Human cytomegalovirus product UL44 downregulates the transactivation of HIV-1 long terminal repeat

Abstract

Objective: 

Human cytomegalovirus (HCMV) is often isolated from HIV-1-infected patients and the two viruses can infect the same cell type giving rise to direct bidirectional interactions. Whereas the long terminal repeat (LTR) transactivation ability of HCMV immediate early gene (IE1/IE2) is well documented, no information is available on the possible role of other HCMV proteins. In this study, the activity of ppUL44, an early DNA-binding protein, on HIV LTR transactivation was investigated.

Methods: 

HIV LTR transactivation by ppUL44 in presence or absence of HIV-1 Tat and HCMV IE1/IE2 was determined in J-Jhan and U973 cells through transient transfection experiments with a series of different expression vectors. Some experiments were also performed on U373-MG astrocytoma cells permanently transfected with UL44 or with another HCMV gene used as a control (UL55).

Results: 

The basal transactivation activity of the HIV LTR was not influenced by the presence of ppUL44. On the contrary, the transactivation observed in the presence of Tat, IE1/IE2 or both factors in synergy was strongly downregulated by ppUL44 in a dose-dependent manner. Deletion constructs of ppUL44 demonstrated that the region of the molecule responsible for the inhibition of the LTR is located within the last 114 amino acids at the carboxyl-terminal region.

Conclusions: 

The results obtained indicate that within the last 114 amino acids of ppUL44 there is a domain that has a negative effect on the ability of HIV-1 LTR to be activated by both its autologous transactivator Tat and the heterologous transactivator HCMV IE1/IE2 functioning individually or synergistically.

Introduction

During the progression of AIDS the diversity of responses to HIV suggests that a number of cofactors contribute to the pathogenesis of the disease. This has implicated differentiating agents, cytokines and mitogens, and a number of DNA viruses. Amongst the DNA viruses, the Herpesviridae family in general and human cytomegalovirus (HCMV) in particular have often been cited as cofactors because they are frequently isolated from HIV-1-infected patients and cause increased morbidity and mortality in this population [^1,2]. Furthermore, the finding that a number of sites in vivo are infected with both HIV-1 and HCMV, and that single cells in the brain, retina and lung may be infected with both viruses [^3–5] has suggested that both direct and indirect interactions between these two viruses could influence their replication and the resulting disease pathogenesis. HCMV encodes regulatory proteins that can transactivate gene expression from HIV long terminal repeat (LTR) [⁶] and may play a role in the activation of latent HIV and in supporting HIV replication [⁷]. Several studies that examined the role of the major HCMV immediate early (IE) gene products in the transactivation of the HIV LTR with expression vectors based on genomic fragments from IE regions 1 and 2 have shown that both IE genes (mainly IE2) code for products that can transactivate LTR [⁸]. Therefore, the incomplete expression of the HCMV genome seems to trigger the activation of HIV LTR because of the transactivation activity of IE1/IE2.

However, other studies have found that the effect of HCMV on HIV-1 replication depends on the permissiveness of the cells to each virus [⁹]. Furthermore, there are data in the literature indicating that in cells completely permissive for HCMV and HIV replication, co-infection results in the inhibition of HIV-1 replication [¹⁰], suggesting the existence of HCMV early/late gene product(s) with a negative effect on HIV.

We recently focused our attention on the open reading frame UL44, which encodes a family of proteins grouped under the name of ICP36. The major UL44 product, ppUL44, is an abundant, phosphorylated DNA-binding protein of about 52 kDa, which is one of the most immunogenic proteins during natural infection [¹¹] and has been shown to be essential for HCMV replication in vitro [¹²]. This protein is expressed initially in the early phase of infection and accumulates in the nuclei of infected cells until and throughout the late phases of the viral replication cycle. ppUL44 has sequence homology with human herpesvirus 6 (HHV-6) P41, which also shares significant similarities with ppUL44 as it has nuclear localization, is expressed early after infection, shows DNA-binding abilities, and binds to viral DNA polymerase [¹³]. An HHV-6 cDNA clone, pcD41, encoding the HHV-6 P41 (U27) has been shown to transactivate HIV LTR via the nuclear factor NF-κB binding sites [¹⁴].

This report examines the effect of HCMV ppUL44 on the transactivation of HIV LTR by HIV-1 Tat and HCMV IE proteins in different cultured mammalian cells.

