Elucidation of the factors involved in host defense against human immunodeficiency viral infection remains pivotal if viral control may be achieved. Toward these ends, we investigated the function of a putative antiretroviral factor, OTK18, isolated by differential display of mRNA from HIV type 1-infected primary human monocyte-derived macrophages. Molecular and immunohistochemical analyses showed that the OTK18 nucleotide sequence contains 13 adjacent C2H2-type zinc finger motifs, a Krüppel-associated box, and is localized to both cytosol and nucleus. Mutational analyses revealed that both the Krüppel-associated box and zinc finger regions of OTK18 are responsible for the transcriptional suppressive activities of this gene. OTK18 was copiously expressed in macrophages following HIV type I infection and diminished progeny virion production. A mechanism for this antiretroviral activity was by suppression of HIV type 1 Tat-induced viral long terminal repeat promoter activity. Our findings suggest that one possible function of OTK18 is as a HIV type 1-inducible transcriptional suppresser.
Mononuclear phagocytes (MP3; infiltrating perivascular and resident tissue macrophages and microglia) are target cells and vehicles for dissemination of HIV-1 (reviewed in Ref. 1). Although immunosuppression and opportunistic infections result from virus-induced destruction of CD4+ T-lymphocytes, infection of brain MP plays a central role in the pathogenesis of HIV-1-associated dementia (HAD). Indeed, viral replication in brain MP affects leukocyte infiltration, astrocytosis, formation of microglial nodules, and neuronal injury/death, the pathological hallmarks of HIV encephalitis and HAD. Together, HAD is a metabolic encephalopathy, fueled by disordered MP immunity and progressive viral infection (reviewed in Refs. 2, 3, 4).
A critical question in the pathogenesis of HAD is what triggers viral replication late in the course of disease. Certainly, a plethora of viral and cellular factors regulate HIV-1 transcription and affect viral production in MP, lymphocytes, and other cells types. In this regard, viral accessory proteins, including, but not limited to, negative factor (Nef), viral protein R (Vpr), virus infectivity factor (Vif), and the regulatory proteins Tat and Rev actively control viral expression. Nef increases the efficiency of viral replication, enhances virion infectivity, and affects viral expression (5). Vpr and Vif regulate viral production (6, 7) and affect HIV-1-induced apoptosis (8, 9). Tat is perhaps the most studied of all viral proteins modulating HIV replication and is an absolute requirement to complete the viral life cycle (10). Rev facilitates transport of viral transcripts from the nucleus to the cytoplasm along with other functions required for completion of the viral life cycle (11). The host-cell factors that enhance viral replication are commonly cytokines and chemokines. These are produced soon after viral infection and are regulated by groups of genes that respond to HIV-1 gp120 cell signaling (12). In addition, a number of cellular transcriptional factors can repress virus often through the viral long terminal repeat (LTR). These factors include the c-myc promoter binding protein (MBP-1), the tumor suppressor p53, the transcriptional factor YY-1, the regulatory protein T lymphocyte-1, the transcriptional factor E2F1, and the NF-κB (12, 13, 14, 15). All are, in part, regulated by HIV-1 and affect viral production. For example, MBP-1 can suppresses HIV-1 replication (14) and Krüppel-associated box (KRAB) domains repress distinct regions of the HIV-1 LTR and affect viral transcription (16).
Cellular differentiation, activation, and viral infection alter HIV-1 expression in susceptible cells. The best known examples are found in studies of HIV-1-macrophage interactions. In this study, the susceptibility of viral infection parallels monocyte-macrophage differentiation and has four potential outcomes: abortive, latent, restricted, and permissive. Monocytes are abortively infected and the viral life cycle goes only to its conclusion after the cells differentiate into macrophages (17). Indeed, HIV-1 infection of freshly isolated monocytes results in an abortive infection with low viral titers and minimal cytopathic effects. HIV-1 infection of mature macrophages is permissive and involves a number of transcriptional factors important for both monocyte-macrophage differentiation and the completion of the HIV life cycle. For example, the close correlation between monocyte maturation and viral gene expression is linked in a causal fashion to the induction of specific cellular nucleic acid-binding proteins (18). How such cellular and viral factors modulate virus infection is pivotal as it may underlie the considerable viral restriction that occurs throughout much of the course of HIV-1 infection. This is especially relevant in the nervous system where permissive viral replication in MP correlates, in part, with disease progression (19). Cellular factors that regulate HIV-1 expression in the nervous system are thus a fertile area of investigation (20, 21, 22, 23). This is true not only for MP but also in astrocytes where a more restricted infection predominates. A number of transcriptional factors in astrocytes have been uncovered including, the orphan nuclear receptor chicken OVA upstream promoter transcriptional factor. This was shown to function as an HIV-1 transcriptional activator by directly interacting with the viral LTR (21), and is facilitated by direct interaction with either Sp1 (22) or Tat (23). Despite the accumulation of this large body of knowledge for regulation of HIV infection, few studies have served to characterize the transcriptional factors operative in MP. These may serve important roles in permitting ongoing viral replication and dissemination in the infected human host. Later in disease, such factors may also affect the levels of secreted toxins in the CNS, which contribute in large measure to disease (22).
In attempts to better understand the molecular events in viral-MP interactions that contribute to restricted viral infection, we used mRNA differential display to identify cellular genes differentially regulated following HIV-1 infection. We hypothesized that alterations in innate MP immunity serve as a compensatory mechanism to curtail viral growth and inevitably, as intracellular controls break down, to induce dysregulation of cellular homeostasis leading to tissue damage and clinical disease. Such a loss of intracellular controls would result in sustained viral replication and secretory activities and could have particular relevance to CNS disorders where infected MP play a principal role in disease pathogenesis (1). In this study, we identified a transcriptional factor, OTK18 (24), which was induced during HIV-1-infection of human monocyte-derived macrophages (MDM) and served to restrict viral infection. The OTK18 sequence contains 13 zinc fingers and has a KRAB domain, which is identified as one type of motif responsible for transcriptional suppression (25). The data, in toto, suggest that one function of OTK18 is to act as an innate MP transcriptional factor serving to modulate HIV-1 replication.
