min , followed by an incubation with the biotin-labeled secondary antibody, horse anti-mouse immuno-

Ž .globulin Vector Labs; 30 min followed by treatment withŽ . Ž .the aminoethylcarbasole AEC solution 10 min , and

Ž .counterstained with Mayer’s hematoxylin 1 min . Irrele-vant isotype matched antibody was used in place of theprimary antibody as the negative control.

2.5. RNase protection analysis

The radioactively labeled RNA antisense probes wereprepared following the manufacturer’s protocol. Using the

Ž .In Vitro Transcription Kit Pharmingen, San Diego, CA ,Ž .10 ul of 32 P UTP 3000 Cirmmol, 10 mCirml NEN

.Research, Wilmington, DE , 1 ul GACU pool were added

( )F.M. Hofman et al.rJournal of Neuroimmunology 94 1999 28–3930

Ž . Žto the RNase Protection Assay RPA Template Set HCK-.5 which is a human chemokine multi-probe set including

IL-8 and the housekeeping gene, glyceraldehyde-3-phos-Ž . Ž .phate dehydrogenase GAPDH Pharmingen ; also in-

cluded were T7 polymerase, DTT, RNAsin and transcrip-tion buffer as suggested by the manufacturer. The reactionwas terminated by adding 2 ul of DNase for 30 min at378C. The probe was then extracted using tris-saturated

Ž . Žphenol: chloroform: isoamyl alcohol 25:24:1 GIBCO.BRL and chloroform: isoamyl alcohol sequentially, and

then ethanol precipitated. The radiolabeled RNA pellet wasair dried and solubilized with hybridization buffer. RNAfrom 2–3=106 cultured cells per experimental group wasisolated, and prepared according to a modification of the

Žacid phenol method using the Trizol reagent Life Tech-.nologies, Gaithersburg, MD as specified by the manufac-

Ž . 5turer. Total RNA 10 ug together with 6.2=10 cpm ofprobe was heat denatured at 908C and then hybridizedovernight at 568C. Subsequently, the samples were treatedwith the RNase cocktail, followed by proteinase K cock-tails, then precipitated using ammonium acetate andethanol. Air dried samples were solubilized in 1= loadingbuffer, denatured at 908C for 3 min and then placed on ice.The protected fragments were resolved in 5% acrylimider8

Ž .M urea gel 24 cm length ; the gel was dried and exposedŽto Hyper film Amersham Life Science, Arlington Heights,

.IL at y708C. The protected bands were observed for IL-8Ž . Ž .181 bp and GAPDH 96 bp . The manufacturer recom-mended yeast RNA negative control, and a positive controlwas included in every RPA.

2.6. Neutrophil migration assay

Ž .Polymorphonuclear cells PMN were isolated fromwhole blood of healthy human volunteers using a mixtureof sodium metrizoate and a Ficoll gradient centrifugationŽ1 step Polymorphs, Accurate Chemical and Scientific,

.Westbury, NY . Approximately 5 ml of whole blood waswithdrawn into a syringe containing 0.5 ml of 3.8% sodiumcitrate as an anticoagulant. The blood was then layeredover 3.5 ml of Polymorphprep in a 12 ml tube and spun at500=g for 35 min at 208C. After centrifugation 2 leuco-cyte bands were visible, and the lower band of PMNs wasresuspended in RPMI supplemented with 0.05% FCS and0.1% BSA. PMNs were then washed twice for 10 min at400=g. Cell purity, analyzed using morphologic criteria,following hematoxylin staining was 94%–98%; and viabil-ity was greater than 95% as determined by trypan blue dyeexclusion.

Migration of PMNs was conducted in cell culture in-Ž .serts Becton Dickinson, Franklin Lakes, NJ , using a 3

mm pore P.E.T. track-etched membrane, 24 well format.After 72 h, media from the differently treated CNS-ECgroups, cultured in RPMI with 0.05% FCSr0.1% BSA,were added to the lower compartment. Media which wasnot exposed to cells was used as the control. In the uppercompartment, 3=105 PMNs in culture media were added.The chambers were incubated for 30 min at 378C in 5%CO . At the conclusion of this time period, the chambers2

were collected, fixed and stained with the Diff-QuickŽfixative and staining solution Baxter Healthcare Corpora-

ŽFig. 1. Tat induces the production of IL-8 in CNS-EC. CNS-EC were cultured to confluency and treated with tat at a range of concentrations 10, 30, 100.ngrml ; after 72 h incubation, the culture supernatants were collected and tested for secreted IL-8 protein using the ELISA assay. The data are presented as

the mean"SEM of quadruplicated samples and expressed in ngr106 cells. These data are obtained from one of three representative experiments.


.tion, McGraw Park, IL . Filters were detached from theirinserts using a scalpel blade and placed on a microscopeslide. The cells attached on the upper side of the filter wereremoved using a cotton swab; a cover glass with mountingmedia was then added to the filter before observation.Cells were counted in 30 fields at =40 magnification. Theresults were obtained from one experiment which wasrepresentative of three different experiments and are ex-pressed as the increase in percent of cells migrating in the

experimental groups divided by the number of cells mi-grating when exposed to control media multiplied by 100.

2.7. Statistics

Values were presented as the mean"standard error ofŽ .the mean SEM , unless otherwise stated. Statistical signif-

icance was evaluated using the Student’s t-test for pairedcomparison; p-0.05 was considered significant.

Fig. 2. Immunocytochemistry of tat treated CNS-EC. Cell preparations of treated and untreated cultures were stained with monoclonal anti-human IL-8Ž . Ž . Ž . Ž . Ž .after 24 h A treated , B control . Dark precipitate indicates positive cells 400= magnification .