Materials and methods

Cell cultures and treatments

U373-MG astrocytoma cells (kindly provided by S. Michelson, Paris, France), were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco, Paisley, Scotland) supplemented with 5% fetal calf serum (FCS). Non-adherent human J-Jhan and U937 (kindly provided by M.C. Re, Bologna, Italy) CD4+ cell lines were grown in RPMI-1640 medium (Mascia Brunelli, Milan, Italy) supplemented with 10% FCS.

Reporter vectors

The HIV LTR construct was based on the original LAV-1_Bru LTR chloramphenicol acetyltransferase (CAT) construct [¹⁴] (gift from A. Rabson, National Institutes of Health, Bethesda, Maryland, USA). The LTR-luc plasmid carried the luciferase (luc) reporter gene under the control of HIV (LAV-1_Bru) LTR (fragment BglII −485 to HindIII +78).

Expression vectors

The Tat expression vector (pLTRtat) contained cDNA encoding the HIV transactivator Tat under the control of HIV-1 LTR and was a generous gift from S. Michelson [¹⁵]. The Tat expression vector (pRNeocTAT) contained cDNA encoding Tat under the control of simian virus 40 (SV40) early region promoter and was kindly provided by A. Caputo (Ferrara, Italy) [¹⁶]. Plasmids expressing HCMV IE proteins were a gift from R.L. LaFemina (West Point, Pennsylvania, USA), G. Hayward (Baltimore, Maryland, USA), and C. Davrinche (Toulouse, France) and have been described in detail elsewhere [¹⁶]. These plasmids include the complete HCMV IE transcriptional control region and encode the major IE1 and IE2 proteins (pRL45, exons 1–5) [¹⁷].

Construction of pUL44 expression plasmids

The eukaryotic expression vector pCDNA3 (Invitrogen Corporation, San Diego, California, USA), a plasmid expressing a selection marker for resistance to neomycin, was linearized by digestion with HindIII and filled in, then ligated to a DNA fragment carrying nucleotides 1–1302 of UL44 coding for the whole protein, plus 10 base-pairs of the 3′-untranslated sequence obtained by polymerase chain reaction amplification performed on the viral genome (Towne strain) with the following primers: up44, 5′-CTTCGCTCGAGGG ATGGATCGCAAGACGCGCCTCTC-3′; down44, 5′-CTAAGGATCCTAGCCGCACTTTTGCTTG GTG-3′. The resulting construct, p433, placed UL44 sequences under the control of the major IE HCMV enhancer/promoter region, followed by the polyA region of bovine growth hormone gene BGH (Fig. 1), as confirmed by conventional dideoxy-sequencing. The eukaryotic expression vector pRc/RSV (Invitrogen Corporation), was linearized by digestion with NotI and filled in, then ligated to a DNA fragment carrying nucleotides 1–1302 of UL44 (coding for the whole protein) obtained from the expression vector pHIL- IN5 [¹⁸], after digestion with AgeI and BstbI and filling in. The resulting construct, RSV-44, placed the UL44 sequence under the control of the enhancer/promoter sequences from the Rous sarcoma virus (RSV) LTR.

Fig. 1: . Construct p433 expressing the whole ppUL44 and its deletion constructs. CMV-MIEP, major immediate early enhancer/promoter of human cytomegalovirus (HCMV); BGHpA, bovine hormone gene polyA; RSV-LTR, Rous sarcoma virus long terminal repeat.

Transfection conditions for stable expression

U373-MG cells were permanently transfected with p433 DNA, and 3 × 10⁶ cells in log-phase growth were resuspended in 800 µl DMEM containing 10 µg ppUL44 plasmid DNA. The mixture was transferred to a 0.4 cm electroporation cuvette and pulsed once using a Gene Pulser set (BioRad, Richmond, California, USA) at 960 mF and 300 V. Transformants were selected in medium containing the neomycin analogue G418 (400 mg/l), starting 1 week after electroporation. The stably transfected population was called UL44/U373.

A cell population harbouring the vector plasmid alone (pCDNA3) was obtained (pCDNA3/U373) in parallel and used as a control in all experiments. A cell line expressing HCMV gB (gB/U373; kindly provided by M. Reschke, Marburg, Germany) was used in some experiments as a further control cell line [¹⁹].

