Absence of DICER in Monocytes and Its Regulation by HIV-1^*

Cell Culture

CEM, ACH2, 8E5, Jurkat, J1.1, 293T, U1, and U937 cell lines were obtained from the AIDS Reagent Program. HCT-116 WT and HCT-116 DICER^−/− cell lines were a gift from Dr. B. Vogelstein and were generated as described previously (32). Suspension cell lines were maintained in RPMI 1640 with l-glutamine and penicillin/streptomycin with 10% FBS. Adherent cell lines were cultured in DMEM supplemented with l-glutamine and penicillin/streptomycin with 10% FBS. HCT-116 WT and DICER^−/− cells were cultured in McCoy's 5A modified medium supplemented with l-glutamine and penicillin/streptomycin with 10% FBS. For experiments, cells were always harvested during log phase growth. For experiments, cells were always harvested during log phase growth. To induce maturation of monocytes, cells were stimulated once with 250 nm PMA and allowed to differentiate for 96 h.

Peripheral blood mononuclear cells were obtained from the blood of a healthy donor (YW) and purified by centrifugation through a layer of lymphocyte isolation medium. Cells were resuspended in serum-free RPMI and plated on a culture dish for 1 h at 37 °C. Nonadherent lymphocytes were removed, and the adherent monocytes were cultured in RPMI plus 10% heat-inactivated FBS. Macrophages were further differentiated by incubating in 10 ng/ml M-CSF or PMA for 1 week with medium change every 2 days.

Western Blots

Cell extracts were resolved by SDS-PAGE on 4–20% Tris-glycine gels (Invitrogen). Proteins were transferred to Immobilon membranes (Millipore) at 200 mA for 3 h. Membranes were blocked with Dulbecco's phosphate-buffered saline (PBS) + 0.05% Tween 20 + 3% BSA. Primary antibody against specified antibodies was incubated with the membrane in PBS + 0.1% Tween 20 + 3% BSA overnight at 4 °C. Membranes were washed three times with PBS + 0.1% Tween 20 and incubated with HRP-conjugated secondary antibody for 2 h at 4 °C. HRP luminescence was elicited with SuperSignal West Dura Extended Duration Substrate (Pierce) and visualized on an Eastman Kodak Co. 1D image station. Antibodies used for Western blots were obtained from Abcam (DICER, AB1460; DROSHA, AB12286; AGO1, AB5070; AGO2, 57113), Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) (β-actin, sc-1615), and Abnova (PIWIL4, H00143689-B10P). Densitometry for Western blots was performed using the Carestream software package from Eastman Kodak Co. Levels of β-actin were used to normalize all densitometry data.

Multiple Western blots were performed for the various miRNA-related proteins using various cell lines. Semiquantitative data for protein expression levels were obtained by running multiple Western blots with increasing amounts of whole cell extract until the protein (e.g. DICER or DROSHA) was observable. Expression was converted to percentages by comparing protein expression in all other cell lines with protein expression in CEM cells, with the number rounded to the nearest integer. Proteins that were not observable even in 195 μg of whole cell extract were marked as having 0% expression. ND indicates that no data are present for a particular cell line.

Electroporations and Transfections

For electroporations, U937 cells were resuspended at 3 million cells in 250 μl of RPMI. A total of 5 μg of either pGFPmax (Amaxa) or pNL4-3 was then added to the cell suspension. Cells were pulsed a single time at 210 V, 800 microfarads, and low resistance. Electroporated cells were then immediately transferred to RPMI 1640 with l-glutamine and penicillin/streptomycin with 10% FBS and plated in 6-well plates. Twenty-four hours after electroporation, cells were either treated with 250 nm PMA or an equal volume of vehicle and allowed to differentiate for 96 h. The vehicle for PMA is 100% ethanol. Adherent cells were harvested with trypsin-EDTA and cell scraping. For transfections, 293T cells were seeded in a 24-well culture plate at 40,000 cells/well. The following day, the cells were transfected with 400 ng of the appropriate empty psiCheck2 or cloned plasmid using Attractene (Qiagen) lipid reagent. For the functional DICER assay in monocytes and T-cells, the following shRNA luciferase was used: 5′-GGUGGCUCCCGCUGAAUUGGAGAAC; 3′-UUCCACCGAGGGCGACUUAACCUGAAG. The si-PIWIL4 was SMARTpool siRNA D-004644-01 to -04, and the si-DICER was SMARTpool siRNA D-003483-01 to -04 from Thermo Scientific.

