Induction of cell differentiation by human immunodeficiency virus 1 vpr

Cell, Vol. 72, 541-550, February26, 1993, Copyright 0 1993 by Cell Press

Induction of Cell Differentiation by Human lmmunodeficiency Virus 1 vpr David N. Levy,“+ Laura S. William V. Williams,* and *Institute of Biotechnology and Advanced Molecular Department of Medicine University of Pennsylvania tWistar Institute Philadelphia, Pennsylvania

Fernandes,’ David B. Weiner’t Medicine

19104

Summary Cell lines from rhabdomyosarcomas, which are tumors of muscle origin, have been used as models of CD4independent HIV infection. These cell lines can be induced to differentiate in vitro. We report here that the vpr gene of HIV1 is sufficient for the differentiation of the human rhabdomyosarcoma cell line TE671. Differentiated cells are characterized by great enlargement, altered morphology, lack of replication, and high level expression of the muscle-specific protein myosin. We have also observed the morphological differentiation and inhibition of proliferation of two other transformed cell lines. vpr-transfected cells remain fully viable in culture for extended periods. These observations elucidate a potential role for vpr in the virus life cycle and raise the possibility that some aspects of HIV-induced pathologies may be caused by a disturbance of cells by vpr. Introduction The development of mature skeletal muscle cells entails an ordered process of cellular differentiation from musclecommitted myocytes (presumptive myoblasts), to postmitotic myoblasts, to mature multinucleated myotubes possessing a functional muscle-contractile apparatus. Embryonal rhabdomyosarcoma is a cancer of cells resembling presumptive myoblasts and may originate from muscle satellite cells (Bruni, 1979). Rhabdomyosarcoma cell lines have been used in studies of muscle differentiation and tumorigenesis, and they can be induced to differentiate from a rapidly dividing population of cells (myoblastlike) that express low amounts of a few mature-muscle proteins to postmitotic, greatly enlarged and elongated, multinucleated (myotube-like) cells that express high amounts of mature-muscle-specific proteins and a functional muscle-contractile apparatus (Aguanno et al., 1990; Hiti et al., 1989; Siegel and Lukas, 1988; Stratton et al., 1989). Rhabdomyosarcoma cell lines, including the related RD and TE671 lines, have additionally been studied as models of CD4 receptor-independent human immunodeficiency virus (HIV) infection and pathogenesis (Clapham et al., 1989, 1991; Haggerty et al., 1991; Srinivasan et al., 1988; Tateno et al., 1989; Weber et al., 1989; Weiss et al., 1988; Werner et al., 1990). As the etiologic agents of acquired immune deficiency

syndrome (AIDS), the HIV1 and HIV2 viruses are responsible, directly or indirectly, for a host of clinical symptoms, including immune system degeneration and wasting as well as muscular and neurological disorders (reviewed in Cohen et al., 199Oc). As for all retroviruses, HIV is dependent on the transcription and translation machinery of the host cell for infection and replication (Cann and Karn, 1989). The ability to interactwith that machinery, however, via its regulatory proteins distinguishes the lentiviruses from most other retroviruses, and the high number of regulatory proteins in the primate immunodeficiency viruses makes them unique among lentiviruses. The conservation of these regulatory genes indicates a vital role in the natural history of the virus. Knowledge of the physiological activities of each regulatory protein will be required to understand more fully the ability of HIV to infect diverse lineages, to infect nonreplicating cells, to establish latent and chronic infections, and to induce disease. Two of the HIV regulatory genes, fat and rev, are required for viral replication in all cell types (Dayton et al., 1986; Fisher et al., 1986; Sodroski et al., 1986). Each of the other, so-called nonessential, regulatory genes, vif, nef, vpr, HIV1/SIVcpZ vpu, and HIV2/SIVuAc vpx, can be deleted without destroying virus replication in many cell lineages in vitro (reviewed in Terwilliger, 1992). However, vif and vpu are required for efficient production of infectious viral particles (Fisher et al., 1987; Strebel et al., 1987, 1988; Terwilliger et al., 1989), and nef and vpr may be necessary for maintenance of infection in vivo (Kestler et al., 1991; Terwilliger, 1992). tat, rev, nef, vpr, and vpx have been shown to influence the genetic regulation of HIV expression (Terwilliger, 1992). Mice transgenic for the tat gene develop tumor-like growths resembling HIV-associated Kaposi’s sarcoma (Vogel et al., 1988), suggesting that expression of viral regulatory proteins during HIV infection might directly lead to morbidity. Since the demonstration in 1987 that the small open reading frame within HIV1 designated R encodes a 15 kd protein (Wong-Staal et al., 1987), relatively little regarding the function of the viral protein R (vpr) has been reported. The vpr open reading frame is conserved within all genames of HIV1 and HIV2 and within most, if not all, simian immunodeficiency virus (SIV) genomes. Vpr is immunogenic in vivo in that a large subset of HIV’ individuals make antibodies that can react with a bacterially produced vpr peptide (Wong-Staal et al., 1987). The most direct evidence for the function of the vpr protein comes from several studies reporting that vpr increases the replication and cytopathogenicity of HlVl, HIV2, and SIV in primary CD4+ T lymphocytes and transformed T cell lines (Ogawa et al., 1989; Shibata et al., 1990a, 1990b), although others have reported no effect of vpr on replication (Dedera et al., 1989). Interestingly, HIV2 mutated for vpr has been reported unable to infect primary monocytelmacrophages (Hattori et al., 1990). Transactivation of the HIV long terminal repeat and heterologous promoters by HIV is in-

Cell 542

Results

Figure 1. Effect of the Transfection domyosarcoma Cells

of HIV1 Genome into TE671 Rhab-

(A) Untransfected TE671 cells. (8 and C) TE671 cells transfected with pNLpuro and selected in puromycin for 6 days (as described in Experimental Procedures). (C)shows a large binucleated cell with a particularly large, radical phenotype. Magnification of all three panels is identical. Note: TE671 cells transfected with murine retroviral vector pBABE-puro and selected in puromycin presented the identical phenotype to (A) (data not shown).

creased about 3-fold in wild-type versus vpf-negative HlVl, though the mechanism through which vpr may upregulate transcription is unknown and may be indirect (Cohen et al., 1990b). The relationship between the effects of vpr on promoter activity and viral infectivity is not clear. Vpr protein is incorporated into the viral particle, and this finding has led to the proposition that vpr functions early in infection, following virus penetration and uncoating, and that vpr may interact with cellular regulatory mechanisms important in the establishment of infection (Cohen et al., 1990a; Yu et al., 1990; Yuan et al., 1990). We report here that HIV expression in a human muscle cell tumor line leads to inhibition of proliferation and activation of the suppressed endogenous cell differentiation program. The vof gene of HIV1 is sufficient for the observed effects and necessary for differentiation of essentially all the cells. These results establish HIV1 vpr as a regulatory protein capable of profound regulation of cell functions, including cell proliferation and differentiation.