Construction of deletion mutants

For generation of amino-terminal deletions in ppUL44, the p433 plasmid was cut with BglI and XhoI to excise a 1054 base-pair fragment which was then cloned in the BamHI-XhoI site of pCDNA3. The resulting construct (p318) expressed the UL44 amino-acid sequence spanning residues 116–433. The p114 construct was obtained by cloning 463 nucleotides 3′ of the UL44 coding sequence. For this purpose a SmaI-XbaI fragment excised from p433 was cloned into pCDNA3 after linearization by double digestion of the vector with EcoRV and XbaI: the resulting construct expresses the last 114 amino acids of ppUL44. For generation of the carboxyl-terminal deletion mutant, p433 was digested with SacI and EcoRV resulting in excision of a fragment of 278 nucleotides and religated. The construct obtained, p341, expressed the first 341 amino acids (Fig. 1).

Transfection procedures

Transient transfection experiments performed on J-Jhan and U937 non-adherent human cell lines were carried out by the diethylaminoethyl–dextran procedure in triplicate using a microtransfection method in 96-well plates, as described by Schwartz et al. [¹⁵]. Briefly, cells were collected by centrifugation and resuspended in transfection buffer (137 mmol/l NaCl, 5 mmol/l KCl, 0.7 mmol/l CaCl₂, 0.5 mmol/l MgCl₂, 0.6 mmol/l Na₂HPO₄, 25 mmol/l Tris-HCl, pH 7.4) containing 200 ng DNA per well and diethylaminoethyl–dextran at a final concentration of 500 mg/ml. Cells were seeded at a density of 2 × 10⁵ cells per well, incubated at room temperature for 20 min and washed with RPMI-1640 medium supplemented with 5% FCS. Cells were incubated at 37°C for 48 h. Transient gene expression in adherent cells (UL44/U373, pCDNA/U373 and gB/U373 cells) was carried out by a variation of the calcium phosphate coprecipitation technique with 1 µg total DNA per 10⁶ cells. Cells were seeded in 24-well plates at a density of 8 × 10⁴ per well, in 1 ml culture medium. One day later, the cells were cotransfected with a mixture of plasmids in transfection buffer for adherent cells (140 mmol/l NaCl, 5 mmol/l KCl, 190 mmol/l glucose, 0.6 mmol/l Na₂HPO₄˙12H₂O, 20 mmol/l Tris HCl pH 7.05) containing 10% of a 0.125 molar CaCl₂ stock solution. After 40 min incubation at room temperature, 110 µl of this mixture were added per well. Each experimental point was carried out in parallel in six wells. Immunofluorescence analysis with specific antibody to Tat, IE1/IE2 and UL44 was used to monitor the level of expression of the transfected plasmid DNA.

At 48 h after transfection, cells were washed with phosphate-buffered saline (PBS) once and then were lysed with 100 ml of lysis buffer (25 mmol/l Tris phosphate, pH 7.8, 8 mmol/l MgCl₂, 1 mmol/l dithiolthreitol, 15% glycerol, 1% Triton X-100, 1% bovine serum albumin), for 40 min at 4°C in agitation.

Luciferase activity assay

Luciferase activity was measured using a luminometer (Digene Diagnostic, Inc., Silver Spring, Maryland, USA), as described elsewhere [¹⁵], and was expressed as relative light units per mg protein (the latter being determined by the BioRad method).

Immunofluorescence

Cells were fixed at −20°C in methanol–acetone (3 : 1) for 30 min, then incubated at 37°C in a humid chamber with the primary antibody for 1 h. Following three 5 min washes in PBS, fluoresceinated secondary antibody was added to the cells and incubation was carried out for 1 h at 37°C in a humid chamber. Three 5 min washes in PBS were subsequently performed. Cells were then counterstained with Evans' blue, washed once briefly in PBS and observed under an ultraviolet microscope. Monoclonal antibody CH16, specific for ppUL44 (Goodwin Institute, Plantation, Florida, USA) was used at a 1 : 200 dilution. Both monoclonal antibody E13, directed against the product of CMV UL123 (Argene-Biosoft, Varilhes, France), and the antibody specific for HIV-1_IIIB Tat (Intracel, London, UK) were used at a 1 : 20 dilution.

