Intracellular Human Immunodeficiency Virus Tat Expression in Astrocytes Promotes Astrocyte Survival but Induces Potent Neurotoxicity at Distant Sites via Axonal Transport

The human immunodeficiency virus (HIV)-Tat protein has been implicated in the neuropathogenesis of HIV infection. However, its role in modulating astroglial-neuronal relationships is poorly understood. Astrocyte infection with HIV has been associated with rapid progression of dementia. We thus initially transfected astrocytes with HIV proviral DNA and confirmed Tat production in these cells. Subsequently, using stably Tat-producing asytocyte cell lines, we observed that Tat promoted astrocyte survival by causing a prominent antioxidant effect and resistance to cell injury in these cells. Tat was released extracellularly where it could be taken up by other cells. Tat remained functionally active following uptake and caused long terminal repeat (LTR) transactivation in lymphocytic and astrocytic cell lines. Tat released from astrocytes caused mitochondrial dysfunction, trimming of neurites, and cell death in neurons. Tat neurotoxicity was attenuated by anti-Tat antibodies, kynurenate or heparan sulfate. The neurotoxic effects of Tat were caused at concentrations lower than that needed to cause LTR transactivation. When Tat-expressing cells were injected into the rat dentate gyrus, Tat was taken up by granule cells and transported along neuronal pathways to the CA3 region where it caused glial cell activation and neurotoxicity. The arginine-rich domain of Tat was essential for both the LTR transactivation and the neurotoxic properties of Tat. Thus HIV-Tat is a potent neurotoxin that may act at distant sites while at the same time it assures its production by preventing cell death in astrocytes where it is produced. The human immunodeficiency virus (HIV)-Tat protein has been implicated in the neuropathogenesis of HIV infection. However, its role in modulating astroglial-neuronal relationships is poorly understood. Astrocyte infection with HIV has been associated with rapid progression of dementia. We thus initially transfected astrocytes with HIV proviral DNA and confirmed Tat production in these cells. Subsequently, using stably Tat-producing asytocyte cell lines, we observed that Tat promoted astrocyte survival by causing a prominent antioxidant effect and resistance to cell injury in these cells. Tat was released extracellularly where it could be taken up by other cells. Tat remained functionally active following uptake and caused long terminal repeat (LTR) transactivation in lymphocytic and astrocytic cell lines. Tat released from astrocytes caused mitochondrial dysfunction, trimming of neurites, and cell death in neurons. Tat neurotoxicity was attenuated by anti-Tat antibodies, kynurenate or heparan sulfate. The neurotoxic effects of Tat were caused at concentrations lower than that needed to cause LTR transactivation. When Tat-expressing cells were injected into the rat dentate gyrus, Tat was taken up by granule cells and transported along neuronal pathways to the CA3 region where it caused glial cell activation and neurotoxicity. The arginine-rich domain of Tat was essential for both the LTR transactivation and the neurotoxic properties of Tat. Thus HIV-Tat is a potent neurotoxin that may act at distant sites while at the same time it assures its production by preventing cell death in astrocytes where it is produced. human immunodeficiency virus long terminal repeat 4-hydroxynonenal green fluorescent protein tetracycline reverse tetracycline transactivator enzyme-linked immunosorbent assay reverse transcription glyceraldehyde-3-phosphate dehydrogenase phosphate-buffered saline 3-nitroproprionic acid glial fibrillary acid protein chloramphenicol acetyl transferase induced nitric oxide synthetase microtubule associated protein-2 The brain is a frequent target in patients with human immunodeficiency virus (HIV)1infection resulting in a dementing illness termed HIV dementia (HIVD). The pathological features associated with HIVD include microglial cell activation, astrocytosis, decreased synaptic and dendritic density, and selective neuronal loss (1Masliah E. Ge N. Morey M. DeTeresa R. Terry R.D. Wiley C.A. Lab. Invest. 