Differentiation of TE671 Following Transfection with Genomic Construct pNLpuro We constructed a drug-selectable env deletion mutant HIV1 genomic plasmid, pNLpuro, based on the HIV1 infectious molecular clone pNL43 (see Experimental Procedures). We transfected pNLpuro into the human rhabdomyosarcoma cell line TE671 (McAllister et al., 1969; Stratton et al., 1969) and selected for stable transfectants. TE671 cells normally grow as small mononuclear round or polygonal fibroblast-like adherent cells about 3-7 urn in size (Figure 1A). Surprisingly, following transfection of TE671 cells with pNLpuro, the phenotype of most of the resulting puromycin-resistant cells was distinctly different from that of untransfected TE671 cells. Over the course of several days the phenotype of the pNLpuro-transfected cells became progressively more radical; many achieved sizes exceeding 120 pm in length and developed long, sometimes branched, processes (Figures 16 and 1C). Some cells became large, flat, and irregularly shaped. Differentiating cells often became bi- or multinucleated (Figure lC), though cells with long processes that resembled myotubes lacked the linear arrangement of many nuclei that is found in true myotubes. The large cells remained fully viable for several weeks. The differentiation of TE671 cells via chemical agents is a well-described phenomenon, as this ceil line has been used as a model for skeletal muscle differentiation (Aguanno et al., 1990; Hiti et al., 1969; Siegel and Lukas, 1968; Stratton et al., 1989). Some agents, including protein kinase C-activating phorbol esters, as well as serumdepleted medium, will induce TE671 cells to undergo alterations in both morphology and growth characteristics that parallel in many aspects the differentiation of myoblasts into myotubes (Aguanno et al., 1990; Siegel and Lukas, 1988; Stratton et al., 1989). Phorbol myristate acetatestimulated TE671 cells increase in overall size and length, are often multinucleated, and display slowed proliferation (Aguanno et al., 1990). On the other hand, no morphologically differentiated pNLpuro-transfected cells were observed to divide when followed for up to 10 days. Determination of HIV1 Elements Sufficient for Differentiation We next sought to define the viral gene(s) responsible for induction of cell differentiation in the rhabdomyosarcoma cell line. We chose to examine individual HIV1 regulatory genes, as some of their protein products have been reported to influence cellular events. We obtained and modified a tat expression vector (Arya et al., 1985) to permit selection of stably transfected cells. The env-rev region of HIV1 was amplified by the polymerase chain reaction (PCR) from an HIV1 molecular clone and subcloned directly into the pCDNAl/neo expression vector. To clone the remaining regulatory genes into expression vectors, we amplified the single open reading frame of each gene using PCR. During the PCR amplification a consensus ribosome initiation sequence (Kozak, 1988) was intro-

HIV1 Cell Differentiation 543

-R.T.

+R.T. vff

vpr

Figure 2. Transfection myosarcoma Cells

vpu

tat

naf Ivtf

vpr

vpu tat

nef

of HIV1 Regulatory Genes into TE671 Rhabdo-

(Top) Transfection of TE671 with HIV1 regulatory genes to determine viral elements involved in differentiation. Cells in log phase growth were transfected with plasmids coding for HIV1 regulatory genes as described in Experimental Procedures. Eight days after transfection the remaining drug-resistant cells were photographed. Plasmids transfected: vpu(A), nef (B), fat (C), vpr(D), vif (E), env-rev (F). (Bottom) Demonstration of regulatory gene expression in TE671 by reverse transcription PCR analysis. TE671 cells transfected with plasmids containing the individual HIV1 regulatory genes vif, vpr, vpu, fat, and nef were subjected to reverse transcription PCR analysis as described in Experimental Procedures. PCR products were run on 2% agarose gels and stained with ethidium bromide for photography. Each band corresponds to the correct length of the respective regulatory gene open reading frames (+R.T. lanes); however, the intensities of the bands cannot be taken as a measure of the relative expression of each gene. As a control against possible DNA carryover in the RNA preparations, RNA that was not subjected to reverse transcription was used as template in PCR (-R.T. lanes).

duced immediately upstream of each start codon. nef, vpr, vif, and vpu were amplified by this method and subcloned into the pBABE-puro expression vector. Each plasmid was transfected into TE671 cells and selected on the appropriate antibiotic. Expression of rev was demonstrated indirectly by showing expression of envelope protein (env) in Western blot and cell fusion assays (data not shown). Since expression of env is dependent on a critical thresh-

old level of rev expression, it can be deduced that physiologically relevant levels of rev were produced. The vpr protein expressed from the vpr vector shows a single 15 kd species by Western blot analysis (data not shown). Cellular expression of tat, vif, vpr, vpu, and nef was demonstrated by reverse transcription PCR analysis (Figure 2, bottom). Expression of vpu (Figure 2A), vif (Figure 2B), tat (Figure 2C), nef (Figure 2E), and rev and env (Figure 2F) failed to induce significant morphological changes in TE671 cells. Vpr expression (Figure 2D), on the other hand, induced profound differentiation in the majority of transfected cells. To verify that TE671 differentiation involves the development of the welldefined muscle phenotype and not a novel program, we stained vpr-transfected TE671 cells with an antibody specific for the heavy chain of fast-twitch skeletalmuscle myosin, which is expressed at high levels only in mature skeletal muscle cells. Antibody MY-32 reacted strongly with vpr-transfected TE671 cells (Figures 3A and 38). The majority of untransfected TE671 cells expressed low levels of myosin, though, as previously reported for TE671, a few untransfected cells stained weakly for myosin (Figure 3C). Staining with an isotype-matched control antibody was negative for both transfected (Figure 3D) and untransfected TE671 cells (data not shown). The transfection efficiency achieved by the vpr vector was equal to the efficiency of transfection of the other vectors. Transfection efficiency was determined by the number of cells remaining after selection in puromycin for 2 days, which is sufficient time to kill all nontransfected cells. After several more days, there appeared to be many fewer cells in the vpr culture than in the non-vpr cultures, owing to the continued replication of the cells in the non-vpr cultures. Not all cells transfected with either the genomic pNLpuro plasmid or with vpralone underwent morphological differentiation, however. This result is consistent with

Figure 3. Expression

of Myosin in vpr-Transfected

TE671 Cells

Cells were plated on glass slides and allowed to adhere. Processing for immunofluorescence was as described in Experimental Procedures. Primary antibodies were MY-32 anti-myosin (A, B, and C) and SIM.4 anti-CD4 isotype-matched control (D). Transfections: HIV1 vpr(A, 6, and D) and untransfected TE671 (C).