3. Results

3.1. Tat protein upregulates IL-8 production in CNS-EC

To determine whether CNS-EC respond to tat proteinby producing IL-8, endothelial cell cultures were grown toconfluency, then treated with tat. Concentrations of tatranging from 10 ngrml to 100 ngrml were tested, witheach group prepared in duplicate. Culture supernatantswere collected at 72 h unless otherwise stated, and evalu-ated for soluble IL-8 protein using the ELISA technique.Each ELISA sample was measured in triplicate; the datapresented was derived from one of three representativeexperiments, unless otherwise noted. The results demon-strated that tat induced an increase in the production andrelease of IL-8 in a concentration dependent manner, withtat increasing soluble IL-8 production 37% at 30 ngrml,

Ž .and 2.6 fold at 100 ngrml Fig. 1 . To determine whenIL-8 protein could first be observed, CNS-EC were treatedwith tat for 6, 24, 48 and 72 h; and then examined withmonoclonal anti-IL-8 using immunocytochemistry. After 6h incubation, there was no detectable difference betweencontrol and untreated cells. Cells treated for 24 h however

Ž .demonstrated 50–70% staining with IL-8 Fig. 2A . Inimmunostaining performed on tat treated cultures after 48and 72 h, the number of positive cells remained constant

Žwhile the intensity of staining appeared to increase data.not shown . Control untreated cultures exhibited -3%

Ž .positive Fig. 2B , suggesting very low levels of constitu-tive IL-8 production. To determine whether tat regulatesIL-8 at the level of mRNA, the RPA was performed,where the IL-8 and GAPDH genes were present in thesame template set. The kinetics of tat induced IL-8 mRNAexpression was examined by treating CNS-EC with tat for

Ž .1, 3 and 6 h. The results Fig. 3A showed that by 1 h,there was a 71% increase in IL-8 mRNA in tat treated

Žcells, with optimal expression occurring after 3 h 2.2 fold.increase ; by 6 h IL-8 mRNA levels were similar to

Ž .control values 17% increase . At 22 h post treatment withŽtat, IL-8 mRNA levels returned to control values data not

.shown . In order to determine whether this effect was tatspecific, tat treated cultures were incubated in the presenceor absence of polyclonal anti-tat antibody. The datademonstrated that anti-tat antibody completely blocked

Ž .IL-8 mRNA upregulation by tat Fig. 3B , while irrelevantŽ .antibody did not block tat activity data not shown .

To determine whether CNS-EC were unique in theirresponse to tat, microvessel endothelial cells from othersources were examined. Primary cultures of human lung

Ž .microvessel endothelial cells MEC were treated with tatfor 72 h; supernatants were collected and assayed for IL-8production. Fig. 4A shows that these capillary EC re-sponded to tat with a 4.8 fold increase at tat 100 ngrml. Incontrast to endothelial cells, primary cultures of retinal

Ž .pigmented epithelial RPE cells, a cell type that producedIL-8 in response to TNF or IL-1b, did not appear to

Ž .Fig. 3. Tat upregulates IL-8 mRNA expression. A . Confluent CNS-ECŽ .were incubated with tat 100 ngrml for 1, 3 or 6 h or left untreated for

Ž .the appropriate times. The RNase Protection Assays RPA were per-formed to analyze IL-8 mRNA. Total RNA was isolated from the cells,then probed with 32 P labeled riboprobes for IL-8 or GAPDH. The resultsshow the autoradiograph with the protected size corresponding to IL-8Ž . Ž .181 bp and GAPDH 96 bp . Percentages are calculated as the ratio ofspectrophotometric density measurement of the IL-8 band divided by the

Ž .corresponding GAPDH band. B . CNS-EC cultures as described aboveŽ .were treated with tat 100 ngrml for 3 h in the absence or presence of

polyclonal anti-tat antibody, anti-tat antibody alone, or left untreated. Theautoradiographic data is presented, and the ratio of IL-8rGAPDH iscalculated as described.

Ž .respond to tat under these conditions Fig. 4B . Theseresults demonstrated that tat stimulated IL-8 production inCNS as well as systemic microvessel endothelial cells;however specialized epithelial cells, RPE, did not appearto respond to tat.

3.2. Cytokines modulate tat-induced IL-8 production

A series of experiments were performed to determinewhether tat-induced production of IL-8 at the protein levelcan be modulated by pro- or anti-inflammatory cytokines.

Ž .The results Fig. 5 showed that with tat at 100 ngrmlŽthere was a 4.5 fold increase in IL-8 production, and 3

.fold increase at the suboptimal TNF concentration of 1pgrml; however tat and TNF together exhibit an additive

Ž .increase in IL-8 production 7 fold increase . This increasewas apparent whether the cells were pre-treated with TNF


Ž .Fig. 4. Tat stimulates microvessel endothelial cells to produce IL-8. A . Lung derived microvessel endothelial cells were cultured to confluency; andŽ .treated with tat 30, and 100 ngrml or left untreated for 72 h. Culture supernatants were collected and analyzed for secreted IL-8 protein. The data is

6 Ž .presented as IL-8 ngr10 cells and calculated as the mean"SEM of quadruplicate cultures. These data are representative of two similar experiments. B .Ž . Ž . Ž .RPE cells were cultured to confluency and exposed to tat 30 or 100 ngrml , TNF 10 pgrml , 1L-1b 10 pgrml , or left untreated. After 72 h, culture

supernatants were collected and tested for soluble IL-8 protein. The data are presented as described above.