Materials and Methods
Isolation and culture of primary human MDM
Monocytes were obtained from leukopheresis of HIV-1, -2, and hepatitis B seronegative donors, and purified by counter current centrifugal elutriation (26). Cell suspensions were documented >98% monocytes by criteria of cell morphology in Wright-stained cytosmears and CD68 immunolabeling. Monocytes were cultured as adherent cells at 35 × 106 cells in T-75 flasks (1 × 106 cells/ml) or 3 × 106 cells/well in six-well culture plates. Monocytes were cultured in DMEM (Sigma-Aldrich, St. Louis, MO) with 10% heat-inactivated pooled human serum, 1 mM glutamine, 50 μg/ml gentamicin (Sigma-Aldrich), 10 μg/ml ciprofloxacin (Sigma-Aldrich), and 1000 U/ml highly purified recombinant human M-CSF (a generous gift from Genetics Institute, Cambridge, MA). M-CSF was placed into culture for the initial 7 days of monocyte cultivation (to promote differentiation to macrophages), and then removed. Culture medium was changed every 3 days. All tissue culture reagents were screened before use and found negative for endotoxin (<10 pg/ml; Associates of Cape Cod, Woods Hole, MA) and mycoplasma contamination (Gen-probe II; Gen-probe, San Diego, CA).
HIV-1 infection of MDM
After 7 days in culture, MDM were infected with HIV-1ADA at a multiplicity of infection of 0.1 (26), or left untreated as controls. Reverse transcriptase (RT) activity was determined in culture fluids (n = 3) added to a reaction mixture of 0.05% Nonidet P-40 (Sigma-Aldrich), 10 μg/ml poly(A) oligonucleotide, 0.25 μg/ml oligo(dT) (Amersham Pharmacia Biotech, Piscataway, NJ), 5 mM DTT (Amersham Pharmacia Biotech), 150 mM KCl, 15 mM MgCl2, and [3H]dTTP (5 Ci/mmol; PerkinElmer, Wellesley, MA) in 50 mM Tris-HCl buffer (pH 7.9) for 18 h at 37°C. Radiolabeled nucleotides were precipitated with cold 10% TCA and 95% ethanol in an automatic cell harvester (Skatron, Sterling, VA) and quantitated by liquid scintillation spectroscopy (27). The resulting data were analyzed for statistical significance by both the Student t test and ANOVA with the Newman-Keuls posttest.
mRNA differential display
cDNA from human MDM samples was prepared using single base anchored oligo(dT) antisense primers (RNAimage kit 1; GenHunter, Nashville, TN) according to the manufacturer’s protocol. PCR was conducted in duplicate, using arbitrary primers (H-AP1 through H-AP16) and [35S]dATP (600 Ci/mM; ICN Radiochemicals, Irvine, CA). The resulting 35S-labeled PCR products were run on a 6% PAGE with 6 M urea, dried onto Whatman 3 MM paper (Fisher Scientific, Pittsburgh, PA), and analyzed. Subsequently, the gels were exposed to X-OMAT XAR2 film (Eastman Kodak, Rochester, NY) to localize and cut out the expressed sequence tags (ESTs). Bands selected for EST cloning (those up-regulated in treatment groups vs controls) were isolated, reamplified by PCR, and blunt-end cloned into pCR-TRAP (GenHunter). Tetracycline-resistant colonies were checked for inserts by colony PCR and DNA sequences obtained by using vector-specific primers (Lgh and Rgh) according to manufacturer’s instructions. The DNA sequences were database searched and ESTs (with the exception of housekeeping genes) were subjected to the physiological screen as described in the SYBR Green real-time RT-PCR section. For cloning of OTK18, the 2438-bp fragment was obtained by RT-PCR using oligo(dT) antisense primers and superscript RT (Invitrogen, Carlsbad, CA) in the RT step, and OTK18-specific sense (5′-GAA AAT CCA AAC ACC TAT CC-3′) and antisense (5′-AAG GAC ATT TCT GCT TAC TC-3′) primers in the PCR step using the following thermocycler parameters: 94°C, 3 min; (94°C, 30 s; 60°C, 30 s; 72°C, 3 min)30 cycles, 72°C, 10 min and 4°C, hold. The ends of this fragment were filled in using T4 DNA polymerase (Invitrogen) and cloned into the pCR-TRAP vector (GenHunter). The OTK18 full-length gene was also cloned into the pCDNA3.1+ vector to be used in subsequent transfection studies.
Recombinant adenoviral constructs
Recombinant adenoviruses expressing OTK18 and green fluorescent protein (GFP; R-OTK18 virus) were constructed by the AdEasy system (kindly provided by Dr. B. Vogelstein, Johns Hopkins University, Baltimore, MD) as described (28, 29). Full-length OTK18 cDNA was subcloned into BglII-XhoI sites of the pAdTrack-CMV vector, which dually expresses GFP and target genes. The plasmid was linearized by PmeI digestion and subcloned into the adenoviral DNA backbone (pAdEasy-1) by homologous recombination in Escherichia coli (BJ5183). The recombinant DNA was purified by maxiprep, linearized by PacI digestion, and transfected into human embryonic kidney (HEK) 293 cells using Lipofectamine Plus (Invitrogen). The viruses were harvested from the cells and amplified by repeated infection into HEK293 cells, followed by purification using CsCl banding as described (29). The purified virus was used as 3 × 107 expression-forming units (efu)/ml for infecting 3 × 106 MDM in six-well culture plates and 2 × 106 efu/ml for 1 × 105 MDM in 24-well plates. The infection occurred for 1 h at 37°C, the media were replaced and cells were incubated for an additional 24 h. Control virus-expressing GFP (R-CON virus) was generated by linearizing pAdTrack-CMV. MDM viability was analyzed 3 and 5 days after infection with recombinant adenovirus by the Cell Death Detection ELISAPLUS kit (Roche Applied Bioscience, Indianapolis, IN). This system detects cytoplasmic histone-associated DNA fragments (mono- and oligonucleosomes) produced early in apoptosis. Data were normalized as percent of controls that elicit 100% cell death.