1992; 66: 285-291PubMed Google Scholar). Neuronal damage occurs through toxic substances released from infected microglia or macrophages and possibly astrocytes (2Nath A. Semin. Neurol. 1999; 19: 113-128Crossref PubMed Scopus (111) Google Scholar). The virus predominantly infects microglia and macrophages where it causes a productive infection. Astrocytes are also frequently infected with HIV but produce limited viral replication (2Nath A. Semin. Neurol. 1999; 19: 113-128Crossref PubMed Scopus (111) Google Scholar). The relevance of this viral infection remains unknown. Astrocytes maintain a barrier between blood and brain and provide neuronal support functions. Therefore, astrocytes may serve as a reservoir for the virus or induce neuronal damage by loss of neuronal support functions or release of cellular and viral products. Hence, infection of these cells could potentially have long term consequences on cerebral function. A striking feature is the presence of Tat transcripts in brains of HIV-infected individuals (3Wesselingh S.L. Power C. Glass J.D. Tyor W.R. McArthur J.C. Farber J.M. Griffin J.W. Griffin D.E. Ann. Neurol. 1993; 33: 576-582Crossref PubMed Scopus (403) Google Scholar, 4Hudson L. Liu J. Nath A. Narayan O. Male D. Jones M. Everall I. J. Neurovirol. 2000; 6: 145-155Crossref PubMed Scopus (205) Google Scholar). HIV Tat is a non-structural regulatory protein of 15 kDa that transactivates viral and cellular genes. It is produced in the early phase of infection and is actively released from infected lymphoid cells extracellularly (5Chang H.C. Samaniego F. Nair B.C. Buonaguro L. Ensoli B. AIDS. 1997; 11: 1421-1431Crossref PubMed Scopus (393) Google Scholar). Extracellular Tat may be internalized by uninfected cells, or it may interact with the cell membrane initiating a cascade of events (6Ma M. Nath A. J. Virol. 1997; 71: 2495-2499Crossref PubMed Google Scholar). Tat protein has both direct and indirect neurotoxic activity mediated by interaction with glutamate receptors, by disruption of cytokine network, and by reducing the neuroprotective effects of astrocytes (7Rappaport J. Joseph J. Croul S. Alexander G. Del Valle L. Amini S. Khalili K. J. Leukocyte Biol. 1999; 65: 458-465Crossref PubMed Scopus (144) Google Scholar, 8Koller H. Schaal H. Freund M. Garrido S.R. von Giesen H.J. Ott M. Rosenbaum C. Arendt G. Eur. J. Neurosci. 2001; 14: 1793-1799Crossref PubMed Google Scholar, 9New D.R. Maggirwar S.B. Epstein L.G. Dewhurst S. Gelbard H.A. J. Biol. Chem. 1998; 273: 17852-17858Abstract Full Text Full Text PDF PubMed Scopus (160) Google Scholar, 10Kolson D.L. Collman R. Hrin R. Balliet J.W. Laughlin M. McGann K.A. Debouck C. Gonzalez-Scarano F. J. Gen. Virol. 1994; 75: 1927-1934Crossref PubMed Scopus (47) Google Scholar, 11Johnston J.B. Zhang K. Silva C. Shalinsky D.R. Conant K. Ni W. Corbett D. Yong V.W. Power C. Ann. Neurol. 2001; 49: 230-241Crossref PubMed Scopus (111) Google Scholar). However, neither is the role of intracellularly expressed Tat on astrocyte function known, nor is the relative role of Tat uptake by neurons or its membrane interactions understood. Tat can produce oxidative stress in microglia cells (12Nicolini A. Ajmone-Cat M.A. Bernardo A. Levi G. Minghetti L. J. Neurochem. 2001; 79: 713-716Crossref PubMed Scopus (49) Google Scholar), monocytes (13Foga I.O. Nath A. Hasinoff B.B. Geiger J.D. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 1997; 16: 223-229Crossref PubMed Scopus (42) Google Scholar), and T lymphocytes. In the latter, intracellular expression of Tat leads to down-regulation of mitochondrial superoxide dismutase, thereby causing impaired mitochondrial membrane potential (14Westendorp M.O. Shatrov V.A. Schulze-Osthoff K. Frank R. Kraft M. Los M. Krammer P.H. Droge W. Lehmann V. EMBO J. 1995; 14: 546-554Crossref PubMed Scopus (367) Google Scholar, 15Macho A. Calzado M.A. Jimenez-Reina L. Ceballos E. Leon J. Munoz E. Oncogene. 