Isolation and Hybridization of miRNAs

MicroRNAs were isolated from cells using the mirVana miRNA isolation kit (Ambion) according to the manufacturer's protocol. RNA from treated and untreated U937 cells was separated into two fractions using this protocol, a large RNA fraction and a “miRNA-enriched” fraction. A total of 250 ng of RNA from each miRNA-enriched fraction was hybridized on an Agilent human microRNA V2 microarray. Hybridization, post-hybridization washing, and scanning were performed using standard Agilent protocols (Agilent Technologies). The quality and consistency of individual hybridizations were confirmed by monitoring the following parameters in the Agilent output file: single channel total fluorescence signal 5 times the additive error, signal-to-noise ratio greater than 5-fold, percentage saturated features less than 1%, and percentage coefficient of variation (%CV) for replicate probes less than 10%. MicroRNA annotation and identification was cross-referenced back to various miRNA data bases, such as miRBase (available on the World Wide Web). Data were normalized by quantile normalization and the GeneChip Robust Multi-array Average procedure implemented in the Partek Genomics Suite (Partek Inc.) and GeneSpring GX (Agilent Technologies).

Poly(A) RT-PCR

For poly(A) RT-PCR detection of miRNAs, 400 ng of RNA from the miRNA-enriched fraction was used to generate cDNA using the Quantimir kit (SBI) according to the manufacturer's protocol. Briefly, small RNA species are polyadenylated, and then RT reactions are performed with a company-provided RT primer. For PCR, a universal reverse primer is provided by the manufacturer. Specific miRNA forward primers are identical in sequence to the miRNA of interest. Q-PCR performed using the DyNAmo HS SYBR Green qPCR kit (New England Biolabs) in an ABI Prism 7100 thermal cycler. Data analysis was carried out using the SDS software package. For miRNAs that produced multiple bands, PCR products corresponding to the amplified miRNAs were run on a 3.5% agarose gel and quantified using the Eastman Kodak Co. 1D software package.

RT-PCR and Primers

For mRNA analysis of miRNA-related enzymes following PMA treatments, total RNA was isolated from cells using TRIzol (Invitrogen) according to the manufacturer's protocol. For analysis of HIV-1 mRNA in electroporated cells, total RNA was isolated and separated into two fractions, a large and a small RNA fraction, using the mirVana miRNA isolation kit (Ambion) according to the manufacturer's protocol. A total of 1 μg of RNA from the large RNA fraction was treated with 0.25 mg/ml DNase I for 1 h, followed by heat inactivation at 65 °C for 15 min. A total of 1 μg of total RNA was used to generate cDNA with the iScript cDNA synthesis kit (Bio-Rad) using oligo(dT) reverse primers.

qPCR of RNAi-related Genes

U937 cells were harvested at days 0, 1, and 4 after PMA treatment. RNA was extracted by TRIzol reagent (Invitrogen). One microgram of total RNA was reverse-transcribed using the Superscript III first strand synthesis kit (Invitrogen). Following RT, 1 μl of cDNA was used for qPCR. Quantitative PCR was performed using Brilliant II SybrGreen qPCR master mix (Stratagene) and primers for GADPH (GGAAGGTGAAGGTCGGAGTCAA, CCTTGACGGTGCCATGGAAT), AGO2 (TTCATCGTGGTGCAGAAGAG, CCCAGAGGTATGGCTTCCTT), DICER (TCGCTGGCTGTAAAGTACGA, TTCAAGCAATTCTCGCACAG), and DROSHA (TGCAACTGGTAGCCACAGAG, ACACTGCTGAAGCTGGGATT). PCR was performed on a Bio-Rad CFX384 real-time system. Expression of each transcript was normalized to GAPDH and displayed as -fold change.