SYBR green real-time quantitative RT-PCR
To verify the expression of the OTK18 mRNA, the SYBR green real-time quantitative RT-PCR kit (Applied Biosystems, Foster City, CA) was used on total RNA extracted from MDM or HEK293 cells. OTK18 primers were designed using the Primer Express Software (Applied Biosystems). One hundred nanograms of total RNA were reverse-transcribed using 200 nM of the OTK18 reverse primer (5′-AGG ACG TGA CCG TGG ACT TC-3′) in the RT step and the same reverse primer and the forward primer (5′-ACA TCC CGG TAC AGG CAT CTC-3′) in the PCR step. Real-time PCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems) according to the manufacturer’s instructions. Reactions were conducted in triplicate and performed in a 25-μl volume with 100 nM primers, 3 mM MgCl2, 1 mM nucleotide (dATP, dCTP, dGTP, and dUTP) mix, 0.025 U/μl Amplitaq Gold polymerase, (0.01 U/μl) Amp Erase, and 1× PCR SYBR Green Buffer (all reagents obtained from Applied Biosystems). The reaction cycle parameters were an initial incubation at 50°C for 2 min, denaturation at 95°C for 10 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min. The SYBR green signal was continually monitored and the amplified PCR products were analyzed in the linear range for amplification with standards. To confirm the amplification specificity, the PCR products were subjected to melting temperature dissociation curve analysis. To determine the concentration of OTK18 within a given experimental sample, an OTK18 standard curve was developed from human genomic DNA extracted from monocytes. The standard curve was generated from 100, 10, 1.0, and 0.1 ng of genomic DNA (n = 3) along with no template controls (n = 3). In parallel, no amplification controls were run to rule out the presence of fluorescence contaminants in the sample or in the thermocycler heat block. The resulting data were transformed into copy number and subjected to one-way ANOVA with Newman-Keuls posttest for statistical significance (p < 0.05).
OTK18-GFP fusion protein
The full-length OTK18-GFP fusion protein (pOTK18-GFP) was generated by PCR amplification of OTK18 using the primer set (5′-ATA TGT CGA CAT GCC TGC TGA TGT GAA TTT ATC C-3′, and 5′-ATG GAT CCT TGC TTG GTA AAG CCT TTC ACA GA-3′), Pfu Turbo DNA polymerase (Stratagene, La Jolla, CA), and the following thermocycler parameters: 94°C, 3 min; (94°C, 30 s; 60°C, 30 s; 72°C, 3 min)28 cycles; 72°C, 10 min; and 4°C, hold. The resulting PCR product was digested with SalI and BamHI, ligated to pEGFP-N3 vector (Clontech Laboratories, Palo Alto, CA), which was digested by the same restriction enzymes. The plasmid was sequenced and purified using an endotoxin-free maxiprep kit (Qiagen, Valencia, CA) and subsequently transfected into HEK293 cells by using GenePorter (Gene Therapy Systems, San Diego, CA). pEGFP-N3 vector was used as a control for EGFP expression. The transfected cells were lightly fixed with 2% paraformaldehyde and stained with propidium iodine (PI) for nuclear localization. The expression of the pOTK18-GFP fusion protein was confirmed by both immunoblotting using anti-GFP mAb (Clontech Laboratories) and GFP fluorescent signal. The Hoffman modulation contrast and fluorescent images were captured by Nikon TE-300 inverted fluorescent microscope attached to a digital camera (DVC-1300C; DVC Company, Austin, TX) and using B-2E/C filter.
The human cytokine expression arrays (R&D Systems, Minneapolis, MN) were used for cDNA microarray analysis. Total RNA was extracted from human MDM (n = 3) at 48 h after infection with R-OTK18 or R-CON recombinant adenovirus. Replicate cells were stimulated with human CD40 ligand (1 μg/ml, a generous gift provided by Immunex, Cambridge, MA) for 8 h before harvesting. Probe cDNA was prepared from 10 μg of total RNA using human cytokine-specific primers (R&D Systems), superscript RT (Invitrogen), and [α-32P]dCTP (Amersham Pharmacia Biotech). Prehybridization and hybridization were conducted using ExpressHyb solution (Clontech Laboratories) according to manufacturer’s instructions. Hybridization was conducted by adding 2 × 106 cpm of probe per milliliter to ExpressHyb solution at 65°C overnight. The membranes were washed with 1× SSC containing 0.5% SDS according to the manufacturer’s instructions. The membranes were exposed to a PhosphorScreen and scanned by a PhosphoImager (GeneStorm; Amersham Pharmacia Biotech). Each spot was quantified by ImageQuant (Amersham Pharmacia Biotech). The ratio of each cellular gene investigated (OTK18, for example) to the sum of housekeeping genes (β2 -macrogloblin, β-actin, cyclophilin A, GAPDH, HLA-A 02101 H chain, hypoxanthine-guanine phosphoribosyltransferase, L19, transferrin receptor, and α-tublin) was calculated as a percent ratio. MDM (n = 3) were investigated and results are shown as an average (Table I).
|Gene .||Group .||R-OTK18* .||R-CON+ .||Fold Suppression .|
|CD14||Cell surface protein||0.807||2.201||2.729|
|CD4||Cell surface protein||0.600||1.271||2.118|
|Gene .||Group .||R-OTK18* .||R-CON+ .||Fold Suppression .|
|CD14||Cell surface protein||0.807||2.201||2.729|
|CD4||Cell surface protein||0.600||1.271||2.118|
The ratio of R-CON/R-OTK18 is shown as fold suppression. Genes placed in bold relate to molecules which may play a role in possible mechanisms for OTK18 inhibition of viral infection of MP. R-OTK18* and R-CON+ are the average percent of gene expression compared to the total housekeeping genes. The results shown are the average of three independent experiments from three separate MDM donors.
Monocytes from three separate donors (n = 3) were cultured for 9 days and infected with R-CON, R-OTK18, or left uninfected for 72 h, GFP expression was monitored, and cells were split into tubes containing 2 × 106 cells. Cells from each donor were independently subjected to FACS analysis using PE-conjugated CD4, CXCR4, or CCR5 (BD PharMingen, San Diego, CA) at the FACS Core Facility at the university as described previously (30). The gated PE signal intensity of GFP+ cells were averaged and subjected to Student’s t test and a one-way ANOVA with Newman-Keuls posttest.
