Cellular Localization of Latent Murine Cytomegalovirus

Animals.

Three-week-old, female, specific-pathogen-free BALB/c mice and pregnant BALB/c mice with 15- to 17-day-old embryos were purchased from Jackson Laboratory, Bar Harbor, Maine. Mice were maintained in isolation cages and fed and watered ad libitum. This study protocol was reviewed and approved by the Northwestern University Institutional Animal Care and Use Committee.

Virus and cell culture.

MCMV (Smith strain, ATCC 194-VR) was purchased from the American Type Culture Collection (Rockville, Md.) and propagated in BALB/c mouse embryo fibroblasts (MEFs). Fibroblasts were prepared from 15- to 17-day BALB/c mouse embryos by trypsinization in 0.25% trypsin–0.125% EDTA and passaged in minimal essential medium with Earle’s salts, 10% fetal calf serum (FCS), and antibiotics (penicillin [100 U/ml] and streptomycin [100 μg/ml]) at 37°C in a humidified chamber with 5% CO₂. Large-scale virus propagation was accomplished by infecting 90% confluent monolayers of low-passage (passage 2 to 4) MEFs with virus obtained from clarified supernatants of salivary gland extracts of acutely infected mice. Plaques were visible 11 days after infection, at which time cells and supernatants were harvested, sonicated, and clarified by low-speed centrifugation. Virus was pelleted from the clarified supernatant by centrifugation for 3 h at 18,000 rpm in an SS34 rotor, resuspended in culture medium, sonicated, titered, aliquoted, and stored at −70°C. Titers of 10⁸ to 10⁹ PFU/ml were typically obtained.

Virus infection and establishment of MCMV latency.

Latently infected BALB/c mice were established by intraperitoneal injection with 10⁵ PFU of MCMV purified from infected MEFs. Animals surviving acute infection were maintained for 6 months and analyzed for the presence of serum anti-MCMV antibodies, MCMV DNA, and MCMV IE-1 transcripts. Acutely infected and uninfected animals were maintained as positive and negative controls, respectively.

Plaque assays.

The plaque assay included centrifugal enhancement of infection and omission of the semisolid overlay as previously described (4). MEF cells (2 × 10⁵) were seeded into six-well tissue culture plates (Costar, Cambridge, Mass.) and incubated overnight. Tissue samples obtained from mice ≥6 months after initial MCMV infection were minced and sonicated on ice with an ultrasonic homogenizer (CP-710; Cole Palmer, Chicago, Ill.) for 30 s and stored on ice; 1:10 dilutions of undiluted clarified homogenates from each organ were overlaid on confluent MEF monolayers. Following an initial medium change after 24 h, the cells were maintained in Dulbecco modified Eagle medium with 5% FCS and observed by inverted-phase microscopy (4 to 6 weeks) for cytopathic effects characteristic of CMV infection.

Reactivation of latent virus by spleen explant culture.

Spleens were removed from latently infected mice and minced into 1- by 1-mm fragments. Six fragments were seeded into 75-cm² dishes containing 90% confluent MEFs in 15 ml of Dulbecco modified Eagle medium supplemented with 10% FCS and 1% antibiotics and incubated at 37°C in a humidified chamber with 5% CO₂. Medium supernatant was exchanged and screened for MCMV DNA every fifth day by PCR and Southern blot hybridization.

PCR.

DNA was extracted from tissues of acutely infected, latently infected, or uninfected animals. In initial experiments, MCMV DNA from the IE-1 region was amplified as previously described (1) with primers CH16 and CH17 (18) or MCMV1 and MCMV2 (1) to generate PCR products of 310 or 276 bp, respectively. In later experiments, MCMV DNA from the IE-1 region was amplified by nested PCR. In the first round of amplification, 1 μg of DNA was amplified in standard buffer conditions with 40 pmol of sense primer SY1 (5′TGGATGAGAACCGTGTCTAC3′) and antisense primer SY2 (5′ATATCATCTTGCGTTGTCTT3′) for 20 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s to generate a 549-bp product from the MCMV immediate-early region (25) in a reaction volume of 100 μl; 1 μl of this PCR product was reamplified under the same conditions for 30 cycles with internal primers CH17 and CH16 to generate a 310-bp product. β-Actin DNA was amplified by PCR protocol A as described below or PCR protocol B (40 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s). PCR products were electrophoretically separated on 2% agarose gels, transferred to nitrocellulose with a Turboblotter system (Schleicher & Schuell, Keene, N.H.), hybridized to a 3′-labeled oligonucleotide CH15 probe (18), and detected by enhanced chemiluminescence (Amersham [Arlington Heights, Ill.] ECL 3′ oligolabeling and detection system).

Reverse transcription-PCR (RT-PCR).