1999; 18: 7543-7551Crossref PubMed Scopus (61) Google Scholar). Oxidative damage is difficult to measure directly; however, oxidized endogenous macromolecules such as free protein carbonyl and 4-hydroxynonenal (HNE) can serve as indicators of oxidative damage (16Wilt S.G. Dugger N.V. Hitt N.D. Hoffman P.M. J. Neurosci. Res. 2000; 62: 440-450Crossref PubMed Scopus (22) Google Scholar). Lipid peroxidation leads to the formation of HNE, a lipophilic alkenal that forms stable adducts on mitochondrial proteins (17Picklo M.J. Amarnath V. McIntyre J.O. Graham D.G. Montine T.J. J. Neurochem. 1999; 72: 1617-1624Crossref PubMed Scopus (128) Google Scholar). It has been suggested to be the key mediator of oxidative stress-induced cell death (18Liu W. Kato M. Akhand A.A. Hayakawa A. Suzuki H. Miyata T. Kurokawa K. Hotta Y. Ishikawa N. Nakashima I. J. Cell Sci. 2000; 113: 635-641Crossref PubMed Google Scholar). Hence, in this study, we have determined the effect of Tat on mitochondrial function and other indicators of oxidative stress in astrocytes and neurons. Tat-86 from HIVHXB-2 was cloned from pGEX-Tat into pcDNA3 vector atBamHI/EcoRI driven by a cytomegalovirus promoter. Further, Tat-86 and a deletion mutant of Tat-86 from which amino acids 48–56 were deleted by PCR (ΔTat), was cloned in-frame upstream of green fluorescent protein (GFP) gene driven by cytomegalovirus promoter in pEGFP-N1 (Clonetech) (Fig.1A). The HIV long terminal repeat (LTR)-driven GFP construct was made in pEGFP vector by deleting the cytomegalovirus promoter and inserting the LTR at SalI and SmaI. The PCR-cloned sequences were verified by double strand DNA sequencing. Tetracycline (tet) “on” system was used for generation of inducible constructs. Tat-86 was cloned downstream of a tet responsive element in pTREX vector (Clonetech). Reverse tetracycline transactivator (rtta) was first cloned in pEGFP vector (Clonetech) at BamHI/EcoRI and further subcloned in GFAP promoter-driven vector pGfaLac-1 atBamHI/BglII (Fig. 1B). Human fetal brain specimens were obtained from fetuses of 12–14 weeks gestational age. Neuronal cultures were prepared as described previously (19Magnuson D.S. Knudsen B.E. Geiger J.D. Nath A. Ann. Neurol. 1995; PubMed Scopus Google Scholar). the cells were in with fetal and and The cells were for at to in Human fetal rat cells, human astrocytic of the E. J. Sci. S. A. PubMed Scopus Google Scholar), and cells were in with fetal and cell fetal astrocytes cell or human astrocytes were transfected with using as were for GFP or for Tat using a for and stably transfected cell lines, the cells were in at GFP were a using and for weeks in The cells were used and in We and cell for the Tat in were determined by a where a Tat was at the of the and Tat in was used as the L. Liu J. Nath A. Narayan O. Male D. Jones M. Everall I. J. Neurovirol. 2000; 6: 145-155Crossref PubMed Scopus (205) Google Scholar). The of was Tat protein produced in was used as a (6Ma M. Nath A. J. Virol. 1997; 71: 2495-2499Crossref PubMed Google Scholar). In was from cells by and with of was reverse as described previously I. A. K. J. Neurovirol. 1999; PubMed Scopus Google into using and of reverse at for and for in of was used in PCR with of and The following were and and and and and The were on with and by a in were in for 15 The cells were with phosphate-buffered saline with for at and with in for The cells were at for with of the following in anti-Tat protein a neuronal and The cells were in by with with or for at in specimens were by fluorescent transactivation were using or cells. A stable cell was using the and with for The presence of was confirmed following a with and were transfected with activity was by The from these cells were at and which was to or cells by a The were used Tat Tat-expressing cells were with cells or a lymphocytic cell R.D. J. Acquir. Immune Defic. Syndr. 