GAL4 DNA binding domain-OTK18 (GBD-OTK18) was constructed by subcloning the full-length OTK18 gene into the pFA-CMV vector (Stratagene) using BamHI-XhoI sites. GAL4-luciferase reporter gene (GAL4-Luc; Stratagene), thymidine kinase (TK) promoter-driven Renilla luciferase (pTK-RL; Promega, Madison, WI), and GBD-OTK18 were transfected into HEK293 cells (0.5 × 106 cells/well, Lipofectamine Plus; Invitrogen). Forty-eight hours after transfection, cells were collected and the luciferase activity was measured by luminometer (Berthold Systems, Aliquippa, PA) using the Dual-Luciferase kit (Promega). UAS-TK-LUC, which contains the GAL4-TK minimal promoter-luciferase gene (GAL4-TK-Luc; kindly provided from Dr. B. Lowe, Northwestern University, Evanston, IL), was used instead of GAL4-Luc for monitoring transcriptional suppressor activity. Studies were repeated with deletion mutants of GBD-OTK18, formed by removing portions of OTK18 and KRAB boxes A (27–66) and B (67–89). GBD-OTK18 deletion mutants, 1–58, 1–115, 1–241, and 1–300, were constructed by digesting OTK18 with SacI, HindIII, PstI, or EcoRI, followed by blunt-end ligation of the plasmid DNA. GBD-box A, box AB, and box B were constructed by PCR amplification for OTK18 amino acid 27–66, 27–89, 67–89, and subcloning with a pFA-CMV vector. GBD-Δbox A and Δbox AB, which were deletion mutants of full-length OTK18 at amino acids 27–66 and 27–89, respectively, were constructed by PCR amplification of GBD-OTK18. GBD-OTK18 90–115, 90–241, and 90–300 were constructed by digesting GBD-Δbox AB with HindIII, PstI, or EcoRI, followed by blunt-end ligation of the plasmid DNA.
To assay the effect of OTK18 on HIV-1 LTR activity, we obtained plasmids from the National Institutes of Health AIDS Research and Reference Reagent Program (Rockville, MD) for pSV2tat72 (HIV-1 Tat1–72 expression vector) (31), and a luciferase reporter gene containing the HIV-1 LTR (pLTRwtLITE) (32). OTK18 was subcloned into BamHI-XhoI sites of pcDNA3.1+ (Invitrogen) for expressing full-length OTK18 in mammalian cells. An OTK18 mutant lacking the DNA binding sequence (pcDNA-OTK18Δ) was constructed by restriction digestion of pcDNA-OTK18 with EcoRI, followed by blunt end ligation. HEK293 cells were transfected with pSV2tat72, pLTRwtLITE, pTK-RL, and either pcDNA3.1+, pcDNA-OTK18, or pcDNA-OTK18Δ, using GenePorter (Gene Therapy Systems). Forty-eight hours after transfection, cells were collected and luciferase activity measured as described above. The expression level of recombinant Tat was confirmed by RT-PCR using the primer set (sense: 5′-AAA CTA GAG CCC TGG AAG C-3′ and antisense: 5′-CTT GAT GAG TCT GAC TGT CTT G-3′), which amplified a 181-bp PCR product. The Tat RT-PCR signal intensity was normalized by β-actin mRNA. RT-PCR signal intensity by RT-PCR of Tat in total RNA obtained from pcDNA3.1+, pcDNA-OTK18, and pcDNA-OTK18Δ transfected HEK293 cells (n = 3) was performed. Total RNA (0.5 μg) was reverse-transcribed using the antisense primer (5′-CTT GAT GAG TCT GAC TGT CTT G-3′) in the RT step and the same antisense primer and the sense primer (5′-AAA CTA GAG CCC TGG AAG C-3′) in the PCR step using the following thermal cycle program: 95°C, 2 min; (94°C, 30 s; 50°C, 30 s; 72°C, 3 min)28 cycles; 72°C, 5 min; and 4°C, hold. The same RT-PCR program was used for the amplification of the human β-actin reference gene that was used for normalizing the OTK18 signal. These samples were subjected to 2% agarose gel electrophoresis with a 100-bp DNA ladder (Invitrogen). The ethidium bromide-stained agarose gel was visualized under UV light. Images were digitally captured and band intensities quantitated by ImageQuant software (Amersham Pharmacia Biotech). RT-PCR band signals of unsaturated intensities were used for the statistical analyses. The level of expression of OTK18 from pcDNA3.1+, pcDNA-OTK18, or pcDNA-OTK18Δ transfected HEK293 cells was assessed by SYBR Green Real-time Quantitative RT-PCR as outlined in SYBR Green Real-time Quantitative RT-PCR. All resulting data was subjected to the Student t test and a one-way ANOVA with Newman-Keuls posttest.
OTK18 mAbs were generated by s.c. injection of His-tagged OTK18 1–178(1–178) truncated recombinant protein purified from BL21(DE3)LysS E. coli using a pET-28a+ vector (Novagen, Madison, WI) into female BALB/c mice at the University Monoclonal Ab Facility (33). Splenocytes were isolated from mice and fused with myeloma cells to establish hybridomas. The hybridoma clones were selected by ELISA, immunohistochemistry, and immunoblotting assays.
OTK18: composition and structure
We used mRNA differential display technology (34) to search for up-regulated gene products at 7 days after HIV-1 infection of MDM as compared with uninfected or LPS-stimulated MDM. Of 32 ESTs obtained, one was found to be selectively up-regulated following HIV-1 infection of MDM, which was also confirmed by RT-PCR. The EST corresponds to OTK18 (24) or ZNF175 (GenBank accession numbers D50419 and NM_007147, respectively), which was not characterized in the original report. OTK18 is composed of 711 amino acids and localized to chromosome 19q13.4. The entire 17-kb genomic sequence of OTK18 is composed of six exons and five introns, generating four segments of coding sequence.