Cloning

The psiCheck2 vector was purchased from Promega. Vector was digested with XhoI (New England Biolabs) and NotI (New England Biolabs) for 1 h at 37 °C and then heat-killed at 65 °C for 20 min. The vector was then treated with Antarctic phosphatase (New England Biolabs) at 37 °C for 30 min. The digested vector was purified by phenol/chloroform extraction followed by ethanol precipitation. Inserts corresponding to two regions of DICER cDNA and three regions of the DROSHA cDNA were generated via PCR amplification. PCR products were gel-purified and then digested with XhoI and NotI. Digested inserts were gel-purified to remove enzyme and digest fragments. Ligations were performed using DNA T4 DNA ligase (Promega) at a 1:3 molar ratio of vector to insert.

Luciferase Assay

Luciferase assays were performed with the Dual-Glo luciferase assay system (Promega) according to the manufacturer's recommendations. Briefly, 293T cells were harvested 48 h after transfection, and 75-μl aliquots from each transfection were transferred into a Costar 3610 96-well assay plate. Cells were lysed with an equal volume of Reagent 1 and incubated at 37 °C for 10 min before measuring luminescence at 570 nm. An equal volume of Reagent 2 was then added and incubated at 37 °C for 10 min before measuring luminescence at 460 nm. Luminescence was measured on a BioTEK Synergy HT multimode microplate reader.

Binding Site Prediction

Predictions for miRNA binding sites were generated using the MiRanda Algorithm version 1.0b distributed by the Computational Biology Center of Memorial Sloan-Kettering Cancer Center. The entire CDS for both DICER NM_177438) and DROSHA (NM_013235) were utilized to predict hybridization between miRNA and mRNA transcripts.

Monocytes Are Deficient for Key miRNA Processing Proteins

We have previously observed that some of the components of the RNAi machinery may be either low or undetectable in monocytes and/or macrophages (14). In fact, upon close examination of the literature, we were unable to clearly define if the components of the classic RNAi machinery were detectable in various HIV-1-susceptible cell types, including monocytes, microglial cells, astrocytes, or CD34⁺ hematopoietic stem cells. Indeed, the bulk of the existing literature describes the miRNA machinery in the context of epithelial cell lines (derived from cervical, lung, and breast cancers), endothelial vascular cell lines, and lymphocyte cell lines (33–38). In the current work, we decided to look at the expression of DICER and other miRNA-related protein in monocytes. The results of such an experiment are shown in Fig. 1A, where Western blot analysis of T-cells compared with monocytes showed no detectable levels of DICER, DROSHA, AGO1, and AGO2 (compare lanes 1 and 2). Interestingly, when monocytes were treated with PMA or M-CSF, which induces a differentiation of these cells into macrophages, these RNAi components were detectable (lanes 4 and 5). Treatment with PMA or M-CSF, but not vehicle, also resulted in visible morphological changes to U937 cells, which became adherent and formed pseudopodia as MDMs (data not shown). In the case of AGO1, a band appears after vehicle treatment (lane 3), but the size of the band is larger than the expected size for AGO1, which is observable in lymphocytes (lane 1) and MDMs (lanes 4 and 5). Nevertheless, there is a clear pattern present. Inducing the differentiation of monocytes results in the up-regulation of miRNA-related proteins.

Results in Fig. 1A used a low total of amount of protein extract (50 μg); therefore, it may be possible that our Western blot detection assay was not sensitive enough to detect these components. Therefore, we next performed another experiment with a greater amount of extract to detect whether the components of the RNAi machinery were present in these cells. Results in Fig. 1B show that at increasing concentrations of the extract (65 and 195 μg), all of the components of the RNAi machinery were detectable (albeit at lower levels compared with T-cells) with the exception of the DICER protein. In order to determine if these results were general to all monocytes or unique to U937 cells, we performed Western blots for the miRNA-related proteins in multiple related cell lines. The results are compiled in Table 1, where there is a clear trend separating cell lines derived from the lymphoid lineage and cell lines derived from the myeloid lineage; namely myeloid cells appear to be generally deficient for the proteins related to miRNA biogenesis until terminally differentiated. Although these results are from cell lines, we have previously observed a similar deficiency with extracts from primary monocytes (14). Collectively, these results indicate that monocytes in general express all of the classical miRNA-related proteins at low levels, with the exception of DICER.