1999; PubMed Google in for in fetal with the of In Tat at was for of Tat activity and observed by fluorescent was at time using from the or was using reverse transcription and was at for with for and gene The were by a The were to and of and were expressed as or and cells were with 3-nitroproprionic acid for of time at were in and for The of protein was determined by oxidized protein Biol. PubMed Scopus Google Scholar). were for with and in and with of protein was by the were with with and for at to protein for by to for at were in with and with were by for were by that was used as a function was by a fluorescent to measure mitochondrial as previously described J. C. Zhang J. Liu Y. P.M. I. W. R. Nath A. Neurosci. 2001; PubMed Scopus Google Scholar). When in the cells, a green at membrane potential and at membrane the of the the cells were with at for and in were made with at and at and The of at both were expressed as of from which the of the with the was and using of cell Tat as and cell and were in fetal cells were in the of and neurons in the of the or in fetal cells were induced to and up to with In heparan or kynurenate were to the of the Cell death was by assay as described previously (19Magnuson D.S. Knudsen B.E. Geiger J.D. Nath A. Ann. Neurol. 1995; PubMed Scopus Google Scholar). were and neuronal were determined by to the cells were in was in and were The are expressed as of cells. cells were by a or cells were injected into the of the dentate of using the following and to and to the were for and by with saline by The brains were for in and in for were prepared from the at the and to the were for G. R. Ann. Neurol. 1999; PubMed Scopus Google Scholar). at the were also for Tat using a Tat astrocytes can support the expression of Tat protein and to the of we transfected human astrocytes or cells with DNA Tat expression was by (Fig. and and was further confirmed by LTR transactivation in cells for Tat or LTR transactivation was in cells and of Tat expression was also in human fetal astrocytes transfected with (Fig. The expression of Tat in astrocytes was confirmed by for GFAP and (Fig. Tat expression of Tat was expressed in the and with a in the as by green (Fig. and The were with astrocytes transfected with expression the cell with of or and and were by and by cell The of Tat and expression in these stable cell was to that of transfected astrocytes with the and with The expression of Tat in stable cell was further confirmed by of Tat by or transactivation assay following of these cells with the (Fig. cells of transactivation activity (Fig. of and cells that activity were used for further active Tat can be released from Tat-expressing astrocytes and taken up by cells, cells were with cells. GFP expression in of GFP expression was these cells were with cells (Fig. of or with cells also GFP effect was with or cells (Fig. cells with and Tat of GFP expression of (Fig. and these the LTR transactivation in both astrocytic and lymphocytic cells was for the Tat-expressing cells. further that the effect was to release of Tat from the Tat-expressing cells and uptake by the cells we the in the presence of Tat The cell to cell LTR transactivation activity was by Tat which is of release of Tat The of the to the LTR transactivation is to to Tat by the in of cell to cell The transactivation was for Tat effect was with or by of the cells with cells (Fig. Tat in a from cells was as by However, Tat from the stable Tat-expressing cells at and than transactivation of LTR with the or cells, that of Tat may be for LTR transactivation. is with where the Tat for LTR transactivation and HIV replication was determined to be B. Buonaguro L. G. V. R. R. R. J. Virol. 1993; PubMed Google Scholar). Tat was stably expressed in we determined the effect of Tat expression on as as and cytokine gene expression astrocytes transfected with Tat for that and were were we effect on or which was previously to in astrocytes with Tat protein K. A. Nath A. McArthur J.C. W. Power C. Sci. S. A. 1998; PubMed Scopus Google Scholar). gene gene and were The were also confirmed by Further, and were in the from and cells by the of Tat expression was in these cells by and cells the construct were used as a of cells green of Tat transfected human fetal gene gene and were in a gene gene and were further the effects of Tat on of astrocytes to oxidative we protein and peroxidation the cells with mitochondrial endogenous protein and were