RNA was extracted from tissues following homogenization in guanidine isothiocyanate. Total cellular RNA was isolated as described previously (9), using cesium trifluoroacetate instead of cesium chloride. Typically, equal amounts of total RNA (1.5 μg) were used to synthesize first-strand cDNA with oligo(dT) primers and Moloney murine leukemia virus reverse transcriptase (GIBCO-BRL, Gaithersburg, Md.) as described previously (18). An aliquot of cDNA was used to amplify β-actin with primers BA-1 and BA-2 (18), and the yield of β-actin was used to subject comparable amounts of cDNA to amplification of MCMV IE-1 RNA. All PCRs were subjected to the hot-start method (10) to minimize primer-primer annealing. After initial denaturation at 94°C for 3 min, MCMV IE-1 cDNA was amplified with primers CH16 and CH17 by PCR protocol A: 35 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 45 s, and extension at 72°C for 60 s, followed by a final 10-min extension at 72°C. Predicted sizes of amplimers are 170 bp for β-actin and 310 and 188 bp for MCMV IE-1 DNA and RNA, respectively. PCR products were analyzed as described above.

Detection of MCMV-specific antibodies.

Immunofluorescent detection of anti-MCMV immunoglobulin was performed on animal serum extracted by the standard clot retraction/centrifugation technique (2). The serum was saved undiluted and as a 1:10 dilution in phosphate-buffered saline (PBS) with 0.1% (vol/vol) Tween 20 for analysis. Second-passage MEFs were infected with MCMV (0.1 PFU/cell) as a source of MCMV antigen. Once the MEFs displayed the cytopathic effects of infection, they were mounted, as cytospins, for assay. The cytospins were fixed in acetone, hydrated through graded alcohols, and blocked with nonimmune goat serum, and test serum was aliquoted over the cells and incubated at 37°C for 30 min. Anti-MCMV immunoglobulin G was detected by a fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin G monoclonal antibody (Fisher Biotech, Pittsburgh, Pa.). Negative controls included uninfected MEFs, the omission of test serum, and appropriate isotype controls. Known anti-MCMV serum was used as a positive control.

Tissue processing.

Animals were anesthetized by methoxyflurane (Pitman-Moore, Inc., Mundelein, Ill.) inhalation and sacrificed by cervical dislocation. Whole blood was collected immediately by cardiac puncture, and multiple organs and tissues (including lung, liver, heart, kidney, spleen, and bone marrow) were harvested. These tissues were promptly divided and treated for analysis. Portions collected for plaque assay were homogenized and placed in culture with MCMV-permissive MEFs. Samples dedicated to analysis by PISH were preserved with a nonaldehyde, non-cross-linking tissue fixative (Streck Laboratories, Omaha, Neb.) for in situ preservation of intracellular DNA sequences. Fixed tissues were embedded in paraffin, omitting formalin treatment to prevent DNA cross-linking. Bone marrow specimens were decalcified by 3-day treatment with 0.5 M EDTA prior to embedding.

In situ hybridization.

Standard in situ hybridization was performed by using a modification of the protocol of Haase (16). Modifications included the use of a biotinylated oligonucleotide probe, MCMVP (1), which was detected with a streptavidin-alkaline phosphatase conjugate and nitroblue tetrazolium (NBT)–5-bromo-4-chloro-3-indolylphosphate (X-phosphate) (Boehringer Mannheim Corp., Indianapolis, Ind.).

PISH.

For PISH analysis, 4-μm sections were cut onto a silanized slide (coated with a 3% Elmer’s Glue solution) with a clean, disposable microtome blade. The samples were deparaffinized in xylene and hydrated through graded alcohols. After hydration the samples were washed with PBS and treated with 2.5% bovine serum albumin in reagent-grade water for 20 min at room temperature to inhibit sequestration of the DNA polymerase on the slide and tissues. Sections were digested in proteinase K (20 μg/ml) in lysis buffer (10 mM Tris [pH 8.3], 50 mM KCl, 2.5 mM MgCl₂, 0.5% Tween 20, 0.5% Nonidet P-40) for 25 min to 1 h at 37°C, and the slide was then heated to 95°C for 2 min to inactivate the proteinase K. The slides were washed in PBS, dehydrated through graded alcohols, and air dried. IE-1-specific MCMV DNA was amplified by a modified PCR protocol (with primers CH16 and CH17) as described in reference 31. Each section was incubated in 40 μl of a preheated (94°C) PCR mixture consisting of 1× PCR Buffer II (Perkin-Elmer, Norwalk, Conn.), 4.5 mM MgCl₂, 0.20 mM dATP, 0.20 mM dCTP, 0.20 mM dGTP, 0.20 mM dTTP–digoxigenin-11-dUTP (9:1) mixture (Boehringer Mannheim), and 1 μM each primer. Digoxigenin-11-dUTP was included in the reaction to anchor the product DNA in situ and prevent positive signal from diffusing into uninfected cells. Subsequently, 1.5 U of Taq DNA polymerase (Perkin-Elmer) in 12.5 μl of reaction mixture was added to initiate the hot-start PCR. Multiple samples were thermocycled by using either a Coy (Detroit, Mich.) TempCycler II Slide Cycler and glass cover slip or the GeneAmp In Situ PCR System 1000 (Perkin-Elmer). The thermocycling profile included 24 cycles of 92°C for 1 min, 56°C for 2 min, and 72°C for 2 min for the denaturing, annealing, and amplification steps, respectively.