With the observation that the protein levels for DICER were undetectable by Western blot, we proceeded to perform RT-PCR on monocyte cells to investigate whether the expression of DICER mRNA was dependent upon PMA stimulation. The results of this experiment (Fig. 1C) show that mRNA for DICER was detected in both CEM and untreated U937 cells (lanes 1 and 2, respectively). Furthermore, the results in C show that mRNA for DICER was expressed in monocytes at all time points (lanes 3–6) and did not increase over time after PMA stimulation. Conversely, the results in Fig. 1D clearly show that expression of the DICER protein gradually increased over time after treatment with PMA (lanes 3–6).

Because differentiation of U937 cells using PMA results in increasing expression of AGO2, DICER, and DROSHA, we sought to determine if the levels of the mRNA transcripts for each of these genes changed in response to PMA treatment. Total RNA from U937 cells at days 0, 1, and 4 after treatment was reverse transcribed, and the cDNA was used for qPCR. The absolute quantity of each transcript was normalized to GAPDH and is shown as -fold change from day 0 (Fig. 1E). DICER mRNA levels drop by greater than 50% from day 0 to day 4 following PMA treatment. DROSHA and AGO2 levels fluctuate at 1 day post-PMA treatment but rise to 1.7- and 1.6-fold higher than base line by day 4. As a whole, these results indicate that the DICER mRNA is present in undifferentiated U937 cells and is probably repressed at the level of translation.

Finally, to assess whether functional DICER protein was available in these cells but simply not detectable with our current methods, we utilized a function transfection assay. Here we transfected HIV-LTR-Luc with Tat in the presence of either shRNALuc or siLuc and scored for luciferase activity from these cells. Results of such an experiment are shown in Fig. 1F, where transfection of siLuc and not shRNA was able to down-regulate luciferase protein levels in monocytes. Positive control experiments in T-cells showed down-regulation of luciferase activity using shRNALuc, further indication of a functional DICER in these cells because repression of translation is often mediated by specific miRNA interactions with the mRNA. These results therefore suggest a hypothesis where monocytes might restrict the expression of DICER via the miRNA pathway. This hypothesis is elaborated upon below.

HIV-1 Suppresses the Expression of DICER in MDMs

It is known that infected MDMs produce virus, whereas infected monocytes produce relatively little virus (39–41). Because our earlier results from monocytes and MDMs indicated that DICER expression correlated with differentiation, we hypothesized that viral infection might induce the expression of DICER and the differentiation of monocytes in order to promote efficient viral replication. In order to test this hypothesis, we electroporated the wild type HIV-1 clone pNL4-3 into U937 monocytes and then treated infected monocytes with vehicle or PMA. As a negative control, the pGFPmax plasmid was also electroporated into U937 cells, followed by treatment with either PMA or vehicle. U937 cells were then harvested and processed for both Western blots and RT-PCR. Results for this experiment can be seen in Fig. 2A, where the cell extract was Western blotted for DICER, DROSHA, AGO2, and β-actin. In Fig. 2A, extract from uninfected and infected T-cell lines served as positive controls (lanes 1 and 2, respectively), whereas extract from untreated U937 cells (lane 3) served as a negative control. Cells transfected with the pGFPmax plasmid and then treated with vehicle (lane 4) or with PMA (lane 5) showed that transfection with foreign DNA neither forced cells to differentiate nor suppressed the effects of PMA treatment. However, cells transfected with the pNL4-3 plasmid and then treated with vehicle (lane 6) or PMA (lane 7) showed interesting results. Contrary to our initial hypothesis, transfection of viral DNA did not promote the differentiation of monocytes to macrophages based on the lack of DICER expression (compare lanes 3 and 6) and lack of morphological changes (data not shown). Transfecting cells with pNL4-3 followed by treatment with vehicle neither induced the expression of the RNAi components nor induced cell adherence or pseudopodia formation. Furthermore, transfection with pNL4-3 suppressed the expression of DICER even after PMA treatment (compare lanes 5 and 7). In order to ensure that these results were not an artifact of our experimental design, this experiment was repeated using a different GFP vector and without any plasmid (data not shown). In each case, cells transfected with pNL4-3 and treated with PMA showed a suppression of DICER levels when compared with the untransfected cells. Together, these results suggest that the suppression of DICER protein levels in MDMs is an HIV-1-specific phenomenon.