Results

ppUL44 downregulates Tat-mediated transactivation of HIV-LTR in transient transfection assays

When J-Jhan cells were cotransfected with LTR-luc and constructs expressing ppUL44 under the control of the major immediate early enhancer/promoter (MIEP or p433) or the RSV LTR (RSV-44; Fig. 1), pUL44 expression did not have any effect on the basal level of transactivation of HIV LTR (Fig. 2). To address the question of whether ppUL44 could positively or negatively interfere with the transactivation of HIV LTR by its natural transactivator Tat, we cotransfected J-Jhan cells with HIV tat expression vector (CMV-tat, LTR- tat and pRPneo-cTAT) together with LTR-luc in the presence or absence of p433 or RSV-44. In those experiments in which ppUL44-expressing construct was not used, the control plasmid expressing CMVCAT was added.

Fig. 2: . Transient transfection of J-Jhan cells with HIV long terminal repeat (LTR)-luc together with cytomegalovirus (CMV)-tat, LTR-tat and SV40-tat (pRPneo-cTAT) in the presence or absence of ppUL44 under the control of the major immediate early enhancer/promoter (MIEP or p433) or the Rous sarcoma virus (RSV) LTR (RSV-44). RLU, Relative light units; LUC, luciferase; prot, protein.

As expected, strong LTR transactivation was obtained in the presence of Tat (30-fold more luciferase activity than in the absence of Tat; Fig. 2) coded by CMV-tat or LTR-tat. Tat-induced LTR transactivation was sixfold inhibited when transfections were carried out in the presence of ppUL44-expressing plasmids p433 or RSV-44. As expected a lower level of LTR transactivation was obtained in the presence of Tat coded by pRPneocTAT (15-fold more luciferase activity than in the absence of Tat; Fig. 2). In this case, Tat-induced LTR transactivation was reduced to a basal level in the presence of ppUL44-expressing plasmids p433 or RSV-44.

A series of similar experiments were carried out on U937 cells and the same results were obtained (data not shown).

Dose-dependent effect of ppUL44 on Tat-mediated transactivation of HIV LTR

To establish whether the negative effect of ppUL44 (both under the control of the MIEP and under the control of RSV LTR) on Tat-mediated transactivation of LTR was dose-dependent, increasing amounts of p433 (and RSV-44) were cotransfected in the presence of a constant amount of CMV-tat or LTR-tat. One hundred per cent inhibition of Tat-mediated LTR transactivation was reached by increasing the amounts of p433 four times (Fig. 3).

Fig. 3: . Transient transfection of J-Jhan cells with HIV long terminal repeat (LTR) together with HIV-tat in the presence or absence of ppUL44. Numbers at the bottom of the bars represent the constant amounts of HIV-tat with respect to increasing amounts of ppUL44. The data are the arithmetic means of six independent experiments, each run in triplicate. RLU, Relative light units; LUC, luciferase; prot, protein.

ppUL44 downregulates Tat-mediated transactivation of HIV-LTR in permanent transfection assays

U373-MG cells, an astrocytoma cell line permissive for HCMV replication, was permanently transfected with p433. Several cell clones growing in the selection medium (containing G418) were pooled as one population (UL44/U373) and checked by immunofluorescence with monoclonal antibody CH16 to determine the percentage of cells expressing ppUL44. Approximately 75% of the cells had ppUL44 in the nuclei (Fig. 4). Similarly, a control cell population (pCDNA3/U373) was generated by permanently transfecting U373-MG with pCDNA3 vector DNA.

Fig. 4: . UL44/U373 cells fixed and stained with fluorescein-labelled monoclonal antibody CH16 which recognizes ppUL44.

The two cell populations (UL44/U373 and pCDNA3/U373) were then transiently transfected with LTR-luc together with the constructs expressing HIV-tat. A much stronger LTR transactivation (5.6-fold) was observed when Tat was expressed in pCDNA3/U373 compared with the activation observed in UL44/U373 (Fig. 5).

Fig. 5: . Transcriptional activity of HIV long terminal repeat (LTR)-luc induced by HIV-tat in pRc/U373, UL44/U373 and gB/U373 cells. The data are the arithmetic means of six independent experiments, each run in triplicate. RLU, Relative light units; LUC, luciferase; prot, protein.

As a further control we also transfected a cell line which expressed another CMV gene (UL55 coding for gB) under the control of the MIEP. The level of Tat- mediated transactivation obtained in gB-expressing cells was the same as that observed in pCDNA/U373 (Fig. 5). ppUL44 inhibited HCMV IE1/IE2-mediated HIV LTR transactivation.