Cell 644

Figure 4. Effects of vpr in RD Rhabdomyosarcoma

and D17 Osteosarcoma

Cells and of vpr-negative

HIV1 DNA on TE671 Cells

(A-D) Effects of vpr on RD rhabdomyosarcoma and D17 osteosarcoma cells. RD cells are shown untransfected (A) or transfected with HIV1 vpr (B). D17 cells are shown untransfected (C)or transfected with HIV1 vpr(D). Both RD and D17 cells transfected with vpr were selected in puromycin and photographed 6 days after transfection. Inhibition of proliferation was observed in both lines. Time-lapse video microscopy of Di 7 cells showed them to be very active. The central or perinuclear regions of many cells rotated with a period of approximately 2 hr, frequently resulting in a distinct crescent shape. (E and F) Effect of deletion of vprfrom HIV1 genome on its ability to induce differentiation in TE671. TE671 cells were transfected either with pNLpuro (E) as in Figures 1 Band 1C or the vprdeletion construct pNLpuroAvpr (F) as described in Experimental Procedures. Puromycin-resistant cells were selected and photographed 6 days after transfection.

the heterogeneous response observed in rhabdomyosarcoma lines subjected to differentiation-inducing conditions (Aguanno et al., 1990). More cells remained undifferentiated in the vpr-transfected cultures (lo%-20%) than in the pNLpuro cultures (
for at least 2 weeks and did not proliferate, though some small proliferating cells remained in both the D17 and RD transfectants, as was observed in the TE671 cultures. The D17 osteosarcoma cells did not express increased levels of alkaline phosphatase, however, which is a marker for bone maturation (data not shown). Deletion of vpr from the HIV1 Genome We next examined whether vpr is necessary for HIVlinduced differentiation. To this end we constructed a vpr deletion mutant of the HIV env plasmid pNLpuro called pNLpuroAvpr. A stop codon was introduced after amino acid 22 of vpr, and 122 nt were removed from the coding region of vpr from amino acid 22 to amino acid 62 (see Experimental Procedures). First, we tested whether deletion of the vpr gene affected the expression of HIV1 genes, since such an effect might complicate the interpretation of the experiments. Viral protein expression following transfection with pNLpuroAvpr was equal to expression from pNLpuro, as measured by p24gsgprotein released into the culture medium (data not shown). Since expression of structural genes by HIV is dependent on successful expression of both tat and rev proteins, it is apparent that the mutation introduced into the genome did not result in a general disturbance of HIV transcription or RNA splicing. As a further control to examine the effect of deletion of vpr on HIV1 expression, a vpr deletion mutant was constructed from wild-type pNL43, by the same method as

HIV1 Cell Differentiation 545

Table I. HIV1 Constructs

Used in This Study

Construct

Vector or Derivation

“Pr

Differentiation

p24 Production

Characteristics

pNLpuro w “if “PU nef tar env-rev pNLpuroAvpr pNL43 pNLAvpr

HIV1 pNL43 pBABE-puro pBABE-puro pBABE-puro pBABE-puro pcv-1 pcDNAl/neo pNL43 pNL43 pNL43

+ +

+++ +++

+++ N/A N/A N/A N/A N/A N/A +++ +++ +++

env deletion mutant

-

+

-I+” NT NT

em+, syncytia:++ env deletion mutant em+, syncytia in TE671rq:++ em+, syncytia in TE671V:++

p24 release into the culture medium was measured 46 hr following transfection, as described in Experimental Procedures. a Transient increase in size in some cells, with reversion of phenotype observed after approximately 6 days, and differentiation (So/a-10%). NT, not tested; N/A, not applicable.

that used for the vpr deletion mutant of pNLpuro, to yield an env+ construct, pNLAvpr. When transfected into TE671 w, a CD4’ transfectant of TE671 constructed in our laboratory (see Experimental Procedures), syncytia were efficiently produced by both the vpr and vpr’ constructs. Despite nearly equivalent HIV1 protein production between pNLpuro and pNLpuroAvpr, the outcome of transfection with the vpr deletion mutant, with respect to differentiation, was clearly different from that of transfection with the vpr+ HIV1 genome. While cells transfected with pNLpuro differentiated as discussed above (Figure 4E), the majority of the TE671 cells transfected with the vpr deletion mutant pNLpuroAvpr showed either no change or a small and transient increase in size and length (Figure 4F). Though a few myosin-staining morphologically differentiated cells were produced in each transfection, the efficiency of this effect varied from experiment to experiment and was never seen to exceed 10% of the cells remaining after drug selection. Taken together these results(summarized in Table 1) demonstrate that HIVl-induced differentiation of TE671 cells is a function primarily of the vpr gene. p24gw production in pNLpuroAvpr continued for 2-3 weeks following transfection and subculture, whereas p24 released from pNLpuro-transfected cells was eliminated following subculture. Subculture effectively eliminated the large differentiated cells, leaving only the replicating undifferentiated cells intact. Therefore, only the differentiated cells released virus in the pNLpuro-transfection experiments cells. Exposure of the transfectants to the protein kinase C-activating phorbol ester phorbol myristate acetate, which has been shown to stimulate HIV1 expression in chronically infected cells (Harada et al., 1986) resulted in a 3-fold increase in p24 release from the pNLpuroAvprtransfected cells but no measurable p24 release from the undifferentiated pNLpuro-transfected cells. This result indicates that, in the presence of vpr, HIV1 production in TE671 cells appears to be incompatible with their replication, whereas in the absence of vpr, HIV1 expression can continue in replicating cells. pNLpuroAvpr-transfected cells retain the ability to differentiate in response to various agents (data not shown) and thus remain relatively unaffected by HIV1 expression.

in a small subset

Demonstration That Infection with HIV1 Induces Differentiation To examine the relevance of HIV-induced differentiation to HIV infection, we next studied the effects of HIV on TE671qr cells following infection with HIV. For these experiments we used the cell line TE871~. TE671~ expresses high levels of CD4 on its cell surface (Figure 5A) and can be infected with HIV at very high efficiency, resulting in a high level of viral production (Figure 58). Infection of TE671yr cells at or near confluence results in cell fusion into giant multinucleated syncytia, owing to the fusion of cell membranes following coexpression of HIV envelope proteins and their receptor, CD4 (data not shown). To allow infection and maintenance in culture of unfused cells for several days following infection, TE671 w was plated at low cell density, typically 5% confluence or less. Cells plated at low cell density and left unexposed to HIV1 did not differentiate and continued to replicate (Figure 5C). Cells infected with HIV1 (strains RF or MN) differentiated in a manner very similar to that observed following transfection with the pNLpuro viral genome (Figure 5C). These results demonstrate that HIV infection can directly induce cell differentiation.

