Rapid Complementation Assays Measuring ... - Journal of Virology

3 downloads 0 Views 1MB Size Report
Feb 2, 1990 - AND JOSEPH SODROSKIl*. Division ofHuman ..... Fisher, A. G., B. Ensoli, L. Ivanoff, M. Chamberlain, S. Pette- way, L. Ratner, R. Gallo, and F.
Vol. 64, No. 5

JOURNAL OF VIROLOGY, May 1990, p. 2416-2420 0022-538X/90/052416-05$02.00/0 Copyright C) 1990, American Society for Microbiology

Rapid Complementation Assays Measuring Replicative Potential of Human Immunodeficiency Virus Type 1 Envelope Glycoprotein Mutants EIRIK HELSETH,1 MARK KOWALSKI,' DANA GABUZDA,1 UDY OLSHEVSKY,1 WILLIAM HASELTINE,2 AND JOSEPH SODROSKIl* Division of Human Retrovirology, Dana-Farber Cancer Institute, Department of Pathology, Harvard Medical School,' and Department of Cancer Biology, Harvard School of Public Health,2 Boston, Massachusetts 02115 Received 24 October 1989/Accepted 2 February 1990

Rapid assays which measure the ability of mutant human immunodeficiency virus type 1 envelope glycoproteins to mediate cell-free and/or cell-to-cell transmission of virus are described. By using these assays, envelope glycoprotein mutants with varying degrees of syncytium-forming ability were tested for ability to complement viral replication in trans. As expected, mutants that dramatically affect association of the gpl20-gp4l envelope subunits, CD4 binding, or membrane fusion were unable to form syncytia or to support cell-free or cell-to-cell transmission. Surprisingly, some membrane fusion-defective mutants significantly attenuated in syncytium-forming ability were able to complement viral replication. Conversely, mutations in the carboxyl terminus of gp4l transmembrane glycoprotein, although not affecting syncytium-forming ability, significantly attenuated both forms of virus transmission. These results indicate that syncytium formation is not sufficient for cell-to-cell transmission of human immunodeficiency virus type 1. Furthermore, virus transmission appears to be less sensitive to inhibition of membrane fusion than is syncytium formation. Human immunodeficiency virus type 1 (HIV-1) is the etiologic agent of acquired immunodeficiency syndrome (AIDS) (1, 6, 16). AIDS is characterized by a progressive depletion of CD4-positive lymphocytes (7, 12). HIV-1 exhibits a tropism for CD4-positive cells in tissue culture (8). The basis for this tropism is a specific interaction between CD4, the viral receptor, and the gp120 exterior envelope glycoprotein (4, 9, 15). Following receptor binding, the HIV-1 envelope glycoproteins, gp120 and gp4l, mediate a fusion event that allows entry of the viral core and nucleic acids into the target cell (22). In addition to cell-free virion-mediated transmission, HIV-1 can spread in a poorly understood cell-to-cell manner (5, 23). The HIV-1 envelope glycoproteins are expressed in the infected cell and mediate at least some of the cytopathic effects that accompany viral infection. The formation of multinucleated giant cells or syncytia is mediated by the HIV-1 envelope glycoproteins (14, 21). Syncytium formation requires proteolytic processing of the gp160 envelope precursor, cell surface expression, association of the gp120 and gp4l subunits, and CD4 binding (10). Processes following receptor binding, referred to herein as membrane fusion events, are necessary for syncytium formation and are affected by changes in the gp4l amino terminus (10). While the effects of mutations in the HIV-1 env gene on syncytium formation are readily assessed, analysis of the effects of such mutations on virus transmission is much more laborious. Typically, the mutation is introduced into an infectious proviral clone and virus replication is assayed for several days or weeks following transfection of the provirus into susceptible target cells. The true relative effect of the mutation on replication is often difficult to assess in this system, since at the time of assay, the wild-type and mutant viruses have usually undergone several rounds of replication. In addition, effects of the mutation on cis-acting se*

quences, which are scattered throughout the HIV-1 genome, are difficult to distinguish from effects on trans-acting envelope glycoproteins. Finally, the effects of mutations on the replicative function of the HIV-1 gp4l carboxyl terminus can