Because it has repeatedly been shown that HCMV IE1/IE2 transactivate HIV LTR, we investigated the effect of ppUL44 on IE1/IE2-mediated LTR transactivation cotransfecting J-Jhan cells with LTR-luc in presence or absence of pRL45 and the construct coding for ppUL44 (either p433 or RSV-44).

The results obtained are shown in Fig. 6 and represent the mean of six different experiments each run in triplicate. As expected, IE1/IE2 gave a 20-fold transactivation of LTR. This IE1/IE2-mediated LTR transactivation was strongly (fivefold) inhibited by the coexpression of ppUL44. The same results were obtained when transfection experiments were carried out on U937 cells (data not shown).

Fig. 6: . Transient transfection of J-Jhan cells with HIV long terminal repeat (LTR) together with human cytomegalovirus (HCMV) IE1/IE2 in the presence or absence of ppUL44 under the control of the major immediate early enhancer/promoter (p433) or the Rous sarcoma virus (RSV) LTR (RSV-44). The data are the arithmetic means of six independent experiments, each run in triplicate. RLU, Relative light units; LUC, luciferase; prot, protein.

In order to determine the expression levels of Tat, IE1/IE2 and UL44, Western blots of cell extracts used in the luciferase assays were performed, and bands were stained with specific monoclonal antibodies. The bands detected on the blots were so faint that no clear conclusion on the expression level of the different proteins could be drawn.

ppUL44 inhibits the synergistic IE1/IE2-Tat transactivation of HIV LTR

We recently observed that coexpression of Tat with the IE1/IE2 complex yields a strong synergistic LTR trans-activation [²⁰]. For this reason, we determined whether ppUL44 (both p433 and RSV-44) can have a negative effect on LTR when this was transactivated by both factors in synergy. For this purpose we cotransfected J-Jhan cells with LTR-luc, pRL45 and LTR-tat.

As expected (Fig. 7), the transactivation obtained by the IE1/IE2 complex was greatly enhanced in the presence of Tat (980-fold). This synergistic transactivation was strongly (fivefold) inhibited when J-Jhan cells were cotransfected with the ppUL44 expression vector. The same results were obtained in U937 cells (data not shown).

Fig. 7: . Transient transfection of J-Jhan cells with HIV long terminal repeat (LTR) together with HIV tat and human cytomegalovirus IE1/IE2 in synergy in presence or absence of ppUL44. The data are the arithmetic means of six independent experiments, each run in triplicate. RLU, Relative light units; LUC, luciferase; prot, protein.

ppUL44 downregulates via a C-terminal domain

In order to identify the ppUL44 region responsible for LTR downregulation, three p433 deletion mutants were constructed (Fig. 1) and used in cotransfection experiments. When plasmid p341 (expressing amino acids 1–341) was cotransfected with LTR-luc in the presence of LTR-tat, no significant variation in the transactivation of the LTR was observed (Fig. 8). However, when the constructs p318 (expressing amino acids 116–433) or p114 (expressing amino acids 320–433) were cotransfected with LTR-luc and LTR-tat, a sixfold inhibition of the Tat-induced LTR activation was observed. The inhibition obtained with the construct p114 was the same as that observed with the construct expressing the entire ppUL44.

Fig. 8: . Transient transfection of J-Jhan cells with HIV long terminal repeat (LTR)-luc together with HIV-tat in the presence of ppUL44 (p433), and its deletion constructs: p341(amino acids 1–341), p318 (amino acids 116–433), p217 (amino acids 217–433), and p114 (amino acids 320–433). The data are the arithmetic means of six independent experiments, each run in triplicate. RLU, Relative light units; LUC, luciferase; prot, protein.

Discussion

The global importance of HIV-1 infection and the high prevalence of HCMV infections in AIDS patients have given rise to many studies of the interactions between these two viruses, although the present picture remains complicated and incomplete [^21,22].

When we began this work, our premise was that ppUL44 would activate HIV LTR. This premise was based on the study by Zhou et al. [¹⁴], who showed that the HHV-6 counterpart of HCMV ppUL44 (P41) could activate LTR when linked to an indicator gene.