In order to confirm that HIV-1 mRNA was being produced in the transfected monocytes, RNA was isolated from CEM, ACH2 (an HIV-1-infected cell line), and transfected U937 cells using the mirVana RNA isolation kit. The isolated RNA was treated with DNase I and then used to perform RT-PCR for the viral Nef and Env genes as well as the host GAPDH gene (Fig. 2B). Uninfected CEM cells (lane 1) showed no expression of HIV-1 mRNA, whereas infected ACH2 cells (lane 2) showed expression of both Nef and Env. HIV-1 message for both Nef and Env was observed in cells transfected with pNL4-3 (lanes 5 and 6), whereas no viral mRNA was detected in untransfected cells (lanes 3 and 4). Notably, there was a visible increase in Env message after PMA treatment (compare lanes 5 and 6), consistent with prior reports indicating that PMA treatment stimulates viral production (42, 43). Finally, we utilized primary monocytes (donor YW) and asked whether the expression of DICER increased once cells were differentiated into macrophages. The results shown in Fig. 2C indicate that differentiated cell extracts (150 μg) showed an abundance of DICER (lane 2), and the levels were mostly diminished in the presence of pNL4-3 construct (lane 4). Collectively, these results indicate that increased DICER levels in differentiated cell line or primary cells are regulated by HIV-1.

HIV-1 Suppresses miRNA Production in MDMs

The lack of DICER in undifferentiated monocytes and the suppression of DICER in infected MDMs raise important questions about the state of miRNA biogenesis in both of these cell types. Results from animal models with conditional knock-outs of DICER have shown that a DICER^−/− genotype leads to developmental defects, such as failure to differentiate, disruption of heterochromatin, and premature senescence, indicating that the production of miRNAs is important for cell survival and proliferation (44, 45). Additionally, RNAi initiated by DICER has been implicated in the processes of restricting viral replication and promoting viral latency (30, 31, 46). The observed data raised two important and related questions. First, although DICER could not be detected in undifferentiated monocytes, can any cellular miRNAs be detected in undifferentiated monocytes? Second, although HIV-1 transfection suppressed the expression of DICER in MDMs, does this HIV-1-specific suppression result in a significant drop in miRNA levels?

To address both of these questions, we isolated small RNAs from transfected and untransfected monocytes and MDMs and then hybridized the RNA to an Agilent V2 human miRNA chip capable of detecting 722 unique human miRNAs. Total RNA was isolated from four differentially treated groups: cells electroporated without plasmid plus vehicle (treatment 1) or plus PMA (treatment 2) and cells electroporated with pNL4-3 plus vehicle (treatment 3) or plus PMA (treatment 4). Isolated RNAs were column-purified to capture small RNA species using the mirVana miRNA isolation kit and then end-labeled and hybridized on an Agilent human miRNA V2 microarray. The microarray probes possess a secondary structure (resembling shepherd's crooks) that results in preferential binding for correctly sized miRNA only and prevents binding by unprocessed hairpins and large RNA molecules. The results from this microarray indicated that miRNAs were indeed present in undifferentiated U937 monocytes and that the expression of miRNA was altered by both PMA treatment and by transfection with the HIV-1 clone (Table 2). Furthermore, the results from this microarray were consistent with prior publications concerning MDMs and miRNA production. For instance, miR-155 is required for proper myelocyte development, and this miR was detected in both monocytes and MDMs (47). Similarly, the expression of miR-146b was shown to be up-regulated by inflammatory cytokines (48), and a similar up-regulation was observed in our PMA-treated cells, as seen in Table 2. When comparing detectable miRNAs among the different treatments, we observed that the expression of DICER after PMA treatment significantly increased the abundance of several dozen miRNAs. The first column in Table 2 provides the name of each detected miRNA, followed by its -fold increase in expression after the addition of PMA. The third column in Table 2 displays the -fold change in miRNA expression between transfected monocytes (treatment 3) and transfected MDMs (treatment 4). From Table 2, it is clear that cells transfected with the pNL4-3 clone produce significantly fewer mature miRNAs than untransfected cells. The observed drop in miRNA expression also correlates with the observed suppression of DICER protein levels shown in Fig. 2. Taken as a whole, these results suggest that the expression of DICER after PMA treatment resulted in an increase in miRNA production in MDMs and that transfection with HIV-1 suppressed general miRNA biogenesis in infected MDMs.