be difficult to study, since the mutations often disrupt the overlapping tat- and rev-coding sequences. Here we describe an assay for the single-step replication of a defective HIV-1 genome dependent on trans-complementation by envelope glycoproteins. The assay is simple and quantitative and overcomes the above-mentioned problems. By using this trans-complementation assay, we provide evidence that syncytium formation and virus transmission are related, but dissociable, events. The trans-complementation assay uses two plasmids, one encoding rev and either wild-type or mutant envelope glycoproteins (pSVIIIenv), the other containing a HIV-1 provirus with a deletion in the env gene (pHXBAenvCAT). The pSVIIIenv plasmid expresses the HIV-1 rev and env genes under the control of the HIV-1 long terminal repeat, which is responsive to the tat transactivator protein (21). The pHXBAenvCAT plasmid contains an HIV-1 provirus with a 580-base-pair deletion in the env gene and a chloramphenicol acetyltransferase (CAT) gene replacing the nef gene (24). Both plasmids contain simian virus 40 origins so that they replicate in COS-1 cells, which express the simian virus 40 T antigen. To examine the ability of the envelope glycoproteins to complement cell-free viral transmission, 5 ,ug each of the pSVIIIenv and pHXBAenvCAT plasmids were cotransfected into COS-1 cells by the DEAE-dextran technique (3) (Fig. 1). At 48 to 72 h after transfection, virions are produced in the transfected COS-1 cell supernatants. These virions contain envelope glycoproteins derived from the envelopeexpressing plasmid and an RNA genome derived from the pHXBAenvCAT plasmid. The COS-1 cell supernatants were filtered (0.45 ,um), 1/10 volume (1 ml) was centrifuged at 13,000 x g, and the reverse transcriptase was measured as

Corresponding author. 2416

VOL. 64, 1990

NOTES

pHXBAGnvCAT pHXB,&envCAT ,DSk Irl-V

35.3

5

2'3

so

8.5

4.8

27

._

rev

pSVIIIenv

FT

Ril I

LTR

||i

1gp20 45

p41

314 363 B9 B9

17

0.29 filter

,0 8 days

t JURKAT

;r

CELLS

CAT assay

0

°

0 0 2 days COS-1

CELLS

18.0

2.2

0.3

0.3

5.1

0.4

._*..

FIG. 2. CAT activity observed following transfer of COS-1 cellproduced virions onto Jurkat cultures. Supernatants from COS-1 cells that were cotransfected with the pHXBA&envCAT plasmid and with either no DNA, pSVIIIenv, or pSVIIIenv containing mutations in the env gene were incubated with Jurkat lymphocytes, and CAT activity was measured. In some experiments, Jurkat lymphocytes were preincubated with azidothymidine (AZT) (Burroughs-Wellcome Co.) at the concentrations shown for 24 h prior to infection. In other experiments, Jurkat cells were treated with 0.25 ,ug of either OKT4 or OKT4a monoclonal antibody (Ortho Diagnostics, Inc.) per ml prior to infection.

l 753

0 e

OKT4 OnKT-A

$100 0 0 0 0 * 0 0 0

_

env

363 517A 51 7B

A7. 2.1 -------

-

X3

5-

2417

JUFIRKAT C

2 days

CAT assay

FIG. 1. Complementation assays for HIV-1 envelope glycoprotein. The pHXBA&envCAT and pSVIIIenv plasmids were cotransfected into either COS-1 cells or Jurkat lymphocytes as described in the text. The location of the mutations in the env gene are shown beneath the pSVIIIenv plasmid, along with the BglII sites marking the deletion endpoints in the pHXBAenvCAT plasmid.