for 30 min or exposed to both reagents simultaneously.CNS-EC were also examined with the anti-inflammatorygrowth factor TGFb. Cells exposed to TGFb did notdemonstrate significant differences in cell number, al-though TGFb has been reported to be an angiogenic as

Žwell an anti-angiogenic factor McCartney-Francis and

. Ž .Wahl, 1994 . Cells treated with TGFb 0.1 ngrml aloneŽ .Fig. 6 demonstrated no increase in IL-8; and often a 20to 30% decrease in IL-8 was detected. However, TGFb

appeared to significantly inhibit tat induced IL-8 produc-tion, from 4.5 fold increase above control in tat treatedcultures to 1.5 fold increase in CNS-EC treated with both


Ž . Ž .Fig. 5. TNF modulates tat-induced IL-8 production by CNS-EC. Cultures of CNS-EC were treated with tat 30 ngrml , TNF 1 pgrml , both reagents, orleft untreated for 72 h. The supernatants were subsequently analyzed for secreted IL-8 protein using the ELISA technique. The data are expressed as IL-8ngr106 cells and calculated as the mean"SEM of quadruplicate cultures. These data are representative of three similar experiments.

Ž .tat and TGFb Fig. 6 . This effect was observed whetherTGFb was added simultaneously with tat or pre-treated for30 min. This inhibitory effect of TGFb was observed at

Ždoses ranging from 0.1 ngrml to 10 ngrml data not

.shown . It should be noted that control, baseline levels ofIL-8 production vary; this may be a reflection of thedifferent tissues used in the preparation of primary cul-tures. These data demonstrate that tat induced IL-8 produc-

Ž . Ž .Fig. 6. TGF b inhibits tat stimulated IL-8 production. Culture supernatants from CNS-EC treated with tat 100 ngrml , TGF b 0.1 ngrml , or bothreagents were analyzed for secreted IL-8 using the ELISA technique. Both reagents were added to the cultures simultaneously, or TGFb was added 30 minbefore the addition of tat. The data are expressed as IL-8 ngr106 cells and calculated as the mean"SEM of triplicate cultures. These data arerepresentative of three similar experiments.


tion can be regulated by cytokines, with TNF enhancingand TGFb downregulating tat activity.

3.3. Tat stimulates production of a functional chemo-attractant

In order to determine whether tat induced the produc-tion of functional IL-8, endothelial cells were treated withtat as previously described. The supernatants from 72 hcultures were added to the lower compartment as describedin Section 2. Supernatants, with this long term accumula-tion of IL-8, provided the maximal effect, because IL-8 isrelatively stable in vitro, as demonstrated by studies usingrecombinant IL-8 incubated for 24, 48 and 72 h with

Ž .minimal loss of chemotactic properties data not shown .The results in Fig. 7 are expressed as the percent increasein the number of cells migrating in the presence of culturesupernatants from the different experimental groups com-

pared to the number of cells present when media alone wasadded. These data show the conditioned media from tat-treated cells increased PMN migration 320% compared tountreated cell cultures. An ELISA analysis of IL-8 in the

Ž 6supernatant from tat-stimulated cultures 21.28 ngr10. Ž 6cells was 65% higher than control cultures 13.74 ngr10.cells . This significant but low increase in IL-8 production

Žwith tat-treatment was due to the reduced serum 0.05%.FCSr0.1% BSA used in the culture media necessary for

preparing supernatants for the migration assay. To deter-mine whether IL-8 was responsible for this migration,

Ž .polyclonal rabbit anti-IL-8 antibody 200 mgrml wasadded to the tat-induced supernatant for 1 h before themigration assay was initiated. The results demonstratedthat anti-1L-8 antibody completely blocked PMN migra-tion by tat-treated culture supernatant; irrelevant antibodyhad no effect on PMN migration. The data in Fig. 7 arerepresentative of three experiments.

Ž . Ž .Fig. 7. Tat stimulates the production of functional IL-8. Supernatants from untreated cultures or cultures treated with tat 100 ngrml , or TNF 5 ngrmlwere collected after 72 h in vitro. Tat or TNF supernatants were either left untreated or incubated with either anti-IL-8 or the irrelevant antibody for 1 hbefore use. The different supernatants were then added to the lower compartment of the chemotaxis chambers; 2=105 PMNs were placed in the upperchamber. After 30 min, the filters were collected, stained and counted as described. The percent increase is calculated as the ratio of the number of PMNson the filter exposed to the experimental supernatants divided by the number of PMNs in the chamber exposed to culture media alone. The data arerepresentative of three similar experiments.


4. Discussion

This study demonstrated that microvessel endothelialcells, when treated exogenously with the HIV tat proteininduced the upregulation of IL-8 both at the level ofmRNA and protein. CNS endothelial cells are a major

Ž .component of the blood-brain-barrier BBB , and haveunique structural and functional properties. These special-ized endothelial cells have tight junctions, a paucity ofintracellular vesicles, and an array of transport moleculesregulating the movement of water-soluble compounds from

Žthe blood to the brain Laterra and Goldstein, 1993; Kale-.ria et al., 1988; Neuwelt et al., 1991 . Several groups have

demonstrated the presence of serum proteins in the brainparenchyma of AIDS patients suggesting an alteration in

Žthe function of the BBB in HIV infection Petito and Cash,.1992; Power et al., 1993 . There are two possible mecha-

nisms of this CNS-EC activation; either these endothelialcells are HIV-infected, or products of HIV infected circu-lating cells stimulate the endothelial cells. HIV infection ofhuman and simian brain endothelial cells have been re-

Ž .ported Moses et al., 1993; Mankowski et al., 1994 , butthe extent of HIV infection of endothelial cells in vivo is

Ž .still controversial Kure et al., 1990; Smith et al., 1990 .There is however, an extensive literature on the activationof endothelial cells by the inflammatory cytokines and

Žviral proteins released by HIV infected macrophages Pober.and Cotran, 1990; Hofman et al., 1993, 1994 . Therefore,

we proceeded to examine the effect of these factors onCNS-EC function.