Contrary to our expectations, when we transiently expressed ppUL44 in HIV LTR-bearing cells we did not observe any increase in the basal activity of HIV LTR. In addition, our data showed that ppUL44, an early protein with DNA-binding properties, has a negative effect on the ability of HIV-1 LTR to respond to both its autologous transactivator Tat, or to heterologous transactivators such as HCMV IE1/IE2 or both factors in synergy. The ability of ppUL44 to down-regulate the Tat-mediated HIV LTR transactivation was also observed in astrocytoma cells permanently expressing ppUL44, although no variation in the LTR transactivation was observed in an astrocytoma cell line expressing HCMV gB under the control of MIEP. These results indicate that the negative effect of ppUL44 is not linked to a transient expression of ppUL44 nor to a particular cell type, and that other HCMV genes under the same experimental conditions do not interfere with HIV LTR.

We also observed that increasing amounts of ppUL44 in the presence of a constant amount of Tat produced a 100% inhibition of Tat-mediated LTR transactivation. This is a dose-dependent effect that could be due to the saturation of interacting sites between ppUL44 with Tat or other cellular factors involved in LTR transactivation. Experiments are in progress to disclose the mechanisms of ppUL44 downregulation of HIV LTR activation and to determine whether ppUL44 interferes with HIV replication.

Our results are interesting because it has repeatedly been shown that HCMV IE1/IE2 can transactivate HIV LTR [⁸], but other HCMV gene products may play a role in the bidirectional interaction between HCMV and HIV. In fact, while incomplete expression of the HCMV genome leads to activation of HIV LTR because of the transactivation activity of IE1/IE2, under conditions in which both HIV-1 and HCMV can undergo fully permissive infection, HCMV can repress HIV-1 gene expression at the transcriptional level [¹⁰]. Moreover, in cells dually infected by HIV-1 and HCMV, HIV-1 production was initially stimulated by HCMV infection and then inhibited as HCMV infection progressed. Therefore, it has been proposed that HCMV can both stimulate and inhibit HIV-1 transcription, with inhibition requiring a critical level of HCMV early gene product. Koval et al. [²¹] and Moreno et al. [²²] showed that HCMV infection influences HIV-1 activity negatively or positively depending on the level of expression of HIV gene products, and that a threshold level of Tat, or of some HIV-1 proteins whose expression requires Tat, is necessary to observe an inhibitory effect of HCMV. In agreement with these findings our experiments indicate that the early gene product ppUL44 can inhibit HIV transcription activated by Tat alone, by the IE1/IE2 complex, or by Tat and IE1/IE2 in synergy. The region of ppUL44 responsible for downregulation is contained within the last 114 amino acids at the carboxyl-terminal region of the molecule. These observations are not in contrast to the report that the HHV-6 homologue of ppUL44 (P41) activates LTR, because the active region of ppUL44 is the carboxyl-terminal portion of the molecule, which does not share significant homology with HHV-6 P41 [¹⁴].

To exclude the possibility that the negative activity of ppUL44 on HIV LTR transactivated by Tat or IE1/IE2 was due to its negative effect on HCMV MIEP, which drives both Tat and IE1/IE2, we repeated the transfection experiments with a construct expressing Tat under the control of the HIV LTR and obtained the same results. Furthermore, it was unlikely that the observed negative effect was due to downregulation of the expression of ppUL44 driven by CMV MIEP, because we obtained the same results with the construct that expressed ppUL44 under the control of the RSV LTR.

Furthermore, to exclude the possibility that the negative activity of ppUL44 on HIV LTR transactivated by Tat was due to a lower amount of Tat produced because of a negative effect of UL44 on the LTR that drives Tat, we repeated the transfection experiments with a construct expressing Tat under the control of the SV40 early promoter, which is not downregulated by UL44, and obtained the same results.

In summary, this is the first report of a CMV gene different from IE1/IE2 that plays a role in HIV gene expression, providing further insight into the molecular interactions that occur between HIV-1 and HCMV. We have shown that the early gene product ppUL44 of HCMV can inhibit the HIV LTR transactivation due to Tat and IE1/IE2, and that the active portion of ppUL44 is localized within the last 114 amino acids.

Speculation on a possible anti-HIV use of peptides from the last 114 amino acids of ppUL44 is tempting, but studies on the mechanism of action of ppUL44 should be completed and the ability of ppUL44 to interfere with HIV replication should be tested before any conclusions are reached.

Acknowledgement

The authors thank S. Michelson for critical reading of the manuscript.