previously described (19). Equivalent reverse transcriptase units (usually 1 x 105 to 2 x 105 cpm) of supernatants derived from COS-1 cells transfected with the wild-type or mutant envelope expressor plasmids were added to Jurkat T lymphocytes (-2 x 106 cells in 10 ml of fresh medium). The Jurkat cells were incubated for 48 to 72 h and then pelleted, lysed, and assayed for CAT activity as previously described (20). Figure 2 shows the typical CAT activity observed in the target Jurkat cells. Approximately 100-fold increases in Jurkat cell CAT activity were observed when the wild-type HIV-1 envelope-expressing plasmid, as compared with a plasmid expressing rev only or with no DNA, was cotransfected into the COS-1 cells with the pHXBAenvCAT plasmid. This observed CAT activity was reduced more than 10-fold by pretreatment of the Jurkat cells with azidothymidine or with OKT4a, but not OKT4, monoclonal antibodies. These results indicate that the transferred CAT activity depended upon exposure of the CD4 epitope to which the HIV-1 gpl20 envelope glycoprotein binds (15) and upon reverse transcription. The results are consistent with the occurrence of a round of Jurkat cell infection, allowing the production of a HXBAenvCAT provirus, which can then be integrated and expressed. Since the pSVIIIenv sequences lack signals important for HIV-1 packaging into virions (13), functional envelope glycoproteins are not expressed in the target Jurkat cells. Thus, only a single round of cell-free infection is likely to occur in this system. In addition to cell-free infection, HIV-1 can spread in a cell-to-cell manner, in which virus is transmitted by contact of infected cells with uninfected cells (5, 23). To assay the ability of mutant envelope glycoproteins to support virus

replication in a context where both cell-free and cell-to-cell transmission might occur, a variation of the cell-free assay described above was used (Fig. 1). In this assay, Jurkat T lymphocytes are cotransfected by the DEAE-dextran procedure (17) with 10 ,ug each of the pSVIIIenv and pHXBAenvCAT plasmids. Transient expression of CAT, which decreases by days 4 to 5 following transfection, can be detected due to introduction of the pHXBAenvCAT plasmid into the transfected cells. The percentage of cells that express the transfected provirus is less than 5% (21). If an envelope glycoprotein competent for either cell-free or cellto-cell transmission is coexpressed in these cells, the HXBAenvCAT provirus can be transmitted and stably integrated into cells surrounding the successfully transfected cells. CAT expression will be relatively stable in such a culture if there is no selective disadvantage associated with expression of the HXBAenvCAT provirus. Figure 3A shows the level of CAT activity observed in Jurkat cultures transfected either with the pHXBAenvCAT plasmid alone or the pHXBAenvCAT plasmid plus the pSVIIIenv plasmid expressing the wild-type HIV-1 envelope glycoproteins. The CAT activity following transfection of the pHXBAenvCAT plasmid alone is 10- to 20-fold lower and only transiently expressed compared with the high level and stable expression of CAT seen in the parallel culture transfected with the pHXBAenvCAT and pSVIIIenv plasmids. The expression of CAT in the latter cultures remained at approximately the same level for at least 11 days following the transfection. The stable, high-level CAT expression was markedly reduced by treatment of the transfected Jurkat cultures with azidothymidine (Fig. 3B). Thus, measurement of CAT activity in the transfected Jurkat lymphocyte culture at day 8 following transfection indicates the amount of stable CAT activity that resulted from either cell-free or cell-to-cell transmission of the HXBAenvCAT provirus. The effect of mutations in the env gene on the transcomplementation assays was assessed. Mutations affecting a variety of HIV-1 envelope glycoprotein functions were selected. The number of the mutation corresponds to the amino acid residue of the HIV-1 HXBc2 envelope glycoprotein (18), modified by either site-directed mutagenesis (11)

2418

NOTES

100 -0-

B

,c0 c

0)

80 F

-----

AZT(IIM)

pHXBAenvCAT + pSVillenv pHXB AenvCAT

2.5

acetylation

E

5

20

80

60.3 3.1

2.1

1.2

ES

o

o-

pSVHllenv

Mock

Plasmid

60 _

0

-6 40 _ c

0) .-

Co

A

20 _ B

1

2

3

4

5

6

7

8

9

10

11

Days

kS

J. VIROL. 517A 517B 363

6.4

1.1

1.0

45

314

1.8

4.9

FIG. 3. CAT activities following direct transfection of Jurkat lymphocytes. (A) The percentage acetylation of ['4C]chloramphenicol for samples containing equal numbers of Jurkat cells transfected on day 0 with either pHXBAenvCAT plus pSVIIIenv (-4--) or pHXBAenvCAT alone (-). (B) CAT assays performed on day 8 after transfection of Jurkat lymphocytes with pHXBAenvCAT and either no DNA, pSVIIIenv, or pSVIIIenv containing env gene mutations. In some cases, azidothymidine (AZT) was added at the shown concentrations following transfection and kept at that concentration until the time of assay for CAT activity. -U-