OTK18 is a transcription suppressor
To initiate the identification of its function, we focused on the 13 zinc finger motif at amino acids 300–711 (Fig. 1,A), which is known be a DNA binding region of transcriptional regulators. Because the target DNA binding sequence is unknown, a GBD-fusion protein/GAL4-luciferase reporting system was used to address its regulatory function on gene expression (35). OTK18 was expressed as a GBD fusion protein that should bind to the GAL4 element of the promoter region of luciferase reporter vector, thus enabling us to monitor the effect of OTK18 on luciferase gene transcription. Recombinant GBD-OTK18 was constructed, then cotransfected with GAL4-Luc or GAL4-TK-Luc, which contains the TK minimal promoter sequence downstream of GAL4 (35). At 24-h posttransfection, GBD-OTK18 failed to activate GAL4-Luc transcription, negating its transcriptional activator function. It, however, suppressed GAL4-TK-Luc gene transcription in a DNA dose-dependent manner (up to 54% inhibition, Fig. 1 B). Control TK-RL activity, which does not have a GAL4 DNA sequence at the 5′ end of the TK promoter region, was not affected by either GBD or GBD-OTK18 expression, serving as a negative control for the gene suppression (data not shown). This demonstrated that OTK18 could function as a transcriptional suppressor.
OTK18 contains KRAB sequences
To further characterize the potential transcriptional suppressor region of OTK18, amino acids 58–115 were subjected to basic local alignment search tool subroutines to search for equivalent amino acid sequences present in other mammalian cDNAs. Several zinc finger proteins (Homo sapiens zinc finger 41 (HSZNF41), kidney, ischemia, and developmentally regulated gene 1A (KID1A), ZNF157, gonadotropin-inducible transcription repressor(GIOT)-3, and retinoblastoma gene product-associated Krüppel-associated box repressor (RBAK); GenBank accession numbers X60155, M96548, U28687, AB021643, and AF226869, respectively) were found to contain sequences similar to OTK18. These proteins all share a KRAB motif, critical for transcriptional suppression and found in 30% of all known zinc finger proteins (Fig. 2) (36). An extension of KRAB B (box B1) was found in both OTK18 and HSZNF41 sequences. However, there is no motif or homologous sequence on OTK18 at amino acids 90–300, distinguishing OTK18 from other KRAB-containing zinc finger proteins. Thus, the KRAB box is a potential region responsible for the suppressive function.
Determination of OTK18 suppressive activities
To confirm whether the KRAB box is a region responsible for suppression, we generated truncation mutants of GBD-OTK18 (amino acid 1–58, 1–115, 1–241, and 1–300) to determine the stretch of the OTK18 sequence responsible for the suppressive function of the gene (shown in Fig. 3,A). As shown in Fig. 3,B, mutant 1–58 did not suppress GAL4-TK-Luc transcription. However, mutants 1–115, 1–241, and 1–300 suppressed transcription at levels equivalent to the full-length 1–711 constructs. These data indicate that amino acids 58–115 contain one of the regions responsible for the suppressive functions of OTK18 (Fig. 3 B). None of the GBD-OTK18 mutants had an effect on TK-RL gene expression, which was used as a negative control.
To determine the effect of KRAB on GAL4-TK-Luc activity, we constructed KRAB mutations of GBD-OTK18. These consisted of deletions of box A or boxes A and B (Δbox A or Δbox AB), and deletion of the remainder of OTK18 with retention of one or both KRAB boxes (box A, box B, or box AB) (Fig. 4,A). Using the luciferase assay, we demonstrated that neither GBD nor GBD-box B alone suppressed GAL4-TK-Luc transcription. However, GBD-box A alone partially suppressed transcription. The other mutants (Δbox A and Δbox AB) suppressed it equally as well as full-length OTK18. This data indicated that expression of box AB is sufficient for gene suppression activity and complete box A is necessary for the suppressive activity as shown in Fig. 4,A. The suppressive effect of GBD-Δbox AB suggests that another transcriptional suppressive region may be present between amino acid 90–711 of OTK18 (Fig. 4,B). We tested another series of OTK18 deletion mutants (90–115, 90–240, and 90–300), which do not contain KRAB but other regions of OTK18. As shown in Fig. 4 C, GBD-OTK 90–115, 90–240, and 90–300 failed to suppress transcription, suggesting the zinc finger region (301–711) contained a secondary suppressive domain. In summary, these studies demonstrated that transcriptional suppressor function is located on both KRAB and zinc finger regions of OTK18. None of the OTK18 KRAB mutants had an effect on TK-RL gene expression, which served as a negative control.
The subcellular localization of OTK18 was determined in HEK293 cells expressing the OTK18-GFP fusion protein in comparison with control GFP protein expression. OTK18-GFP (Fig. 5,B, green) was mainly colocalized with PI-positive nuclei (Fig. 5, C and D). OTK18 has a putative nuclear localization signal, RKKP, which was located at position 359 between zinc fingers 1 and 2, which might be involved in the nuclear localization. Control GFP signal was mainly localized in cytosol (Fig. 5,F) and did not demonstrate marked overlapped with the PI signal (Fig. 5,H). To further characterize the possible functions of OTK18, we developed a mAb against the N terminus of OTK18 for immunohistochemistry and immunoblotting. As shown in Fig. 6, A–D, MDM express endogenous OTK18 (Fig. 6,B, red), which is mainly localized in the nucleus (Fig. 6, C and D) but can also be detected in the cytoplasm. To further characterize the possible functions of OTK18, we generated dual-expression adenoviral constructs for GFP either with (R-OTK18) or without (R-CON) full-length OTK18 to be used in subsequent analyses. To assess the viral efficiency of the adenoviral constructs in human MDMs, we equally infected these cells with either R-OTK18 or R-CON and determined the expression levels by monitoring the GFP signal (Fig. 6, E and F). These panels demonstrate the equal and ubiquitous distribution of GFP signal throughout the infected MDM, which equated to 60–70% infection efficiency. The expression of recombinant OTK18 by the adenovirus was also confirmed by immunoblotting. As shown in Fig. 6,G, R-OTK18 virus infection expresses recombinant OTK18 at 75, 65, and 35 kDa bands (Fig. 6,G, lane 3). This suggests that full-length OTK18 might have posttranslational processing to generate different fragments, which might exhibit different patterns of subcellular localization as shown in Fig. 6,B. Actually, truncated OTK18 (1–300), which does not contain zinc finger region or nuclear localization signal, is mainly distributed in cytosol, suggesting its different subcellular localization mechanism upon protein processing (data not shown). Uninfected (Fig. 6,G, lane 1) or R-CON infected (Fig. 6 G, lane 2) cells express no detectable OTK18.