In comparing CNS-EC with systemic microvessel en-dothelial cells, studies showed that endothelial cells de-rived from lung microvessels, responded to tat with IL-8production at similar doses, suggesting that the bindingsites and activation processes for tat may be similar forthese endothelial cell types. In contrast epithelial cells,RPE, a critical component of the blood retina barrier,showed a lack of response to tat, although TNF and IL-1 b

stimulated cultures induced significant amounts of IL-8Ž .Elner et al., 1997 . Thus IL-8 induction by tat may bemore universal for microvessel endothelial cells but stillcell type dependent.

Experiments presented here showed that tat acts as anextracellular agent whose functions can be modulated bycytokines, at the site of secretion. TNF, produced by HIVinfected leukocytes as well as activated astrocytes andmicroglia, was shown here to act additively with tat in theproduction of IL-8. This effect was observed most dramati-

Ž .cally at low concentrations of TNF 1 pgrml . This sug-gests that at low levels of both TNF and tat, these proteinscan function together in amplifying the local production ofIL-8. TNF has been shown to act synergistically with otherfactors to upregulate IL-8 production, particularly with

Ž .Interferon IFN-g ; while IFN-g alone does not increaseIL-8, TNF and INF-g together significantly enhance

Ž .chemokine production Yasumoto et al., 1992 . The regula-

tion by both TNF and IFN-g appears to be mediatedthrough the binding of NF-kB and AP-1 like proteins tothe regulatory region of IL-8 gene. The mechanism ofTNF-tat interaction in IL-8 production in endothelial cellsremains to be explored.

The data presented here show that TGF-b downregu-lated the production of IL-8. TGF-b functions as both apro- and anti-inflammatory growth factor. TGF-b has beenshown to suppress cell proliferation, induce apoptosis,regulate the expression of growth factor receptors, de-crease endothelial cell migration, and downregulate adhe-

Žsion molecule expression Mooradian and Digluo, 1990;McCartney-Francis and Wahl, 1994; Fabry et al., 1995;

.Chen et al., 1997; Lee et al., 1997 . TGF-b also functionsin enhancing angiogenesis and extracellular matrix protein

Ž .production Penttinen et al., 1988 . This growth factor isnormally present in the brain, and synthesized by endothe-

Žlial cells, astrocytes and microglia Constam et al., 1992;.Weller and Fontana, 1995 . The in vivo relevance of

TGF-b has been documented with its upregulation in HIVŽ .infection Wahl et al., 1991; Zauli et al., 1992 , supporting

the function of TGF-b as immune suppressive in the CNS.Although TGF-b can directly stimulate neutrophil and

Žleukocyte chemotaxis Drake and Issekutz, 1993; and Wahl.et al., 1987 , this and other studies have reported that

TGF-b acting directly on basal level or TNF-activatedendothelial cells caused an inhibition of IL-8 productionŽ .Chen and Manning, 1996; Smith et al., 1996 , thereby

Žresulting in an inhibition of neutrophils migration Smith.et al., 1996 . In other cell types, e.g., RPE, lung alveolar

epithelial and endometrial stromal cells, TGF-b appears toŽenhance IL-8 mRNA and protein production Kuppner et.al., 1995; Arici et al., 1996; Kumar et al., 1996 . Thus the

regulatory role of TGF-b may reflect the local microenvi-ronment, including the cell phenotype, the extracellularmatrix, and the composition and concentration of the localcytokines and growth factors present at that specific organsite. Therefore in the CNS, particularly associated with theblood vessels of the BBB, TGF-b may have anti-in-flammatory effects, which function as homoeostatic regula-tor to regulate inflammation.

Several studies have shown that there is an increase inIL-8 production in HIV-infected patients. Elevated serumlevels of IL-8 were demonstrated in HIV-infected individu-

Ž .als Matsumoto et al., 1993 . Furthermore, patients withŽHIV infected monocytes express IL-8 mRNA Tsai et al.,

.1991 , and secrete IL-8 protein, as well as other proinflam-Žmatory cytokines such as TNF, IL-1b and IL-6 Esser et.al., 1991; Esser et al., 1996; Goletti et al., 1996 . Simian

microglia as well as macrophages when infected with SIVŽ .produce IL-8 Sopper et al., 1996 . Bronchial lavage fluid

and tissue from HIV-infected individuals demonstrated asignificant increase of IL-8 production over control prepa-

Ž .rations Denis and Ghardirian, 1994 . These reports indi-cate that the upregulation of IL-8 is one manifestation ofpathology in AIDS, however the function of IL-8 in this


disease is not clear. IL-8 is known to be a potent chemo-Ž .attractant and activator of PMNs Baggiolini et al., 1989 ,

but significant PMN infiltration into tissue in AIDS is nota hallmark of this disease. Studies have shown that indeed,PMN function in later stages of HIV infection is impaired,with accelerated apoptosis of these cells further contribut-ing to an already existing debilitated immune responseŽ .Pitrak et al., 1996 . However, the correlation between theabnormal function of the PMNs and the overproduction ofIL-8 is speculative at this time. In addition to its activityon PMNs, IL-8 is also a potent chemoattractant for CD4

Ž .and CD8 T cells Jinquan et al., 1993 . Cytotoxic TŽ .lymphocytes CTL , CD8 positive cells, have been shown

to be involved in AIDS dementia. CTLs have been foundŽin the CSF of HIV infected individuals Jassoy et al.,

.1992 ; and virus specific CTL have been found in the CSFŽand brains during early stages of simian AIDS von Her-

.rath et al., 1995 . Furthermore, both T cell populationsŽ .constitutively express the IL-8 CXCR2 receptor B; andŽ .the CD4 cells also secrete IL-8 Jinquan et al., 1995 . This

CXCR2 receptor, however, does not play a significant roleŽ .as a co-receptor for HIV-1 infection He et al., 1997 .