as of oxidative stress (16Wilt S.G. Dugger N.V. Hitt N.D. Hoffman P.M. J. Neurosci. Res. 2000; 62: 440-450Crossref PubMed Scopus (22) Google Scholar). Tat-expressing astrocytic cells decreased and protein with the (Fig. effects were for astrocytes as cells Tat have a effect (Fig. the effect of Tat-expressing cells on stable Tat-producing or cells were with neurons using and at time The mitochondrial of neurons was the first but membrane potential decreased in neurons with cells as with cells or cells (Fig. were confirmed by the of expression of Tat in cells (Fig. for of Tat. Hence, in with the we a in mitochondrial at to the was in cultures where was or was to the neuronal cultures with cells where the Tat gene was were further confirmed by from stably Tat-producing cells of and it to neurons for The decreased the neuronal mitochondrial membrane potential at (6Ma M. Nath A. J. Virol. 1997; 71: 2495-2499Crossref PubMed Google Scholar, D.S. Knudsen B.E. Geiger J.D. Nath A. Ann. Neurol. 1995; PubMed Scopus Google have that neurotoxicity is mediated amino acid and Tat to heparan and sulfate. we Tat-expressing astrocytes with neurons in in the presence of of these these attenuated the neurotoxicity as in mitochondrial In cultures where was in the presence of heparan effect was thus that Tat is for the neurotoxicity. the effect of Tat-expressing astrocytes on neuronal we neuronal cultures to cells in neuronal cell death with cells However, cells produce neurotoxicity (Fig. that amino acids 48–56 are essential for neurotoxicity. the neurotoxic properties of cells were in with neurons. cells induced neuronal cell death first at of with neurotoxicity at with where was or with cells with (Fig. Further, neuronal cell death could be using Tat but (Fig. that release of Tat is for neurotoxicity. is a the neurons in to toxic and loss of is a prominent feature of patients with HIV E. Wiley C.A. M. I. Ann. Neurol. 1997; PubMed Scopus Google Scholar). we in neurons following with Tat-expressing of GFP cells with human fetal neurons for induced prominent trimming or loss as in following Tat production was confirmed by direct of GFP Tat-producing cells could neurotoxicity in or cells were injected into the rat from the that or cells produced damage of the dentate granule cells, but with cells damage in the cell (Fig. to the of neuronal injury were observed in with cells, damage to the dentate and cells was observed in with cells (Fig. and Tat at the was to the cells (Fig. that these cells are to and produce Tat in Tat was in the region in cells that both microglial cells and neurons (Fig. the that produced Tat from the cells could be and taken up by cells in the the effect of Tat on we used and inducible expressed Tat in astrocytes in to that used Tat protein A. Geiger J.D. 1998; PubMed Scopus Google Scholar). We used astrocytic cell and for intracellular Tat direct expression through cytomegalovirus promoter and expression using a human GFAP promoter. were obtained with both cell and both for Tat of a that intracellularly expressed Tat was toxic to In it resistance to such as We also of of of has been associated with of however, has effect S. A. PubMed Scopus Google Scholar). is in with that HIV infection of astrocytes is A. V. M. J. Neurol. 1995; PubMed Scopus Google Scholar). We were thus to produce stably Tat-expressing astrocytic cell lines. We expressed Tat in astrocytes Tat can be in HIV-infected astrocytes in L. Liu J. Nath A. Narayan O. Male D. Jones M. Everall I. J. Neurovirol. 2000; 6: 145-155Crossref PubMed Scopus (205) Google and in as in this of Tat was by transactivation and Further, Tat released from these cells was active it in lymphocytic and astrocytic cell lines. are with a D.L. Collman R. Hrin R. Balliet J.W. Laughlin M. McGann K.A. Debouck C. Gonzalez-Scarano F. J. Gen. Virol. 