Effects of OTK18 on MDM gene expression
To further characterize the possible functions of OTK18, we generated dual-expression adenoviral constructs for GFP either with (R-OTK18) or without (R-CON) full-length OTK18 to be used in subsequent analyses. To assess the viral efficiency of the adenoviral constructs in human MDMs, we equally infected these cells with either R-OTK18 or R-CON and determined the expression levels by monitoring the GFP signal (Fig. 6). These panels demonstrate the equal and ubiquitous distribution of the GFP signal throughout the infected MDM, which was ∼60–70% infection efficiency. After infection, total RNA was isolated and subjected to human cytokine cDNA arrays. As would be expected for a transcriptional suppressor, most genes incorporated onto the microarray were down-regulated (from 1.3- to 7.6-fold) when recombinant OTK18 was introduced (Table I). Of these, a diverse group of chemokines (α, β, γ, and δ) was affected. The α-chemokines, IL-8 and growth-related oncogene (GRO) α, β, and γ were down-regulated. These α-chemokines serve as chemotactic stimuli for polymorphonuclear leukocytes and lymphocytes. The β-chemokines, including I-309, macrophage inflammatory protein (MIP)-1α and β, MIP-3α, and RANTES promote migration of monocytes, lymphocytes, eosinophils, and basophils (37). The gene expression of cell surface protein, CD4, and chemokine receptors CXCR4 and CCR5, which are critical for HIV-1 entry in MDM, were down-regulated 47, 70, and 50%, respectively (Table I). To evaluate whether these effects on CD4 and chemokine receptor RNA messages are also seen at the protein level, we performed FACS analyses for CD4, CXCR4, and CCR5 on MDM infected with R-OTK18, R-CON, or uninfected (n = 3). In cells infected with R-OTK18, the levels of CD4 and CCR5 in GFP+ cells were down-regulated ∼15 and 23%, respectively, whereas the level of CXCR4 in GFP-positive (GFP+) cells was up-regulated ∼19%, as compared with R-CON cells (not statistically significant), suggesting there is minimal effect of OTK18 expression on HIV coreceptor expression on the cell surface (data not shown).
OTK18 suppresses HIV-1 replication
To determine whether OTK18 could affect HIV-1 replication, we infected MDM with R-OTK18 or R-CON at 2 × 106 efu/ml for 1 × 105 MDM in 24-well plates (assessed by GFP signal; Fig. 6, E and F), then infected the cells with HIV-1ADA (a macrophage tropic HIV-1). This viral strain typically results in infection of >80% of MDM after 7 days at a multiplicity of infection of 0.1. To account for the fact that dual viral infection alone decreased HIV-1 replication, and to evaluate the effect at different time-points, we performed a series of experiments with different dilutions of the adenovirus (1/3, 1/5, 1/10, 1/100), and determined HIV-1 RT activity in supernatants (as above), at 3, 5, and 7 days after infection (n = 3). Fig. 7,A demonstrates results for a time-course experiment with adenovirus at 2 × 105 efu/ml for 1 × 105 MDM in 24-well plates. At day 3 after infection with HIV-1, all groups (MDM, R-CON, and R-OTK18) had low levels of HIV-1 RT activity, and no differences were detected among the treatment groups. At day 5, however, while there were no differences between the HIV-1-infected MDMs and the R-CON group, R-OTK18 had statistically lower levels of HIV-1 RT activity vs either the MDM or R-CON group (p < 0.05 for each). To demonstrate that the differences are not due to cell death caused by viral infection, we performed a DNA fragmentation assay at different times postinfection. Fig. 7,B demonstrates that there are no significant differences in the percentage of cell death as measured by histone-associated DNA complex (nucleosome) ELISA after infection by either R-CON or R-OTK18 as compared with uninfected MDM. To show the level of up-regulation of OTK18 mRNA that could be achieved with the adenovirus infections, we performed real-time RT-PCR for OTK18 mRNA (Fig. 7,C). At day 5 following HIV-1 infection, the levels of OTK18 mRNA were significantly elevated in the R-OTK18 group (5-fold higher than the control HIV-1-infected MDM group and 17.5-fold higher than the R-CON group, p < 0.05). The OTK18 protein expression was also confirmed by immunoblotting (Fig. 7,D). MDM express endogenous OTK18, which migrate around 75 kDa (Fig. 7 D, left lane, uninfected). Infection of R-CON had no effect on OTK18 expression (middle lane), whereas R-OTK18 expressed a 6-fold higher amount of OTK18 as compared with uninfected MDM (right lane, 75 and 65 kDa bands). Thus, the OTK18 gene expression induced by R-OTK18 is not superphysiological but modestly higher than the OTK18 expression induced by HIV-1 infection.
OTK18 and HIV-1 LTR activity
Because OTK18 is a transcriptional suppressor, the most likely target of the HIV-1 life cycle is the inhibition of RNA transcription after genomic integration of HIV cDNA. To address this issue, we investigated a possible mechanism for the OTK18-induced inhibition of HIV-1 replication. In these assays, we used a HIV-Tat/LTR transreporting system (32) to monitor the effect of OTK18 on viral transcription in HEK293 cells. The cells transfected with the HIV-1 LTR reporter vector and pSV2tat72 expressing Tat1–72 had a 10.6-fold increase in HIV-1 LTR promoter activity as measured by luciferase activity over what was observed by the LTR alone (pcDNA3.1+; Fig. 8,A, □). However, when cells were cotransfected with pcDNA-OTK18 and pSV2tat72 (Fig. 8,A, ▦), the HIV-1 LTR activity was suppressed to 3.9-fold without changes in reference gene expression (Fig. 8,A). We confirmed the effect of OTK18 on HIV-1-LTR activity by using a pcDNA-OTK18Δ mutant containing amino acids 1–300. Although the mutant contained the KRAB sequence and transcriptional suppressive activity upon DNA binding (shown by GBD-fusion protein assays), it failed to suppress Tat-stimulated HIV-1 LTR promoter activity (Fig. 8,A, ▪). Conversely, the mutant enhanced the HIV-1 LTR-activity, 2.2-fold when compared with pcDNA3.1+. To ensure that the effects observed were not due to OTK18-mediated suppression of Tat gene transcription, we isolated total RNA from transfected cells and assayed Tat expression by RT-PCR in each of the experimental and control groups. These experiments demonstrated that neither pcDNA-OTK18 nor pcDNA-OTK18Δ suppressed Tat expression, confirming that Tat was present in the HIV-1 LTR activation systems tested (Fig. 8,B). pcDNA-OTK18Δ up-regulated Tat expression. The mechanism for this observation is not yet established. By SYBR Green real-time quantitative RT-PCR, we verified that the pcDNA3.1+-transfected HEK293 cells did not express OTK18, whereas those transfected with either pcDNA-OTK18 or pcDNA-OTK18Δ expressed OTK18 to significant levels (p < 0.001) (Fig. 8,C). Expression of OTK18 and OTK18Δ was also confirmed by immunoblotting of HEK293 cell lysates using an anti-OTK18 monoclonal (Fig. 8,D). Endogenous OTK18 was not detected in untransfected cells (Fig. 8,D, left lane), whereas 40 and 27 kDa bands were detected in pcDNA3.1-OTK18Δ transfected cells (Fig. 8,D, middle lane) and ∼75 and 65 kDa bands were detected in pcDNA3.1-OTK18 transfected cells (Fig. 8,D, right lane). These results indicate that OTK18 may have posttranslational processing to generate short OTK18 fragments lacking the 13-kDa N-terminal region, which may not have a KRAB sequence. This was also observed in MDM infected with R-OTK (Fig. 6 G). In toto, the data support the hypothesis that OTK18 suppresses HIV-1 LTR-mediated gene expression by altering Tat-mediated promoter up-regulation.