Finally, it has previously been reported that overexpression of Tat can down-regulate miRNA production in T-cells, but the overexpression of Tat alone did not alter protein levels of DICER (26). Therefore, these experiments suggest that during the normal course of infection in monocytes and MDMs, HIV-1 suppresses RNAi by down-regulating the expression of DICER. To our knowledge, this is the first time that evidence has been presented indicating that HIV-1 has the potential to suppress DICER protein levels in MDMs.

The Viral Protein Vpr Is Required for Suppression of DICER in MDMs

After observing that the HIV-1 virus is capable of suppressing DICER protein levels, we were interested in determining if this observed suppression could be ascribed to a single viral gene. In order to carry out this investigation, we obtained several mutant strains of the pNL4-3 HIV-1 plasmid. These mutant strains were constructed as previously described by Sunamoto et al. (49). Briefly, the mutant strains were generated by replacing the start codon of the individual viral ORFs with a stop codon. The mutant plasmids that we have utilized here include Δrev, Δnef, Δvpr, Δvpu, and Δenv mutations, referred to in this paper as pNL4-3 Δrev, pNL4-3 Δnef, etc. Because a Δtat mutation would not allow efficient viral transcription, no Δtat mutation was generated. Instead, a mutant possessing mutations for all ORFs except tat was used so that the individual effects of Tat could be assessed. This last mutant plasmid can only produce the Tat and is referred to here as pNL4-3 Tat Only.

Next, we repeated our transfections of monocytes with these mutant constructs followed by PMA treatment. Four days after PMA treatment, cells were harvested, and whole cell extract was Western blotted for DICER and β-actin. Western blot data were then quantified via densitometry. DICER protein levels were normalized to β-actin levels (Fig. 3A). DICER protein levels are shown for untransfected and transfected MDMs in lanes 1 and 2. Comparatively, U937 cells that were electroporated with the pNL4-3 plasmid (lane 2) experienced a 5-fold reduction in DICER protein levels compared with levels in transfected cells (lane 1). Interestingly, cells transfected with the pNL4-3 Δvpr plasmid (lane 6) showed a consistent inability to suppress DICER protein levels, thereby indicating that the viral vpr gene may be required for the suppression of DICER.

Although the results for cells transfected with pNL4-3 Δvpr show the most consistent evidence of mediating DICER suppression, there was one other data point worthy of consideration. Cells transfected with the pNL4-3 Δnef plasmid (Fig. 3A, lane 5) were unable to suppress DICER levels. However, the results for the pNL4-3 Δnef plasmid showed a greater S.D. value than the results for pNL4-3 Δvpr plasmid. Nevertheless, these data may indicate a scenario in which Vpr and/or Nef play a part in suppressing the host's expression of DICER.

Finally, we performed a time course experiment where cells were transfected with various pNL4-3 wild types (WT) and mutant constructs, treated with PMA, and processed for suppression of DICER using Western blots. Results in Fig. 3B indicate that Vpr mutant showed the most robust expression of DICER at 48 h and Nef mutant virus showed DICER expression after 96 h post-PMA treatment. The effect of Vpr was further tested in U937 cells transfected with wild type Vpr, where DICER levels significantly decreased after 48 h of transfection in the presence of PMA (Fig. 3C). Collectively, these results indicate that Vpr (and to some extent Nef) may be responsible for the observed suppression of DICER in MDMs.

DICER Is Regulated by miR-106a in Undifferentiated Monocytes

After data from our initial investigations showed that DICER protein levels were absent in monocytes despite the presence of DICER mRNA, we hypothesized that DICER expression in undifferentiated monocytes was blocked at the translation stage. Because miRNA inhibition of translation is a well known mechanism of translational inhibition, we examined our microarray data for any miRNAs that were both down-regulated after PMA treatment and predicted to restrict DICER mRNA translation. To accomplish this, we first consulted the lists of miRNA targets generated by the TargetScan and Pictar-Vert algorithms. Among those miRNAs that were down-regulated more than 2-fold, only miR-18b was predicted by TargetScan to restrict DICER expression.