These data suggest that IL-8 may function as a chemokinefor T cells and thus play a role in the recruitment andtransmigration of these cells into tissue. The role of Tcells, particularly CD8 cells in HIV infected tissue, may beprotective in the initial stage of cell mediated viral clear-ance, and in the suppression of HIV replication through the

Žproduction of viral suppressing cytokines Cocchi et al.,.1995 , and later pathogenic to the virally infected-tissue

Ž .Ando et al., 1993 . Recent evidence has shown that IL-8mediates increased CMV replication, suggesting that thisprotein may play a role in enhancing this opportunistic

Žviral infection common in HIV encephalitis Capobianchi.et al., 1997 . In addition, HHV-8, the human herpes virus

associated with Kaposi’s Sarcoma, shown to express func-tional IL-8 receptors, is activated in response to tat-proteinŽ .Fiorelli et al., 1998 . These new data suggest that IL-8may functions other than as a chemoattractant, and may beinvolved in viral binding or replication.

In conclusion, the studies presented here demonstratethat exogenous treatment with the HIV-derived tat proteininduced CNS-derived endothelial cells to produce andsecrete functional IL-8. This tat-induced IL-8 synthesis canbe upregulated by low levels of the proinflammatory cy-tokine TNF, and downregulated by the pleiotropic growthfactor TGF-b. These data suggest that the local productionof tat, especially in the presence of TNF may be responsi-ble for the increased production of IL-8 reported in AIDS.

Acknowledgements

This work was supported by Grants EYO 8144 and NSŽ .33805 FMH from the National Institutes of Health.

References

Ando, K., Moriyama, T., Guidotti, L.G., Wirth, S., Schreiber, R.D.,Schlicht, H.J., Huang, S., Chisari, F.V., 1993. Mechanisms of class Irestricted immunopathology. A transgenic mouse model of fulminanthepatitis. J. Exp. Med. 178, 1541–1554.

Arici, A., MacDonald, P.C., Casey, M.L., 1996. Modulation of the levelsof interleukin-8 messenger ribonucleic acid and interleukin-8 proteinsynthesis in human endometrial stromal cells by transforming growth

Ž .factor-b . J. Clin. Endo. Metabol. 81 8 , 3004–3009.1

Baggiolini, M., Walz, A., Kunkel, S.H., 1989. Neutrophil-activatingpeptide-1rinterleukin 8, a novel cytokine that activates neutrophils. J.Clin. Invest. 84, 1045–1049.

Breen, E.C., Razai, A.R., Nakajima, K., Beall, G.N., Mitsuyasu, R.T.,Hirano, T., Kishimoto, T., Marinez-Maza, O., 1990. Infection withHIV is associated with elevated IL-6 levels and production. J. Im-munol. 144, 480–484.

Capobianchi, M.R., Barresi, C., Borghi, P., Gessani, S., Fantuzzi, L.,Ameglio, F., Belardelli, F., Papdia, S., Dianizani, F., 1997. Humanimmunodeficiency virus type1 gp 120 stimulates cytomegalovirusreplication in monocytes: Possible role of endogenous interleukin-8. J.

Ž .Virol. 71 2 , 1591–1597.Chen, T.C., Hinton, D., Yong, V.W., Hofman, F.M., 1997. TGF-b and2

soluble p55 TNF R modulate VCAM-1 expression in glioma cells andbrain derived endothelial cells. Neuroimmunology 73, 155–161.

Chen, C.C., Manning, A.M., 1996. TGFb , IL-10 and IL-4 differentially1

modulate the cytokine-induced expression of IL-6 and IL-8 in humanŽ .endothelial cells. Cytokine 8 1 , 58–65.

Cocchi, F., DeVico, A.L., Garzino, D.A., Arya, S.K., Gallo, R.C., Lusso,P., 1995. Identification of Rantes, M1P-1 alpha, and M1P-1 beta asthe major HIV-suppressive factors produced by CD8q T cells. Sci-

Ž .ence 270 5243 , 1811–1815.Constam, D.B., Philipp, J., Malipiero, U.V., ten Dijke, P., Schachner, M.,

Fontana, A., 1992. Differential expression of transforming growthfactor-b, b and b by glioblastoma cells, astrocytes and microglia. J.2 3

Ž .Immunol. 148 15 , 1404–1410.Denis, M., Ghardirian, E., 1994. Dysregulation of interleukin 8, inter-

leukin 10, and interleukin 12 release by alveolar macrophages fromHIV type 1-infected subjects. AIDS Res. Hum. Retroviruses 10,1619–1627.

Drake, W.T., Issekutz, A.C., 1993. Transforming growth factor b en-1

hances polymorphonuclear leukocyte accumulation in dermal inflam-mation and transendothelial migration by a priming action. Immunol-

Ž .ogy 78 2 , 197–204.Elner, S.G., Elner, V.M., Bian, Z.M., Lukacs, N.W., Kurtz, R.M.,

Stuefer, R.M., Kunkel, S.L., 1997. Human retinal pigment epithelialcell interleukin-8 and monocyte chemotactic protein-1 modulation by

Ž .T lymphocyte products. IOVS 38 2 , 446–455.Ensoli, B., Barillari, G., Salahuddin, C.Z., Gallo, R.C., Wong-Staal, F.,

1990. Tat protein of HIV-1 stimulates growth of cells derived fromKaposi’s sarcoma lesions of AIDS patients. Nature 345, 84–86.