The factors involved in the underlying molecular mechanisms that regulate innate CNS immune responses affecting viral replication and influence MP neurotoxic responses are not completely understood. The exact molecular mechanism and all the factors regulating HIV-1 replication in MP is possibly one of the most complex areas of research, as it encompasses all phases of information processing, gene interactions, DNA-protein interactions, protein-protein interactions, transcriptional regulation, and other genetic factors. In this report, we isolated and then described the function of a novel transcriptional suppressor, OTK18, which may play a role in the macrophage control of viral replication during advanced HIV-1 infection. We found that OTK18 mRNA levels in HIV-1-infected MDM were increased 4.6, 5.6, and 1.3-fold on days 5, 7, and 9 after HIV-1 infection when compared with uninfected controls.4 The up-regulation of OTK18 paralleled levels of viral replication.
The data shown in this work suggest that OTK18 may be included among several known MP products that have an effect on viral replication. There are a number of means at the disposal of MP that serve to control HIV-1 production. First and foremost, viral replication is regulated through the abilities of the macrophage to act as a phagocytic cell permitting intracellular killing of a host of microbial pathogens (38). Secondly, regulation can occur through MP effector mechanisms that include the secretion of IFNs (39). Macrophages produce IFN-α and -β in abundance, which serve to inhibit viral entry, in part mediated by secretion of β-chemokines (40, 41, 42, 43). Proinflammatory cytokines, produced by MP following immune activation, also serve to regulate viral growth (42). Indeed, LPS or CD40 ligand-elicited MP activation (44, 45, 46, 47, 48) protect MP from HIV-1 infection. Lastly, cellular factors, such as NF-κB and C/EBPβ, are important in regulating HIV-1 expression in MP. NF-κB is an activator of HIV-1 transcription that is induced in MP upon cytokine-dependent activation and HIV-1 infection (49). C/EBPβ is induced in macrophage differentiation and coordinates gene expression during macrophage activation (50, 51, 52). When MP are activated, NF-κB and C/EBPβ are induced and this induction leads, in part, to increased HIV-1 transcription (53, 54). Such mechanisms work in tandem with humoral and acquired immune responses. The latter is mediated through Ag presentation and T cell expansion. Altogether, the innate, humoral, and acquired immune responses, mediated in large measure through MP, provide the principal means for the regulation of HIV-1 infection in the infected human host (39, 55). The finding that OTK18 inhibits viral replication demonstrates yet another possible factor that may be involved in macrophage control of viral replication. Future experiments are designed to elucidate the intracellular signaling events and parts of the HIV-1 life cycle most likely affected by OTK18.
OTK18 was classified as a transcription factor because of its 13 C2H2-type zinc fingers (24). The specificity of such DNA-zinc finger proteins is largely determined by the conserved structural motifs formed by the interaction of two pairs of cysteines and histidines contained with the zinc ion (56). The consensus sequence of OTK18 is CXE CGK AFX QKS XLX2 HQR XH, which is connected by a TGEKPYX sequence. Our GAL4 reporter assay conclusively demonstrated that OTK18 is a transcriptional suppressor. Mutational analyses revealed that this activity lies within amino acids 26–89. This domain shares homology with a family of zinc finger proteins containing the KRAB motif, a repression domain encoded by a number of transcription factors, including HSZNF41, KID1A, ZNF157, GIOT-3, and RBAK. OTK18 is found on chromosome 19p13 where >40 KRAB-containing zinc finger proteins are clustered together (25, 57). However, amino acids 90–300 have no motif or homology to other proteins of known function, distinguishing OTK18 as a novel transcriptional protein. Our data also suggest that the zinc finger region is involved in transcriptional suppressive activity because deletion of KRAB and the zinc finger domain but not KRAB alone could revert its suppressive activity. More detailed analyses will be necessary to define the zinc finger region responsible for the suppressive activity of the gene. Our OTK18-GFP fusion protein expression study confirmed that OTK18 is a nuclear protein as suggested by its putative nuclear localization signal between the first and second zinc finger. However, endogenous OTK18 expression is detected mainly in the nucleus but can also be found in the cytoplasm (Fig. 6, A–D). Because expressed OTK18 generated multiple protein fragments (mainly 7 and 65 kDa, partly 40, and 35 kDa), it is possible that full-length OTK18 containing the nuclear localization signal is localized in nucleus, whereas processed fragments (40 and 35 kDa) lacking the signal are localized in the cytoplasm. Expression of truncated OTK18 (1–300), which does not contain the signal, was found to be located in the cytoplasm in HEK293 cells (data not shown).
cDNA microarray analyses of MDM demonstrated that the gene expression of a large number of cell surface proteins and chemokine receptors, pivotal for viral infection (58), was down-regulated by OTK18 expression. Of these, CD4, CXCR4, and CCR5 were reduced 2.1-, 2.0-, and 1.4-fold, respectively. FACS analyses of MDM infected with R-OTK18 and R-CON, however, showed modest down-regulation of CD4 and CCR5. Additionally, we evaluated the chemokine receptor levels by ligand binding, using stromal-derived factor-1α to determine the number of CXCR4 receptors, whereas MIP-1α was used for CCR5. We found that infection of MDM with R-OTK18 decreased the number of binding sites for CXCR4 (68% reduction, p < 0.01) compared with R-CON-infected controls, but no statistical difference was observed for that of CCR5, which is the main HIV coreceptor for MP. Because of the variability in these test results, we conclude that OTK18 is unlikely to suppress HIV-1 entry by this gene down-regulation mechanism.