(for mutants 45 W/S and 314 G/W), linker insertion (for mutants 363, 517A, and 517B), or deletion (for the A753 mutant). The 45 W/S mutant contains a single amino acid change, converting the conserved tryptophan at position 45 (in the first conserved region of gp120) to a serine. The 314 G/W mutant contains a substitution of a tryptophan residue for a conserved glycine located within the gp120 variable region 3, the major target for type-specific neutralizing antibodies. The amino acid sequence of the mutants resulting from linker-insertion mutagenesis has been previously published (10). The 363 mutant contains a change in the third conserved region of gp120, while the 517A and 517B mutations affect the amino terminus of gp4l. The A753 mutant contains a deletion of 103 carboxy-terminal gp4l amino acids, with a substitution of glutamine and threonine residues at the terminus of the mutant protein (10). The structure and CD4-binding ability of the mutant envelope glycoproteins were analyzed by labeling COS-1 cells transfected with the envelope-expressing plasmids with 100 puCi of [35S]cysteine per ml for 8 h at 48 h posttransfection. Labeled cell lysates and supernatants were immunoprecipitated by using an AIDS patient serum, 19501, and were analyzed on sodium dodecyl sulfate-polyacrylamide gels as described previously (10). For measurement of CD4binding ability, labeled gpl20 in the transfected cell supernatants was incubated with SupTl CD4-positive lymphocytes and the bound gpl20 was immunoprecipitated as previously described (10). Syncytium-forming ability of the mutant glycoproteins was assessed by transfection of the envelope-expressing plasmids into Jurkat tat cells and scoring for syncytia, as previously described (21). A structural analysis of the mutants is provided in Fig. 4, and the syncytium-forming ability of the mutants is shown in Table 1. The 45 W/S mutant exhibits a dissociation of the gpl20 exterior envelope glycoprotein from the gp4l transmembrane glycoprotein, as evidenced by abundant gpl20 glycoprotein in transfected cell supernatants, but little gpl20 in the cell lysates. Note that gp4l steady-state levels are

reduced, suggesting that the gp4l glycoprotein is rapidly degraded following gpl20 dissociation. The 363 mutant is completely attenuated in the ability to bind to CD4 (10). The 517A and 517B mutations affect a postreceptor binding step in the membrane fusion process, with the 517B mutation being more disruptive of syncytium-forming ability than the 517A mutation (10). The 314 G/W mutation is located in the major neutralization loop of the gpl20 exterior glycoprotein and also disrupts syncytium formation without affecting CD4 binding. The A753 mutation truncates the long intracytoplasmic tail of the gp4l transmembrane glycoprotein; this muta-

B

A ,

1

2

3

6

5

4

1

7

kD

160-

w

120-

C

1

iR

41

.*ff *:

* Om.": :

s

2 3 4

5

7

6

-120

2 3 4

5 6 7 ....

.I......

-20 r..1.1.1..

FIG. 4. Expression and CD4 binding of envelope glycoprotein mutants expressed in COS-1 cells. Immunoprecipitation of COS-1

cell lysates (A) or supernatants (B) with 19501 AIDS patient serum following transfection with no DNA (lane 1), pSVIIIenv (lane 2), pSVIIIenv-45W/S (lane 3), pSVIIIenv-314G/W (lane 4), pSVIIIenv363 (lane 5), pSVIIIenv-517A (lane 6), or pSVIIIenv-517B (lane 7) is shown. Panel C shows the amount of gpl20 bound to cell surface CD4 following incubation of the transfected COS-1 cell supernatants with SupTl CD4-positive lymphocytes.

NOTES

VOL. 64, 1990

TABLE 1. Relative replicative and syncytium-forming abilities of envelope mutants Mutant envelope glycoprotein

Wild type

45W/S 314G/W 363 517A 517B A753

Replication

(o% of wild type)a COS-Jurkat

Jurkat

100 0 11.9 0 13.7 0.3 7.5

100 0 4.2 0 6.7 0 4.7

Syncytium formationb (% of wild type)

100