Discussion We report the unexpected observation that transfection of HIV1 genomic DNA into the embryonal rhabdomyosarcoma line TE671 induced cell growth inhibition and differentiation. Infection of TE671 via a transfected CD4 molecule resulted in the same outcome, indicating that the effects did not result from transfection artifacts and have relevance to natural HIV infection. Transfection and expression of each regulatory gene of HIV1 in the cell line revealed that the vprgene can produce the growth inhibition and morphological differentiation that the whole virus induces. Activation of the endogenous muscle program was demonstrated by showing that the vpr-transfected cells expressed high levels of fast-twitch myosin, while the majority of untransfected cells did not. Transfection of a vprdeletion mutant into TE671 cells resulted in the produc-

546

Cell

p24 gag EL!SA

B

TE671

Figure 5. Analysis of the CD4 Transfectant

TE671q~ and Differentiation

Following Infection

(A) TE671 and TE671~ were analyzed for surface expression of CD4, the cellular HIV1 receptor. Examination of surface expression of CD4 was determined by flow cytometer analysis using the anti-CD4 monoclonal antibody Leu-3a (see Experimental Procedures). The histograms of TE671 and TE671V are displayed individually. Broken line represents staining of irrelevant isotype-matched UPC-21 mouse monoclonal antibody. Solid line represents staining with Leu-3a antiCD4. (B) TE671 and TE671r~r were analyzed for their ability to be infected by HIV1 Following incubation with cell-free virus, target cells were washed once with trypsin and then twice with media to remove residual traces of virus. Infection was determined by detection of ~24~ protein in cell supernatants using a commercial p24 enzyme-linked immunosorbent assay kit. (C and D) Morphological changes of TE6711q induced by HIV1 virus infection. TE6711y cells were plated at low density, either incubated with media alone (C) or inoculated with HIV1 (strain RF) (D) as described in Experimental Procedures, and photographed 9 days after plating.

of large numbers of replicating undifferentiated cells that continued to produce high levels of viral protein. These results indicate that vpr is the primary determinant for differentiation and growth inhibition in TE671 cells. Transfection of vpr into the rhabdomyosarcoma line RD and the osteosarcoma line D17 resulted in cessation of proliferation, gross morphological changes, and profound enlargement. Thus, vpr may be a regulator of cell function in cells of diverse origin. The literature is consistent with a role for vpr early in infection. The findings that vpr is incorporated into the virion (Cohen et al., 1990a; Yu et al., 1990; Yuan et al., 1990) and that vpr-deficient virus fails to infect macrophages (Hattori et al., 1990) and our observation of a profound influence on cellular events all support this hypothesis. Vpx protein, which is structurally and evolutionarily related to vpr (Tristem et al., 1990,1992), is likewise incorporated into virions (Guyader et al., 1969; Kappes et al., 1991). Importantly, virion-associated vpx has been shown to be active in assisting early events following infection (Kappes et al., 1991). The morphology of vpr-negative virions is indistinguishable from that of wild-type virions (Cohen et al., 199Oa), suggesting that vpr does not constitute a structural protein involved in the integrity of the virion. The early events of infection, including reverse transcription and integration, depend on the expression of cellular tion

factors that assist these processes. Vpr may activate expression of cellular factors that assist the virus, or it may suppress factors that would otherwise inhibit virus replication. It is interesting to note that primary monocytelmacrophages have been reported to be infectible in vitro only while they are undergoing differentiation (Schuitemaker et al., 1992) and that HIV infection can result in the selection for more differentiated myelomonoblastic cells (Roulston et al., 1992); thus, in some cell types vpr may assist infection through induction of cell differentiation in vivo. This hypothesis remains to be tested. Muscle differentiation has been well studied in rhabdomyosarcomas and in normal cells (reviewed by Florini et al., 1991; Li and Olson, 1992; Tapscott and Weintraub, 1991; Weintraub et al., 1991). Expression of helix-loophelix transcription factors such as MyoD in normal myoblasts leads to differentiation into mature postmitotic myotubes. In most embryonal rhabdomyosarcomas, despite expression of MyoD, withdrawal from the cell cycle and differentiation are inhibited. Transformation of embryonal rhabdomyosarcomas is linked to expression of an activated ras oncogene, loss of a putative tumor suppressor on chromosome 11, and constitutive expression of autocrine fibroblast growth factor and transforming growth factor 6. In addition, RD (and therefore TE671) has been shown to lack a wild-type ~53 tumor suppressor gene (Fe-

HIV1 Cell Differentiation 547

lix et al., 1992). ~53 expression has recently been associated with cell cycle control and the regulation of DNA repair mechanisms (Livingstone et al., 1992; Yin et al., 1992) cellular events linked to retroviral integration. Osteosarcomas exhibit features of bone matrix-secreting osteoblasts but are thought to arise from multipotential mesenchymal tissue and therefore to represent similarly a disregulation of primitive cells. These tumors also typically display lossof-function ~53 mutations (Miller et al., 1990; Mulligan et al., 1990). Vpr can at least partially overcome the block on differentiation and completely restore inhibition of cell proliferation; therefore, vpr may either replace a function lost during transformation or activate a pathway that overrides the genetic defects. Cells that are subject to abortive infection, owing to an inability of thecell to support virus replication, might nevertheless be disregulated by the activities of vpr. Severe myopathies of unknown etiology, often associated with HIV infection, can precede the onset of AIDS and result in the disturbance of the expression of contractile proteins (Chernoff, 1990; Dalakas and Illa, 1991). Though skeletal muscle cells are not observed to be infected in HIV+ patients (Lamperth et al., 1990), it has been suggested that transient infection of skeletal muscle cells could initiate pathogenesis (Dalakas and Illa, 1991). Cardiac muscle cells have been shown to be infectible in vivo (Grody et al., 1990). The skeletal muscle tumor line TE671 and its parent line RD are infectible with diverse strains of HIV in a CD4independent manner and have been used extensively as models for HIV infection and pathogenesis (Clapham et al., 1989, 1991; Haggerty et al., 1991; Srinivasan et al., 1988;Tatenoetal., 1989; Weberetal., 1989; Weissetal., 1988; Werner et al., 1990). It is tempting to speculate that stem cell populations might be disrupted by vpr and lose the ability to replenish tissues, leading to a gradual decline of mature effector populations. Hematopoietic precursors have been shown to be infectible by HIV (Schnittman et al., 1990; Stanley et al., 1992). Rhabdomyosarcomas are believed to originate from skeletal muscle satellite cells, a stem cell population for the generation of new myotubes in adult muscle. The disruption of the differentiation of satellite cells is the favored model for the progressive nature of DuchennelBecker muscular dystrophy, and disturbance of differentiation by vpr could provide a mechanism for the muscle wasting observed frequently in AIDS patients. These studies directly demonstrate that the HIV1 vpr gene encodes a protein that can function in the regulation of basic cellular events in vitro. The outcome of this regulation is observed here as an inhibition of cell proliferation and the induction of differentiation. The effects of vpr on the normal targets of infection in vivo may be different; however, the biochemical aspects of vpr action can be elucidated using these lines. In light of previous studies demonstrating that vpr assists replication in some cell types, including macrophages and possibly T lymphocytes, we conclude that vpr probably acts via regulation of cellular events involved in their permissiveness to viral replication. Since vpr is incorporated into the virion, an unusual property for a regulatory protein, we suggest that

nonproductively infected cells may be disregulated in the absence of reverse transcription, integration, or residual viral nucleic acid. Therefore, tissues that are normally not observed to be targets of HIV infection in vivo but that are afflicted in AIDS patients should be examined in light of this phenomenon. Experimental