In attempts to discern the role OTK18 may play in the HIV-1 life cycle in MDM, we assayed HIV-1 replication in MDM infected with adenoviral constructs expressing full-length OTK18 (R-OTK18). The amount of OTK18 mRNA by R-OTK18 is 5-fold higher than that by HIV-1 infection, showing that adenovirus-mediated OTK18 gene expression is within the physiological level. These experiments demonstrated that OTK18 expression induced a significant decrease in progeny virion production. Subsequent experiments conducted with the HIV-Tat/LTR reporting system showed one mechanism for viral inhibition by OTK18 is through its interaction with the HIV-LTR. Indeed, OTK18 suppressed Tat-activated LTR promoter activity in a zinc finger-dependent manner without affecting Tat expression, because OTK18Δ lacks the zinc finger domain it failed to suppress the Tat-induced LTR activation (Fig. 8, A and B). The potentiation of HIV-LTR, as well as the Tat mRNA level (driven by SV40 promoter) by OTK18Δ is unexpected, the mutant could be artificially working as a transcriptional activator, which might be able to interact with common sequences between the HIV-LTR and SV40 promoter. The target sequence for OTK18 binding on the LTR promoter is currently under investigation. This work suggests that one of the potential antiretroviral activities of OTK18 was mediated through the viral LTR. The role of KRAB-containing transcriptional factors in regulation of HIV-1 replication is limited. Pengue et al. (59) found that the KRAB domain was able to repress basal HIV-1 promoter activity in HeLa cells. This finding was further supported by the demonstration that a KRAB containing tetracycline-binding protein has been shown to suppress HIV-1 genomic replication through randomly integrated tetracycline response elements within the HIV-1 genomic sequence (16), thereby demonstrating that HIV-1 replication can be suppressed by KRAB-containing transcriptional factors. Ray et al. (14) reported that MBP-1 inhibits HIV-LTR transcriptional activity and HIV-1 replication by overexpression. However, MBP-1 has not been characterized in HIV-1-infected leukocytes, nor is it a family member of the zinc finger proteins or KRAB-containing proteins. Consistent with our results, Reynolds et al. (60) reported that a genetically engineered KRAB-containing C2H2-type zinc finger protein suppressed Tat-mediated HIV-LTR activity. Recently a different CCCH-type zinc finger protein, ZAP, also inhibited HIV replication. This effect was believed to be mediated through depletion of cytoplasmic viral mRNAs (61) and suggests multiple roles of zinc finger molecules on the viral life cycle. Taken together, our data demonstrate that OTK18 is an endogenous transcriptional factor, up-regulated in HIV-1-infected cells, which appears to act as an in vitro replication antagonist of HIV-1, among its many other possible functions. Further study will be necessary to test other possible mechanisms for the role played by OTK18 in the regulation of HIV-1 replication, including suppression of HIV-1 cDNA genomic integration or RT activity.
In toto, our data demonstrate the presence of a transcriptional suppressor produced following viral infection of MDM. It is well-known that HIV-1 integrates into the host cell genome, which precludes its elimination. Currently, the means to eliminate the persistent HIV-1 reservoir within the infected human host or combat ongoing persistent viral replication are limited. This revolves, in part, around deficiencies in the immune surveillance system to eliminate restricted viral infection. Taking advantage of its inhibitory effects on the viral LTR, the potential exists that OTK18 might have therapeutic potential. Nonetheless, the findings presented underscore the potential importance of OTK18 as a factor possibly involved in the cellular regulation of HIV-1 infection.
We thank Larisa Poluektova, Eric Benner, Santhi Gorantla, Tim Moran, Doug Nieman, Shelly Smith, Joel Strominger, Lori Reed, Shannon Wakeley, Greg Weber, James Buescher, and Linda Wilke for research and technical support and lively scientific exchanges. Robin Taylor’s support in providing outstanding editorial assistance is most appreciated. Bert Vogelstein and Bill Lowe of Johns Hopkins and Northwestern Universities, respectively, are acknowledged for providing the AdEasy system and the GAL4(UAS)-TK-LUC plasmids used in these works. The pSV2tat72 (Alan Frankel) and pLTRwtLITE (Steve Zeichner) were obtained from the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases.
This work was supported by the Vada Oldfield Research Foundation (to T.I.) and National Institutes of Health Research Grants K08 MH01552 (to J.L.), R01AI5089401 (to T.I.), P01 MH57556, P01NS31492, P01 NS43985, R01NS34239, R37NS36136 (to H.E.G.), and NCRR P20RR15635 (to T.I. and H.E.G). U.S. patent pending.
Abbreviations used in this paper: MP, mononuclear phagocyte; HAD, HIV-1-associated dementia; LTR, long terminal repeat; MBP, c-myc promoter binding protein; KRAB, Krüppel-associated box; MDM, monocyte-derived macrophage; RT, reverse transcriptase; EST, expressed sequence tag; GFP, green fluorescent protein; HEK, human embryonic kidney; efu, expression-forming unit; GBD, GAL4 DNA binding domain; TK, thymidine kinase; GIOT, gonadotropin-inducible transcription repressor; RBAK, retinoblastoma gene product-associated KRAB repressor; PI, propidium iodine; R-CON, recombinant adenovirus expressing GFP; R-OTK18, recombinant adenovirus expressing OTK18 and GFP; GRO, growth-related oncogene; MIP, macrophage inflammatory protein.
K. A. Carlson, J. Limoges, G. D. Pohlman, L. Y. Poluektova, D. Langford, E. Masliah, T. Ikezu, and H. E. Gendelman. OTK18 expression in brain mononuclear phagocytes parallels the severity of HIV-1 encephalitis. Submitted for publication.