Ensoli, B., Buonagoro, L., Barillari, G., Fiorelli, V., Gendelman, R.,Morgan, R.A., Wingfield, P., Gallo, R.C., 1993. Release, uptake, andeffects of extracellular human immunodeficiency virus type 1 tatprotein on cell growth and viral transactivation. J. Virology. 67,277–287.

Esser, R., Glienke, W., von Briesen, M., Rubsamen-Waigmann, H.,¨Anderson, R., 1996. Differential regulation of proinflammatory andhematopoeitc cytokines in human macrophages after infection with

Ž .human immunodeficiency virus. Blood 88 9 , 3474–3481.Esser, R., von Briesen, H., Brugger, M., Ceska, M., Glienke, W., Muller,

S., Rehm, A., Rubsamen-Waigmann, H., 1991. Secretory repertoire of¨HIV-infected human monocytesrmacrophages. Pathology 59, 219–222.

Fabry, Z., Topham, D.J., Fee, D., Herlein, J., Carlino, J.A., Hart, M.N.,


Sriram, S., 1995. TGF-b decrease migration of lymphocytes in vitro2

and homing of cells into the central nervous system in vivo. J.Immunol. 155, 325–332.

Fauci, A.S., 1988. The human immunodeficiency virus type 1: infectivityand mechanism of pathogenesis. Science 239, 617–622.

Fiorelli, V., Gendelman, R., Sirianni, M.C., Chang, H.-K., Columbini, S.,Markham, P.D., Monini, P., Sonnabend, J., Pintus, A., Gallo, R.C.,Ensoli, B., 1998. g-Interferon produced by CD8q T cells infiltratingKaposi’s sarcoma induces spindle cells with angiogenic phenotypeand synergy with human immunodeficiency virus-1 tat protein: animmune response to human herpesvirus-8 infection?. Blood 91, 956–967.

Frankel, A.D., Pabo, C.O., 1988. Cellular uptake of the tat protein fromhuman immunodeficiency virus. Cell 55, 1189–1193.

Gabrielian, K., Osusky, R., Sippy, B.D., Ryan, S.J., Hinton, D.R., 1994.Effect of TGFB on interferon-a-induced HLA-DR expression inhuman retinal pigment epithelial cells. Invest. Ophthalmol. Vis. Sci.35, 4253–4259.

Gallo, P., Frei, K., Rordorf, C., Lazdins, J., Tavolato, B., Fontana, A.,Ž .1989. Human immunodeficiency virus type 1 HIV-1 infection of the

central nervous system: An evaluation of cytokines in cerebrospinalfluid. J. Neuroimmunol. 23, 109–116.

Goletti, D., Kinter, A.L., Hardy, E.C., Poli, G., Fauci, A.S., 1996.Modulation of endogenous IL-1b and IL-1 receptor antagonist resultsin opposing effects on HIV expression in chronically infected mono-

Ž .cytic cells. J. Immunology 156 9 , 3501–3508.He, J., Chen, Y., Farzan, M., Choe, M., Ohagen, A., Gartnor, S.,

Busciglio, J., Yang, X., Hofmann, W., Newman, W., MacKay, C.R.,Sodroski, J., Gabuzuda, D., 1997. CCR3 and CCR5 are co-receptorsfor HIV-1 infection of microglia. Nature 385, 845–849.

Hofman, F.M., Wright, A.D., Dohadwala, M.M., Wong-Staal, F., Walker,S.W., 1993. Exogenous tat protein activates human endothelial cells.Blood 82, 2774–2780.

Hofman, F.M., Dohadwala, M.M., Wright, A.D., Hinton, D.R., Walker,S.M., 1994. Exogenous tat protein activates central nervous system-derived endothelial cells. J. Neuroimmunol. 54, 19–28.

Jassoy, C., Johnson, R., Navia, B., Worth, J., Walker, B., 1992. Detectionof a vigorous HIV-1-specific cytotoxic T lymphocyte response incerebrospinal fluid from infected persons with AIDS dementia com-plex. J. Immunol. 149, 3113–3119.

Jinquan, T., Deleuran, B., Gesser, B., Maare, H., Deleuran, M., Larsen,C.G., Thestrup-Pederson, K., 1995. Regulation of human T lympho-cyte chemotaxis in vitro by T-cell derived cytokines IL-2, IFN-g,IL-4, IL-10, IL-13. J. Immunol 154, 3742.

Jinquan, T., Larsen, C.G., Gesser, B., Matsushima, K., Thestrup-Peder-sen, K., 1993. Human IL-10 is a chemoattractant for CD8q Tlymphocytes and an inhibitor of IL-8-induced CD4q T lymphocyte

Ž .migration. J. Immunol. 151 9 , 4545–4551.Kaleria, R.N., Gravina, S.A., Schmidley, J.W., Perry, G., Harik, S.I.,

1988. The glucose transporter of the human brain and blood-brainbarrier. Ann. Neurol. 24, 757–764.

Karakurum, M., Shreeniwas, R., Chen, J., Pinsky, D., Yan, S.D., Ander-son, M., Sunouchi, K., Major, J., Hamilton, T., Kuwabara, K., Rot,A., Nowyrod, R., Stern, D., 1994. Hypoxic induction of interleukin-8gene expression in human endothelial cells. J. Clin. Invest. 93,1564–1570.