Procedures

Cell Lines and Cultivation The human embryonal rhabdomyosarcoma TE671 line (ATCC HTB 139) and the canine osteosarcoma 017 line (ATCC CCL 183) were obtained from the American Type Culture Collection. TE671 was originally classified as a medulloblastoma line (McAllister et al., 1977) but has recently been shown (Chen et al., 1989; Stratton et al., 1989) to be a subclone of the rhabdomyosarcoma cell line RD (ATCC CCL 136) (McAllister et al., 1969). RD cells were provided by Dr. A. Srinivasan. All cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, penicillin-streptomycin, and sodium pyruvate and maintained in a 5%-6% CO* atmosphere at 37%. Constructlon of the TE671y Cell Line TE671 cells were infected with replication-defective murine retrovirus containing the human CD4 retroviral expression vector T4-pMV7 (a generous gift of Dr. Daniel Littman, University of California at San Francisco) as previously described (Weiner et al., 1991). Clonal populations were analyzed for CD4 expression by flow cytometry as previously described (Weiner et al., 1989). In brief, cells were incubated with either Leu-Sa, a murine monoclonal antibody specific for the human CD4 cell surface molecule, or UPC-21, an irrelevant isotypematched murine monoclonal antibody. Secondary antibody was a flu@ rescein-labeled goat anti-mouse antibody. A stable CD4’ clone was selected for further analysis and designated TE671v. Plasmids and Cloning Strategies The HIV1 genomic clone pNL43 was obtained through the National Institutes of Health (NIH) AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, NIH, from Dr. Malcom Martin (Adachi et al., 1986) and was used as the starting material for most of the genetic constructs used in this study. This construct consists of HIV1 proviral DNA plus 3 kb of host sequence from the site of integration cloned into pUC18. All cloning design was greatly facilitated by the Macintosh computer program “DNA Strider” (Marck, 1988). Construction of pNLpur0 To simplify further cloning steps, the Stul site within the non-HIV 5’ flanking human DNA of pNL43 was destroyed by partial digestion with Stul followed by digestion of the free ends with Escherichia coli polymerase I. The linear plasmid was filled, then self-ligated, leaving a unique Stul site within the HIV genome. This plasmid, pNLAstu, was then digested with the blunting enzymes Stul and BsaBI, which eliminated a large section of the coding sequence for gp120. The SV40 promoter and puromycin resistance coding region (puromycin acetyltransferase) were isolated from pBABE-puro (Morgenstern and Land, 1990; kindly provided by Dr. Hartmut Land of the Imperial Cancer Research Fund) using EcoRl and Clal. This fragment was blunted, then cloned into the Stul-BsaBI-digested pNLAstu. A clone was selected with the SV40-puro fragment in the correct orientation so that the 3’long terminal repeat of HIV could provide poly(A) functions for the puromycin acetyltransferase message. This plasmid was designated pNLpuro. H/W Regulatory Genes nef, vpu, vpr,and vif were amplified via PCR from the HIV1 genomic plasmid pNL43. PCR was performed under conditions that yield DNA amplification with high fidelity (Ling et al., 1991). Three PCR primers were used to amplify each gene. One primer, called a universal start/ cloning primer (USP), encodes appropriate restriction sites for cloning the PCR product into a vector, a consensus sequence determined by Kozak (1986) to promote strong initiation of translation, and an ATG start site. This primer was used in the amplification of each gene. The USP was placed in a PCR with a second short primer consisting of a reverse complement of the 3’ end of the USP plus approximately 15

Cell 548

bp of the S’end of the target gene. The double-stranded product of this PCR was used as a 5’ primer in a second reaction with an appropriate 3’ primer specific for the gene of interest. The 3’ primer introduced restriction sites for cloning the PCR product. The resulting PCR products were cloned into the retroviral vector pEABE-puro. The sequences of the primers are as follows: USP: ggcggctcgaggatccgccgccaccatg. vpr primers: 5’ linker primer, complementary lo the USP and the vpr open reading frame 5’ end: ggggcttgttccatggtggc. 3’ cloning primer, complementary to the 3’ end of the vpropen reading frame plus the BamHl cloning site: ccgcggatcctaggatctactggc. Primers designed to amplify the remaining regulatory genes (vif, vpu, and ner) were constructed by the same design principle. H/W env-rev Plasmid The region encoding the two exons of rev and the vpuand env open reading frames of HIV1 HXBP was amplified via PCR and cloned into the expression vector pCDNAl/neo (purchased from Invitrogen). Expression of rev and env proteins was demonstrated by Western blot analysis and by the ability of cells transfected with this construct lo fuse with CD4+ cell lines (data not shown). tat HIV1 tat expression plasmid pCV1 was obtained through the AIDS Research and Reference Reagent Program from Dr. Flossie WongStaal (Arya et al., 1985). A region from the vector pBABE-hygro (Morgenstern and Land, 1990) expressing hygromycin resistance was subcloned into this plasmid to make pCVl-hygro. Alternatively, pCV1 was cotransfected with pBABE-puro at a ratio of 1OO:l. Identical results were obtained with both methods. Cloning Strategy for Deletion of the vpr Gene from the HIV Genome A region from just upstream of the unique PflMl site to just after the vif termination codon was amplified via PCR using primers that introduced a nonconservative amino acid change (Glu - Val) at amino acid 22 of vpr, a stop codon in the vpr reading frame immediately after amino acid 22, and an EcoRl site immediately following the new stop codon. This PCR fragment was substituted for the PflMI-EcoRI fragment of pNLpuro or pNL43. This substitution resulted in the deletion of 122 nt of the open reading frame of vpr, thus eliminating the possibility of reversion. The resulting plasmids, pNLpuroAvpr and pNLAvpr, encode the first 21 natural amino acids of vpr plus a valine plus all other remaining HIV1 genes and splice junctions in their native form. Determination of Regulatory Gene Expression by Reverse Tranacrlptlon PCR TE671 cells (0.5 x IO6 to 1.0 x 109 were transfected using DOTAP (Boehringer Mannheim) with expression vectors encoding individual regulatory genes. Forty-eight hours later cells were lysed in situ using RNAzol B (Biotecx Laboratories, Inc.), and total cellular RNA was prepared according to standard methodology (Chomczynski and Sacchi, 1967). cDNA was prepared by reverse transcription using random 6-mer primers and Moloney murine leukemia virus reverse transcriptase. An aliquot of cDNA was used as a template in PCR amplification. As a control for possible genomic DNA contamination, aliquots of RNA not subject to reverse transcription were used as templates in PCR. Analysis of HIV1 ~24~“’ Antigen Production For analysis of the infectivity of TE671 and TE671~, cells were grown to 80% confluence in tissue culture flasks, then incubated with filtered (0.2 pm pore size) supernatants from HUT-78 cells chronically infected with HIV1 (strain RF). One day later cells were washed once with phosphate-buffered saline containing trypsin (2.5 mglml). then twice with culture medium to remove residual virus used to infect the cells. Supernatants were collected at 24 hr intervals with the first collection occurring immediately after the wash step. Detection of ~24~ antigen was performed using an HIV1 p24 antigen assay kit (Coulter Immunology, Coulter Corporation) as per the manufacturer’s instructions. This method employs an antigen-capture enzyme-linked immunosorbent assay. Wells were analyzed for absorbance at 450 nm on a Dynatech MR5000 enzyme-linked immunosorbent assay reader. Transfections for Dlfferentlation Studies For differentiation experiments TE671 cells were transfected either by eledroporation or with the lipid-mediated method using DOTAP.