1994; 75: 1927-1934Crossref PubMed Scopus (47) Google where Tat was taken up by neurons and was to the A of Tat expression was intracellularly in Tat was in the by the and of the amino acids 48–56 in the and that this region of Tat is for its S. A. R. K. R. R. C.A. J. Virol. PubMed Google Scholar, Full Text PDF PubMed Scopus Google Scholar, 1998; PubMed Scopus Google Scholar). EMBO J. PubMed Scopus Google Scholar, H. A. D.G. M. 1998; PubMed Scopus Google Scholar, E. B. J. Biol. Chem. 1997; Full Text Full Text PDF PubMed Scopus Google have that the arginine-rich domain of Tat acids is for or expression or the of into the it to In astrocytic cells, LTR transactivation occurs by interaction of Tat with transactivation element of LTR or interaction of Tat with sites in the LTR L. J.M. M. 1997; PubMed Scopus Google Scholar). The arginine-rich domain is thus essential for both of LTR transactivation and as also essential for neurotoxicity. this region may be target for of The of Tat to the remains we were to effect of intracellularly expressed Tat on and production in astrocytes using which was further confirmed by and for and it was that transfected Tat induced and produced in cell M. S. S. J. C. K. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). However, gene and on astrocytes as as the astrocyte cell support these a cell stably Tat of by J. Gen. Virol. PubMed Scopus Google Scholar). In A. Conant K. C. J. Biol. Chem. 1999; Full Text Full Text PDF PubMed Scopus Google Scholar), and we have that Tat has effects on cytokine and production by Thus and Tat may have We of to that extracellularly released Tat is from the Tat-expressing cell produced neurotoxicity as by of and neuronal cell The neurotoxicity could be by heparan and to that the neurotoxicity is mediated Tat. the effect of Tat on on neurons (19Magnuson D.S. Knudsen B.E. Geiger J.D. Nath A. Ann. Neurol. 1995; PubMed Scopus Google Scholar), heparan and to Tat (5Chang H.C. Samaniego F. Nair B.C. Buonaguro L. Ensoli B. AIDS. 1997; 11: 1421-1431Crossref PubMed Scopus (393) Google Scholar, M. M. M. J. Biol. Chem. 2001; Full Text Full Text PDF PubMed Scopus Google Scholar). that Tat protein was toxic to neurons (19Magnuson D.S. Knudsen B.E. Geiger J.D. Nath A. Ann. Neurol. 1995; PubMed Scopus Google Scholar). However, the of Tat to cause the neurotoxicity was lower in the study, that Tat produced by cells is a potent J. C. Zhang J. Liu Y. P.M. I. W. R. Nath A. Neurosci. 2001; PubMed Scopus Google that to Tat could neurotoxicity produced by from HIV-infected the relevance of these We cell to cell of Tat astrocytes and from astrocytes to using expression cells that were to Tat-producing function was also amino acids 48–56 effect was with this of Tat-expressing cells with the cell lines, from the stably Tat-expressing cells to induce LTR transactivation. was to lower concentrations of Tat in the However, Tat released from the same cell was to cause that the concentrations of Tat for causing neurotoxicity are lower than that needed for transactivation of may in be by the that neurotoxicity interaction of Tat with the neuronal cell membrane J. Nath A. Knudsen B. S. Geiger J.D. M. 1998; PubMed Scopus Google Scholar), transactivation of LTR Tat uptake by the cell Y. Jones M. G. N. R.D. Nath A. 2000; 6: PubMed Scopus Google Scholar). in was that Tat released from the astrocyte cell was taken up by granule cells in the dentate and transported to the region where it caused glial cell activation and neurotoxicity. Thus Tat can cause pathological at sites distant to the of the following of HIV replication in Tat is produced as early It the astrocytes to oxidative thus it reservoir for survival of the viral Tat to the but are released Tat may be taken up by cells in the brain or cells through the these cells are infected with the it cause transactivation of the viral and viral taken up be neurons it may be transported along neuronal pathways and cause neurotoxicity and glial cell activation at distant of Tat are to cause neuronal damage that mitochondrial dysfunction, and neuronal cell death in are in neurites, the mitochondrial stress may be for HIV-infected astrocytes a role in the neuropathogenesis of HIV infection. The Tat and in the arginine-rich region of as a target for HIV infection. We G. for M. for D. for cells, W. for cells, and C. for We the of and for and and cells. We for and for to Tat. We C. R. and for