Kobayshi, S., Yamamoto, Y., Kobayashi, N., Yamamoto, N., 1990.Serum level of TWFa in HIV-infected individuals. AIDS 4, 169.

Kolson, D.L., Buchhalter, J., Collman, R., Hellmig, B., Farrell, C.F.,Debouck, C., Gonzalez-Scarano, F., 1993. HIV-1 tat alters normalorganizations of neurons and astrocytes in primary rodent brain cellcultures: RGD sequence dependence. AIDS Research Human Retro-

Ž .viruses 9 7 , 677–685.Kumar, N.M., Rabadi, N.H., Sigiordson, L.S., Schunemann, H.J., Live-

buga-Mukasa, J.S., 1996. Induction of interleukin-1 and interleukin-8mRNAs and problems by TGFb in rat lung alveolar epithelial cells.1

Ž .J. Cell Physiol. 169 1 , 186–199.

Kuppner, M.C., McKillop-Smith, S., Forrester, J.V., 1995. TGFb andIL-1b act in synergy to enhance IL-6 and IL-8 mRNA levels and IL-6production by human retinal pigment epithelial cells. Immunology 84Ž .2 , 265–271.

Kure, K., Lyman, W.D., Weidenheim, K.M., Dickson, D.W., 1990.Cellular localization of an HIV-1 antigen in subacute AIDS encephali-tis using an improved double-labeling immunohistochemical method.Am. J. Pathol. 136, 1085–1092.

La Frenie, R.M., Wahl, L.M., Epstein, J.S., Yamada, K.M., Dhawan, S.,1997. Activation of monocytes by HIV-tat treatment is mediated bycytokine expression. J. Immunol. 159, 4077–4083.

Laterra, J., Goldstein, G.W., 1993. Brain microvessels and macrovascularŽ .cells in vitro. In Pardridge, W.M. Ed. , The Blood-Brain Barrier:

Cellular and Molecular Biology, Raven Press, New York, 1–24.Laverda, A., Gallo, P., DeRossi, A., Sivieri, S., Cogo, P., Pagliaro, A.,

Chieco-Bianchi, L., Tavolato, B., 1994. Cerebrospinal fluid analysisin HIV-1 infected children: immunological and virological findingsbefore and after AZT therapy. Acta Pediatr. 83, 1038–1042.

Lee, Y.J., Lu, H.T., Nguyen, V., Qin, H., Howe, P.H., Hocevar, B.A.,Boss, J.M., Ransohoff, R.M., Benbeniste, E.N., 1997. TGF-b sup-2

pressing IFN-g induction of class II MHC gene expression by inhibit-ing class II transactivation messenger RNA expression. J. Immunol.158, 2065–2075.

Lepe-Zuniga, J.L., Mansell, P.W.A., Hersh, E.M., 1987. Idiopathic pro-duction of interleukin-1 in acquired immunodeficiency syndrome. J.Clin. Microbiol. 25, 1695.

Mankowski, J.L., Spelman, J.P., Ressetar, H.G., Strandberg, J.D., Lat-erra, J., Carter, D.L., Clements, J.E., Zink, M.C., 1994. Neurovirulentsimian immunodeficiency virus replicates productively in endothelialcells of the central nervous system in vivo and in vitro. J. Virol. 68,8202–8208.

Matsumoto, T., Mike, T., Nelson, R.P., Trudeau, W.L., Lockey, R.F.,Yodoi, J., 1993. Elevated serum levels of IL-8 in patients with HIVinfection. Clin. Exp. Immunol. 93, 149–151.

McCartney-Francis, N.L., Wahl, S.M., 1994. Transforming growth factorb: A matter of life and death. J. Leukocyte Biol. 55, 401–409.

Mooradian, D.L., Digluo, C.A., 1990. Effects of epidermal growth factorand transforming growth factor b and rat heart endothelial cell1

anchorage-dependent and independent growth. Exp. Cell Res. 186,122–129.

Moses, A.V., Bloom, F.E., Pauza, C.D., Nelson, J.A., 1993. Humanimmunodeficiency virus infection of human brain capillary endothe-lial cells occurs via a CD4rgalactosylceramide-independent mecha-nism. Proc. Natl. Acad. Sci. USA 90, 10474–10478.

Neuwelt, E.A., Pagel, M.A., Dix, R.D., 1991. Delivery of ultraviolet-in-activated 35S-herpesvirus across an osmotically modified blood-brainbarrier. J. Neurosurg. 74, 475–479.

Oppenheim, J.J., Zachariae, C.O.C., Mukaida, N., Matsushima, K., 1991.Properties of the novel proinflammatory supergene intercrine cytokinefamily. Ann. Rev. Immunol. 9, 617–648.

Penttinen, R.P., Kobayashi, S., Bornstein, P., 1988. Transforming growthfactor beta increases mRNA for matrix proteins both in the presence

Ž .and in the absence of changes in mRNA stability. PNAS 85 4 ,1105–1108.

Petito, C.K., Cash, K.S., 1992. Blood-brain barrier abnormalities in theacquired immunodeficiency syndrome immunohistochemical localiza-tion of serum proteins in postmortem brain. Ann. Neurol. 32, 658–666.

Pitrak, D.L., Tsai, H.C., Mullane, K.M., Sutton, S.H., Stevens, P., 1996.Accelerated neutrophil apoptosis in the acquired immunodeficiency

Ž .syndrome. J. Clin. Invest. 98 12 , 2714–2719.Plata-Salaman, C., Borkoski, J., 1993. Interleukin-8 modulates feeding by

direct action in the central nervous system. Am. J. Physiol. 265,R877–882.