Though DOTAP produced more efficient transfections than electroporation, identical results were obtained with both methods with respect to differentiation. RD and D17 cells were transfected using DOTAP. In brief, electroporation was performed with a Bio-Rad GenePulser and Pulse Controller on 2 x lo6 to 5 x lo6 cells harvested in log phase growth. DOTAP transfection was performed as per the manufacturer’s instructions in tissue culture flasks on 0.5 x 1O’to 1 x 1 O6cells in log phase growth. In either case selection medium was added 48-60 hr after transfection, and cells were maintained in selection for the duration of the experiments. Cells transfected with plasmids containing the puromycin resistance gene (puromycin acetyltransferase) were selected in 1 pglml (D17) or 2.5 vglml (TE671 and RD) puromycin (Sigma). TE671 cells transfected with the tat-hygromycin expression vector were selected in 25 uglml hygromycin. Neomycin selection was in 1 mglml G418. Anti-Myosin Photomicrographic lmmunofluorescence Assay TE671 cells were transfected with the vpr expression vector, and 48 hr later the cells were trypsinized, transferred lo glass slides, and selected with puromycin for 5 days prior to staining. Rapidly proliferating untransfected TE671 cells were grown on glass slides for 2-4 days before staining. Permeabilization and fixation were performed with 100% methanol at -20°C for 10 min. The remaining steps were performed in PHEM buffer, which consists of 25 m M HEPES, 80 m M PIPES, 10 m M EGTA, and 2 m M MgClz (pH 6.9). The fixed cells were washed three times with PHEM in between each step. Cells were first blocked with 5% normal goat serum to reduce nonspecific staining. Murine monoclonal antibody MY-32, which is specific for the myosin heavy chain of fast-twitch (type II) skeletal muscle (Sigma number M-4276), was then incubated with the specimens. As a negative control, an isotype-matched antibody (SIM.4 anti-CD4, obtained through the AIDS Research and Reference Reagent Program from Dr. James Hildreth) was used as primary antibody on some cells. Rhodamineconjugated goat anti-mouse immunoglobulin G (TAGO) was used as secondary antibody, or alternatively a peroxidase-conjugated secondary antibody was used (Boehringer Mannheim). The cells were washed with PHEM, then examined under a fluorescence microscope and photographed. Infection Assay CD4’ TE671~ or CD4- TE671 cells were plated into tissue culture at very low confluence (<5%). One day later, supernatant and infected cells were added from HIV1 (strains RF or MN)-infected CD4+ HUT-78 T lymphoma cells. The following day the infected cells were washed from the culture, and fresh medium was added to the cells. Cells were examined for differentiation beginning on the second day after infection.

This work was supported by grants from the Council for Tobacco Research, the M. L. Smith Charitable Trust, and the NIH. The authors wish to thank the following persons for useful discussions and/or technical assistance: Yosef Refaeli, Brian MacDonald, Alice Sate, Nathan Levy. Jr., Billie Levy, Rebecca Shaina, Lindsey Anne, Elizabeth Hannah, Jennifer Punt, and Kristin Kelley. The authors also wish lo thank the reviewers for useful comments and suggestions. Received June 12. 1992; revised December

8, 1992

References Adachi, A., Gendelman, H. E., Koenig, S., Folks, T., Willey, FL, Rabson, A., and Martin, M. A. (1986). Production of acquired immunodeficiency syndrome-associated retrovirus in human and nonhuman cells transfected with an infectious molecular clone. J. Viral. 59, 284-291. Aguanno, S., Bouche, M., Adamo, S., and Molinaro, M. (1990). 12-Otetradecanoylphorbol-13-acetate-induced differentiation of a human rhabdomyosarcoma cell line. Cancer Res. 50, 3377-3382. Arya, S. K.. Guo, C., Josephs, S. F., and Wang-Staal, F. (1985). Transactivation gene of the human T-lymphotrophic virus type Ill (HTLV-III). Science 229, 69-73.

HIV1 Cell Differentiation 549

Bruni, C. (1979). Mitotic activity of muscle satellite cells during the early stages of rhabdomyosarcomas induction with nickel subsulfide. In Muscle Regeneration, A. Mauro, ed. (New York: Raven Press), pp. 265-273. Can% A. J., and Karn, J. (1989). Molecular biology of HIV: new insights into the virus life-cycle. AIDS (Suppl.) 3, 19-34. Chen. T. Ft., Dorotinsky, C., Macy, ht., and Hay, Ft. (1989). Cell identity resolved. Nature 340, 106. Chernoff, D. (1990). Myositis in HIV-infected patients. In The AIDS Knowledge Base, D. I. Abrams, P. T. Cohen, M. A. Sande, and P. A. Volberding, eds. (Waltham, Massachusetts: The Medical Publishing Group), pp. 5.14.1-5.14.2. Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156-159. Clapham, P. Ft., Weber, J. N.. Whitby, D., McIntosh, K., Gagleish, A. G., Maddon, P. J., Deen, K. C., Sweet, R. W.. and Weiss, R. A. (1969). Soluble CD4 blocks the infectivity of diverse strains of HIV and SIV for T-cells and monocytes but not for brain and muscle cells. Nature 337, 368-370. Clapham, P. R., Blanc, D., and Weiss, R. A. (1991). Specific cell surface requirements for the infection of CDCpositive cells by human immunodeficiency virus types 1 and 2 and by simian immunodeficiency virus. Virology 181, 703-715.