The miR-18b gene is located on the X chromosome, where it belongs to a polycistronic cluster of six miRNAs transcribed on the same mRNA. We refer to this miRNA cluster as the ChrX 17–92 cluster because it possesses strong sequence similarity to the better characterized 17–92 cluster located on chromosome 13. The members of the ChrX 17–92 cluster are depicted in Fig. 4A, along with the -fold change in expression following treatment of cells with PMA. Each member of the cluster was down-regulated following PMA treatment, which is expected because all miRNAs in this cluster are derived from the same mRNA transcript. However, we observed that the degree of down-regulation was not equal for all cluster members but was related to the position of each miRNA within the cluster. Most miRNAs are derived from RNA polymerase II transcripts, but because this unusual pattern of down-regulation would not be expected of a polymerase II transcript, we explored the possibility that the ChrX 17–92 cluster may be transcribed by RNA polymerase I or III. However, miR-106a and miR-18b were both found to be sensitive to low concentrations of α-amanitin, indicating that this cluster is probably transcribed by RNA polymerase II (data not shown). Therefore, the reason for the uneven down-regulation of cluster members after PMA treatment is not immediately apparent and will require future research. Fig. 4B shows the detected level of each miRNA after each treatment (lanes 1, untransfected cells; lanes 2, untransfected cells plus PMA; lanes 3, pNL4-3-transfected cells; lanes 4, pNL4-3-transfected cells plus PMA). Transfection with pNL4-3 did not appear to affect levels of each miRNA (compare lanes 1 with lanes 3 or lanes 2 with lanes 4).

Because all of the miRNAs of the ChrX 17–92 cluster were down-regulated by PMA treatments, all six members were further checked for binding to DICER mRNA using the MiRanda version 1.0b algorithm. The sequence for the DICER mRNA was RefSeq NM_177438, whereas the sequence for each miRNA was obtained from the Sanger Sequence Database. Using these sequences, the MiRanda algorithm provided three potential binding sites for miR-18b and a single potential binding site for miR-106a (Fig. 4C).

In order to verify these predictions, we PCR-cloned the targeted regions of the DICER mRNA into the 3′-UTR of the Renilla luciferase gene in the psiCheck2 plasmid. As a negative control, we cloned a region of the 5′-UTR of the DROSHA gene into the psiCheck2 reporter. Specific knockdown by a miRNA results in a drop in Renilla luciferase activity. The plasmid constructs were transfected into 293T cells and then harvested and assayed for Renilla luciferase and firefly luciferase activity 72 h later (Fig. 4D). As a negative control, a PCR fragment from the 5′-UTR region of the DROSHA gene was inserted into the vector. The construct bearing the proposed target for miR-18b showed Renilla luciferase activity comparable with the negative control, but the construct bearing the proposed binding site for miR-106a showed a significant drop in enzyme activity.

We then performed an siRNA experiment against miR-106a in undifferentiated primary monocytes and observed a reversal of the endogenous DICER levels in these cells after 72 h (Fig. 4E). Although PMA-treated cells (lane 3) at day 3 showed a macrophage-like phenotype, we observed no apparent change in the structural shape of monocytes after simiR-106a treatment. Collectively, these results indicate that the expression of DICER protein is probably restricted by miR-106a in undifferentiated monocytes.

Evidence That PIWIL4 Can Play a Part in the Generation of miRNAs

Last, because our initial results showed that DICER is undetectable in undifferentiated monocytes but small RNA species were still detected, we hypothesized that proteins other than DICER may be involved in the production of mature miRNAs. As mentioned earlier, eukaryotic cells generate another set of small RNAs, which are known as piRNAs. The generation of piRNAs is carried out by the members of the PIWI protein family in a DICER-independent fashion. PIWI proteins do not possess enzymatic activity but instead protect short segments of double-stranded RNA from degradation by an unknown ribonuclease (50). Although the expression of PIWI family members is often said to be restricted to reproductive tissues, several reports have presented evidence that PIWI proteins are expressed in somatic tissues as well (18–21). Based on these previous reports, we focused on PIWIL4 because mRNA for this protein was reported to be ubiquitously expressed in human somatic tissues (18).