Pober, J.S., Cotran, R.S., 1990. The role of endothelial cells in inflamma-Ž .tion. Transpl. 50 4 , 537–544.

Power, C., Kong, P.A., Crawford, T.O., Wesselingh, S., Glass, J.D.,McArthur, J.C., Trapp, B.D., 1993. Cerebral white matter changes in


acquired immunodeficiency syndrome dementia: alterations of theblood-brain barrier. Ann. Neurol. 34, 339–350.

Sabatier, J.M., Vives, E., Mabrouk, K., Benjouad, A., Rochat, H., Duval,A., Hue, B., Bahraoui, E., 1991. Evidence for neurotoxic activity oftat from human immunodeficiency virus type 1. J. Virology. 65,961–967.

Sippy, B.D., Hofman, F.M., Wallach, D., Hinton, D.R., 1995. Increasedexpression of TNF-a receptors in the brains of patients with acquiredimmune deficiency syndrome. J. Acquired Immune Deficiency Syn-drome Hum. Retrovirology 10, 511–524.

Smith, W.B., Noack, L., Khew-Goodall, Y., Isenmann, S., Vadas, M.A.,Gamble, J.R., 1996. Transforming growth factor-b inhibits the pro-1

duction of IL-8 and the transmigration of neutrophils through acti-vated endothelium. J. Immunol. 157, 360–368.

Smith, T.W., DeGuolami, U., Henin, D., Bolgert, F., Hauw, J.J., 1990.Ž .Human immunodeficiency virus HIV leukoencephalopathy and the

microcirculation. J. Neuropathol. Exp. Neurol. 49, 357–370.Sopper, S., Demuth, M., Stahl-Hennig, C., Hunsmann, G., Plesker, R.,

Coulibaly, C., Czub, S., Ceska, M., Koutsilieri, E., Reiderer, P.,Brinkmann, R., Katz, M., Ter Meulen, V., 1996. The effect of simianimmunodeficiency virus infection in vitro and in vivo on the cytokineproduction of isolated microglia and peripheral macrophages fromrhesus monkey. Virology 220, 320–329.

Tsai, W., Hirose, K., Nara, P., Kuang, Y., Conley, S., Li, B., Kung, H.,Matsushima, K., 1991. Decrease in cytokine production by HIV-in-fected macrophages in response to LPS-mediated activation. Lym-phokine Cytokine Res. 10, 421–429.

Tyor, W.R., Glass, J.D., Griffin, J.W., Becker, S., McArthur, J.C.,Bezman, L., Griffin, D.E., 1992. Cytokine expression in the brainduring the acquired immunodeficiency syndrome. Ann. Neurol. 31,349–360.

Viscidi, R.P., Mayur, K., Lederman, H.M., Frankel, A.D., 1989. Inhibi-tion of antigen-induced lymphocyte proliferation by tat protein fromHIV-1. Science 246, 1606–1608.

von Herrath, M., Oldstone, M.B.A., Fox, M.S., 1995. Simian immunode-

Ž .ficiency virus SIV -specific CTL in cerebrospinal fluid and brains ofSIV infected rhesus macaques. J. Immunol. 154, 5582–5589.

Wahl, S.M., Hunbt, D.A., Wakefield, L.M., McCartney-Francis, N.,Wahl, L.M., Roberts, A.B., Sporn, M.B., 1987. Transforming growthfactor type beta induces monocyte chemotaxis and growth production.

Ž .Proc. Natl. Acad. Sci. USA 84 16 , 5788–5792.Wahl, S.M., Allen, J.B., Francis, N.M., 1991. Macrophage and

astrocyte-derived transforming growth factor b as a mediator ofcentral nervous system dysfunction in acquired immune deficiencysyndrome. J. Exp. Med. 173, 981–991.

Weller, M., Fontana, A., 1995. The failure of current immunotherapy formalignant glioma. Tumor-derived TGF b T-cell apoptosis, and the1

w x Ž .immune privilege of the brain. Brain Res. Review 21 2 , 128–151.Wesselingh, S., Glass, J., McArthur, J., Griffin, J., Griffin, D., 1994.

Cytokine dysregulation in HIV-associated neurological disease. Adv.Neuroimmunol. 4, 199–206.

Wright, S., Jewett, A., Mitsuyasu, R., Bonavida, B., 1988. Spontaneouscytotoxicity and tumor necrosis factor production by peripheral bloodmonocytes from AIDS patients. J. Immunol. 141, 99–104.

Yasumoto, K., Okamoto, N., Mukaida, S., Murakami, M.M., Mat-sushima, K., 1992. Tumor necrosis factor a and interferon-g syner-gistically induce interleukin-8 production in a human gastric cancercell line through acting concurrently on AP-1 and NF-KB-like bind-ing sites of the interleukin-8 gene. J. Biol. Chem. 26731, 22506–22511.

Zauli, G., Gibellini, D., Milan, D., Mazzoni, M., Borgatti, P., La Placa,M., Capitani, S., 1993. Human immunodeficiency virus type-1 tatprotein protects lymphoid, epithelial, and neuronal cell lines fromdeath by apoptosis. Cancer Res. 53, 4481–4485.

Zauli, G., Davis, B.R., Carla, M., Visani, G., Gurlini, G., La Placa, M.,1992. Tat protein stimulates production of transforming growth fac-tor-b1 by marrow macrophages: a potential mechanism for humanimmunodeficiency virus-1-induced hematopoietic suppression. Blood

Ž .80 12 , 3036–3043.

HIV1 tat protein induces the production of interleukin-8 by human brain-derived endothelial cells

Documents