Harada, S., Koyanagi. Y.. Nakashima, H., Kobayashi, N., and Yamamoto, N. (1966). Tumor promoter, TPA, enhances replication of HTLVIIIIlAV. Virology 154, 249-256. Hattori, N., Michael% F., Fargnoli, K., Marcon, L., Gallo, R. C.. and Franchini, G. (1990). The human immunodeficiency virus type 2 vpr gene is essential for productive infection of human macrophages. Proc. Natl. Acad. Sci. USA 87, 6080-8064. Hiti, A. L., Bogenmann, E., Gonzales, F., and Jones, P. A. (1989). Expression of the MyoDl muscle determination gene defines differentiation capability but not tumorigenicity of human rhabdomyosarcomas. Mol. Cell. Biol. 9, 4722-4730. Kappes, J. C., Conway, J. A., Lee, S. W., Shaw, G. M., and Hahn, B. H. (1991). Human immunodeficiency virus type 2 vpx protein augments viral infectivity. Virology 784, 197-209. Kestler, H. W., Ill, Ringler, D. J., Mori, K., Panicali, D. L., Sehgal, P. K., Daniel, M. D., and Desrosiers, R. C. (1991). Importance of the nef gene for maintenance of high virus loads and for development of AIDS. Cell 65, 651-662. Kozak, M. (1986). Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44, 283-292. Lamperth. L., Illa, I., and Dalakas, M. (1990). In situ hybridization in muscle biopsies from patients with HIV-associated polymyositis (HIVPM) using labeled HIV RNA probes. Neurology (Suppl.) 40, 121.

Cohen, E. A., Dehni, G., Sodroski, J. G., and Haseltine, W. H. (1990a). Human immunodeficiency virus vpr product is a virion-associated regulatory protein. J. Virol. 64, 3097-3099.

Li, L., and Olson, E. N. (1992). Regulation of muscle cell growth and differentiation by the MyoD family of helix-loop-helix proteins. Adv. Cancer Res. 58, 95-119.

Cohen, E. A., Terwilliger, E. F., Jalinoos. Y., Proulx, J., Sodroski, J. G., and Haseltine, W. H. (1990b). Identification of HIV-1 vpr product and function. J. Acquir. Immune Defic. Syndr. 3, 11-16.

Ling, L. L., Keohavong. P., Dias, C., and Thilly, W. G. (1991). Optimization of the polymerase chain reaction with regard to fidelity: modified T7. Taq, and Vent DNA polymerases. PCR Meth. Appl. 7, 63-69.

Cohen, P. T., Sande, M. A., and Volberding, P. A. (199Oc). The AIDS Knowledge Base (Waltham, Massachusetts: The Medical Publishing Group). Dalakas, M., and Illa, I. (1991). HIV-associated myopathies. In Pediatric AIDS: The Challenge of HIV Infection in Infants, Children, and Adolescents, P. A. Pizza and C. M. Wilfert, eds. (Baltimore, Maryland: Williams and Wilkins), pp. 420-429.

Livingstone, L. R., White, A., Sprouse, J., Livanos, E.. Jacks, T., and Tlsty, T. D. (1992). Altered cell cycle arrest and gene amplification potential accompany loss of wild-type ~53. Cell 70, 923-935.

Dayton, A. I., Sodroski, J. G., Rosen, C. A., Goh, W. C., and Haseltine, W. A. (1986). The frans-activator gene of the human T cell lymphotrophic virus type Ill is required for replication. Cell 44, 941-947. Dedera, D., Hu, W., Vander Heyden, N., and Ratner, L. (1969). Viral protein R of human immunodeficiency virus types 1 and 2 is dispensable for replication and cytopathogenicity in lymphoid cells. J. Virol. 63, 3205-3206. Felix, C. A., Chavez Kappel, C., Mitsudomi, T., Nau. M. M., Tsokos, hf., Crouch, G. D., Nisen, P. D., Winick, N. J., and Helman, L. J. (1992). Frequency and diversity of ~53 mutations in childhood rhabdomyosarcoma. Cancer Res. 52, 2243-2247. Fisher, A. G.. Feinberg, M. B., Josephs, S. F., Harper, M. E., Marselle, L. M., Reyes, G., Gonda, M. A., Aldovini, A., Debouk, C., Gallo, R. C., and Wong-Staal, F. (1986). The transactivator gene of HTLV-III is essential for virus replication. Nature 320, 367-371. Fisher, A. G., Ensoli, B., Ivanoff, L., Chamberlain, M., Petteway, S., Ratner, L., Gallo, R. C., and Wong-Staal, F. (1967). The sor gene of HIV-l is required for efficient virus transmission in vitro. Science 237, 866493. Florini, J. R., Ewton, D. Z., and Magri, K. A. (1991). Hormones, growth factors, and myogenic differentiation. Annu. Rev. Physiol. 51, 201216. Grody, W. W., Cheng, L., and Lewis, W. (1990). Infection of the heart by human immunodeficiency virus. Am. J. Cardiol. 66, 203-206. Guyader, M., Emerman, M., Montagnier, L., and Peden, K. (1989). Vpx mutants of HIV-2 are infectious in established cell lines but display a severe defect in peripheral blood lymphocytes. EMBO J. 8. 11691175. Haggerty, S., Dempsey, M. P., Bukrinsky, M. I., Guo, L., and Stevenson, M. (1991). Posttranslational modifications within the HIV-1 envelope glycoprotein which restrict virus assembly and CD4-dependent infection. AIDS Res. Hum. Retroviruses 7, 501-510.

Marck, C. (1986). “DNA Strider”: a “c” program for the fast analysis of DNA and protein sequences on the Apple Macintosh family of computers. Nucl. Acids Res. 6, 1829-1836. McAllister, R. M.. Melnyk, J., Finklestein, J. Z., Adams, E. C., Jr., and Gardner, M. B. (1969). Cultivation invitrooncellsderivedfrom a human rhabdomyosarcoma. Cancer 24, 520-526. McAllister, R. M., Isaac% H., Rogney, R., Peer, M., Au, W., Soukup, S. W., and Gardner, M. B. (1977). Establishment of a human medulloblastoma cell line. Int. J. Cancer 20, 206212. Miller, C. W., Aslo, A., Tsay, C., Slamon, D., Ishizaki, K., Toguchida, J., Yamamuro, T.. Lampkin, B., and Koeffler, H. P. (1990). Frequency and structure of ~53 rearrangements in human osteosarcoma. Cancer Res. 50,7950-7954. Morgenstern, J. P., and Land, H. (1990). Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucl. Acids Res. 16, 3567-3596. Mulligan, L. M., Matlashewski, G. J.. Scrable, H. J.. and Cavenee, W. K. (1990). Mechanismsof ~53 loss in human sarcomas. Proc. Natl. Acad. Sci. USA 87,5863-5687. Ogawa, K., Shibata, R., Kiyomasu, T., Higuchi, I., Kishida. Y., Ishimoto, A., and Adachi, A. (1989). Mutational analysis of the human immunodeficiency virus vpr open reading frame. J. Virol. 63, 41104114. Roulston, A., D’Addario, M., Boulerice, F., Caplan, S., Wainberg, M. A., and Hiscott, J. (1992). Induction of monocyticdifferentiation and NF-r&-like activities by human immunodeficiency virus 1 infection of myelomonoblastic cells. J. Exp. Med. 775, 751-763. Schnittman, S. M., Denning, S. M., Greenhouse, J. J., Justement, J. S., Easeler, M., Kuttzberg, J., Haynes, B. F., and Fauci, A. S. (1990). Evidence for susceptibility of intrathymic T-cell precursors and their progeny carrying T-cell antigen receptor phenotypes TCRaf3+ and TCRTS+ to human immunodeficiency virus infection: a mechanism for CD4+(T4) lymphocyte depletion. Proc. Natl. Acad. Sci. USA 87,77277731. Schuitemaker,

H., Kootstra, N. A., Koppelman,

M. H. G. M., Bruisten,

Cdl 550

S. M., Huisman. H. G., Tersmette, M., and Miedema, F. (1992). Proliferation-dependent HIV-1 infection of monocytesoccurs during differentiation into macrophages. J. Clin. Invest. 89, 1154-I 160.