In order to confirm that PIWIL4 is expressed in the cell lines used in our previous experiments, we extracted protein from multiple cell lines and Western blotted for PIWIL4 and β-actin. The results shown in Fig. 5A demonstrate that the full-length PIWIL4 protein can indeed be detected in infected and uninfected T-lymphocytes (CEM, ACH2, 8E5, Jurkat, and J1.1 cells), in monocytes (U937 and U1), and in the colorectal cancer cell line HCT-116. Uninfected cell lines (CEM, Jurkat, and U937) showed protein levels comparable with those of their infected clones (ACH2, 8E5, J1.1, and U1), indicating that infection with HIV-1 has no discernable effect on PIWIL4 levels. Furthermore, the absence of DICER has no visible effect on PIWIL4 levels, as evidenced by the presence of PIWIL4 in both HCT-116 cells and the knock-out HCT-116 DICER^−/− cell line.

Because our previous results indicated that miRNA-related proteins are induced by PMA, we were interested to know if PIWIL4 expression was similarly affected by PMA levels. The results of such an experiment are presented in Fig. 5B. U937 cells were treated with PMA and harvested at 24-h intervals. The protein extract was Western blotted for PIWIL4 and β-actin. Extract from untreated cells (Fig. 5B, lane 1) showed that PIWIL4 is detectable prior to PMA treatment and that protein levels were only slightly altered by PMA treatment over the course of 4 days (lanes 2–5). These results suggest that unlike proteins in the miRNA pathway, the expression of the PIWIL4 protein was not dependent upon monocyte differentiation.

Our results confirmed that PIWIL4 is present in monocyte cells at time points when DICER cannot be detected. In order to test our hypothesis that PIWIL4 may contribute to miRNA production in U937 monocytes, we performed poly(A) RT-PCR with the Quantimir miRNA detection kit to look for multiple bands or a shift in the size of various miRNAs, thereby indicating that processing by PIWIL4 had occurred. The results from this analysis did not show obvious shifts in size but instead showed that mature miRNAs produced either single or paired bands on an agarose gel. Two miRNAs representative of our findings are shown in Fig. 5C. Results for miR-18b (top) are representative for an miRNA that produces only a single band on a gel. Interestingly, mature miRNA for miR-18b was detected in U937 cells both prior to and after PMA treatment (compare lanes 1 and 3 with lanes 2 and 4). However, many other miRNAs identified by poly(A) RT-PCR in monocytes showed evidence of multiple bands. Among those miRNAs that produced a pair of bands, miR-572 showed the clearest evidence of processing by PIWIL4, as shown in the bottom half of Fig. 5C. The miR-572 miRNA is derived from the 3′-end of the pre-572 RNA hairpin, and cleavage by DICER would normally produce an miRNA roughly 19 bp in length, corresponding to the smaller band visible in C. However, the larger band visible in C corresponds to a 32-bp RNA molecule. Small RNA molecules in the range of 26–33 bp are typically classified as piRNAs. These results suggest that the pre-572 hairpin can possibly be processed by both DICER and PIWIL4.

Because the presence of multiple bands makes it difficult to perform qPCR analysis on the products of pre-572, we obtained densitometry data for these bands (Fig. 5C, bottom). Consistent with the previous trend seen in Table 2, miRNA processing is suppressed in infected MDMs because the intensity of the miRNA band is comparatively weaker in the infected cells than in untransfected cells (compare lane 1 with lane 3 and lane 2 with lane 4).

Finally, we performed a similar experiment in U937 cells with PMA, pNL4-3, and si-PIWIL4, but the cells rapidly underwent apoptosis (data not shown). However, we were able to transfect U937 cells with si-PIWIL4 and with PMA and observed a dramatic decrease in the 32-bp piRNA and a slight increase in the 19-bp miRNA (Fig. 5D). These results therefore provide validation for our earlier conclusions that 1) miRNAs are produced in both monocytes and MDMs using proteins other than DICER and 2) that miRNA biogenesis is impaired in MDMs when transfected with an HIV-1 clone.