McKnight, A., Dalgleish, A. G., Sattentau, Q. J., and Weber. J. N. (1988). Human immunodeficiency viruses: neutralization and receptors. J. Acquir. lmmmune Defic. Syndr. 7, 536-541.

Shibata, R., Adachi, A., Sakai, H., Ishimoto, A., Miura, T., and Hayami, M. (1990a). Mutational analysis of Simian immunodeficiencyvirus type 2. J. Med. Primatol. 79, 217-225.

Werner, A., Winskowsky, G., Cichutek, K., Norley. S. G., and Kurth, R. (1990). Productive infection of both CD4’and CD4- human cell lines with HIV-l, HIV-2 and SIV,,,. AIDS 6, 537-544.

Shibata, Ft., Miura, T., Hayami. M., Ogawa, K., Sakai, H., Kiyomasu, T., Ishimoto, A., and Adachi, A. (1990b). Mutational analysis of the human immunodeficiency virus type 2 (HIV-2) genome in relation to HIV-1 and simian immunodeficiency virus SIV,I~~. J. Virol. 84, 742747.

Wong-Staal, F., Chanda, P. K., and Ghrayeb, J. (1987). Human immunodeficiency virus: the eighth gene. AIDS Res. Hum. Retroviruses 3. 33-39.

Siegel, H. N., and Lukas, R. J. (1988). Morphological and biochemical differentiation of the human medulloblastoma cell line TE671. Dev. Brain Res. 44, 269-280. Sodroski. J. G., Goh, W. C., Rosen, C., Tartar, A., Portetelle, D., Burney, A., and Haseltine, W. A. (1988). Replication and cytopathic potential of HTLV-IIILAV with sor gene deletions. Science 237, 1549-1553. Srinivasan, A., Goldsmith, C. S., York, D., Anand, R., Luciw, P., Schochetman, G., Palmer, E., and Bohan. C. (1988). Studies on human immunodeficiency virus-induced cytopathic effects: use of human rhabdomyosarcoma (RD) cells. Arch. Virol. 99, 21-30. Stanley, S. K., Kessler, S. W., Justement, J. S., Schnittman, S. M., Greenhouse, J. J., Brown, C. C., Musongela, L., Musey, K., Kabita, B., and Fauci, A. S. (1992). CD34’ bone marrow cells are infected with HIV in a subset of seropositive individuals. J. Immunol. 749, 689-697. Stratton, M. R., Darling, J., Pilkington, G. J., Lantos, P. L., Reeves, B. R., and Cooper, C. S. (1989). Characterization of the human cell line TE671. Carcinogenesis 10, 899-905. Strebel, K., Daugherty, D., Cohen, D., Folks, T., and Martin, M. (1987). The HIV “A” gene product is essential for virus infectibility. Nature 38, 728-730. Strebel, K., Klimkait. T., and Martin, M. A. (1988). A novel gene of HIV-l, vpu. and its 16-kilodalton product. Science 241, 1221-1223. Tapscott, S. J.. and Weintraub, H. (1991). MyoD and the regulation of myogenesis by helix-loop-helix proteins. J. Clin. Invest. 87, 11331138. Tateno, M., Gonzalez-Scarano, F., and Levy, J. A. (1989). Human immunodeficiency virus can infect CDCnegative human fibroblastoid cells. Proc. Natl. Acad. Sci. USA 86, 4287-4290. Terwilliger, E. F. (1992). The accessory gene functions of the primate immunity viruses. In AIDS Research Reviews, Volume 2, W. C. Koff, F. Wang-Staal, and R. C. Kennedy, ads. (New York: Marcel Dekker, Inc.). pp. 3-27. Terwilliger, E. F., Cohen, E. A., Lu, Y., Sodrosky, J. G., and Haseltine, W. A. (1989). Functional role of human immunodeficiency virus type 1 vpu. Proc. Natl. Acad. Sci. USA 86, 5163-5167. Tristem, M., Marshal, C., Karpas, A., Petrik, J., and Hill, F. (1990). Origin of vpx in lentiviruses. Nature 347, 341-342. Tristem, M., Marshal, C., Karpas, A., and Hill, F. (1992). Evolution of the primate lentiviruses: evidence from vpx and vpr. EMBO J. 17, 3405-3412. Vogel, J., Hinrichs, S. H., Reynolds, R. K., Luciw, P. A., and Jay, G. (1988). The HIV tat gene induces dermal lesions resembling Kaposi’s sarcoma in transgenic mice. Nature 335, 606-611. Weber, J., Clapham, P. R.. McKeating, J., Stratton, M., Robey, E., and Weiss, R. (1989). Infection of brain cells by diverse HIV isolates: role of CD4 as receptor. J. Gen. Virol. 70, 2853-2660. Weiner, D. B., Kokai, Y., Wada, T., Cohen, J. A., Williams, W. V., and Greene, M. I. (1989). Linkage of tyrosine kinase activity with transforming ability of the p185neu oncoprotein. Oncogene 4, 1175-l 163. Weiner, D. B., Huebner, K., Williams, W. V., and Greene, M. I. (1991). Human genes other than CD4facilitate HIV-1 infection of murine cells. Pathobiology 59, 361. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause, M., Benezra, R., Blackwell, T. K., Turner, D., Rupp, R., Hollenberg, S., Zhuang, Y., and Lassar, A. (1991). The myoD gene family: nodal point during Specification of the muscle Cell lineage, Science 251, 761-766. Weiss,

R. A., Clapham,

P. R., McClure,

M. O., McKeating,

J. A.,

Yin, Y., Tainsky, M. A., Bischoff. F. Z., Strong, L. C., and Wahl, G. M. (1992). Wild-type ~53 restores cell cycle control and inhibits gene amplification in cells with mutant p53 alleles. Cell 70, 937-948. Yu, X.-F., Matsuda, M., Essex, M., and Lee, T.-H. (1990). Open reading frame vpr of simian immunodeficiency virus encodes a virionassociated protein. J. Virol. 64, 5688-5693. Yuan. X., Matsuda, Z., Matsuda, M., Essex, M., and Lee, T.-H. (1990). Human immunodeficiency virus vpr gene encodes a virion-associated protein. AIDS Res. Hum. Retroviruses 6, 1265-1271.