Following thermocycling, the sample coverslip or AmpliCover Disc (Perkin-Elmer Cetus, Norwalk, Conn.) was removed, and the PCR mixture was saved for the detection of diffused product DNA. The slides were washed twice for 5 min each time in PBS, and endogenous biotin was blocked in all tissues by using an avidin-biotin blocking kit (Vector Laboratories, Inc., Burlingame, Calif.). Virus-specific product DNA was hybridized in situ with an internally conserved, biotin-labeled (as biotin-dTTP, every 10 to 12 nucleotides) oligonucleotide probe (CH15P [18]). The hybridization buffer (consisting of 50% formamide, 4× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 1× Denhardt’s solution, 0.1% sodium dodecyl sulfate, and 1 mg of salmon sperm DNA per ml, with a probe concentration of 100 ng/ml) was applied to the sections, which were sealed with a Probe-Clip Press-Seal incubation chamber (Sigma Chemical Co., St. Louis, Mo.). The sections were heated to 95°C for 5 min on the Coy Slide Cycler and incubated overnight at room temperature. Following removal of the incubation chamber, the hybridization buffer was saved for the detection of diffused product DNA and DNA-probe hybrids. The sections were washed once with posthybridization wash buffer (2× SSC, 0.01% sodium dodecyl sulfate, 0.1% bovine serum albumin) and twice with PBS for 5 min and then blocked with 10% sheep serum. Hybridized probe was detected with a streptavidin-alkaline phosphatase conjugate (Boehringer Mannheim). The slides were washed twice with PBS and then incubated for 30 min with NBT (337.5 μg/ml)–X-phosphate (175 μg/ml) in alkaline phosphatase buffer (100 mM Tris, 100 mM NaCl, 50 mM MgCl₂ [pH 9.5]) at ambient temperature in the dark. Endogenous alkaline phosphatase was blocked during detection on all sections by the addition of 1× levamisole solution (Zymed Laboratories, Inc., San Diego, Calif.) to the alkaline phosphatase buffer. Slides were then washed in PBS once, counterstained with fast green or aqueous eosin, and mounted by using Aqua-Mount (Lerner Laboratories, Pittsburgh, Pa.). Positive cells yielded a black precipitate.

Controls for the PISH assay included tissue from uninfected mice, serial sections of slides with positive signal (acutely infected tissues) to which no Taq DNA polymerase or primers were added, the use of nonspecific (human CMV [HCMV]) primers or probe (45), and omission of MCMV-specific probe and/or streptavidin-alkaline phosphatase. More than 80 control slides were tested in parallel under these conditions, and positive cells were not detected in control sections. Sections in which MCMV-positive cells were found were confirmed as positive by examining more than 20 sections. The postamplification PCR mixtures and hybridization buffers were analyzed for diffusion of oligonucleotides by agarose gel electrophoresis and ethidium bromide staining.

Immunohistochemistry.

Immunohistochemistry was performed on the tissue section immediately adjacent (contiguous) to the section found positive for the latent MCMV IE-1 DNA by PISH. Sections were deparaffinized and rehydrated through graded alcohols, washed in PBS, and blocked with serum corresponding to the species in which the primary antibody was raised. The following primary antibodies were used at a dilution of 1:100: rat anti-mouse CD31 (PECAM-1, clone MEC 13.3), specific for endothelial cells (ECs) (55); and rat anti-mouse Mac-3 (clone M3/84), specific for macrophages (PharMingen, San Diego, Calif.) (19, 44). The primary antibodies were detected with biotinylated mouse anti-rat κ light-chain secondary antibody (PharMingen) and a streptavidin-fluorescein isothiocyanate conjugate, or alkaline phosphatase in the case of light microscopic analysis. The isotype control antibodies were purchased from PharMingen. Sections were counterstained with methyl green and eriochrome black and mounted with Aquamount containing 10% triethylenediamine (Sigma). Light microscopic sections were counterstained with metanil yellow (Sigma). Control staining included omission of primary antibody, isotype control antibodies, and identically processed, positive control tissue sections of thymus, lymph nodes, and spleen. Hematoxylin and eosin staining of slides for histological comparison was performed by the standard protocol (11). All images of gels, in situ hybridization, PISH, and immunofluorescence were scanned by using Adobe Photoshop.

Virus infection and establishment of MCMV latency in animal subjects.

Latency for MCMV has been defined as the inability to detect replicating virus in target organs such as salivary gland or spleen 6 months after acute infection (24). Latently infected BALB/c mice were established by intraperitoneal injection of 10⁵ PFU of MCMV (Smith strain), and animals surviving acute infection were maintained for 6 months. All animals were screened for the presence of MCMV-specific antibodies in the serum. Anti-MCMV immunoglobulin was detected in the sera of all infected animals, while uninfected animals were seronegative. Infected animals were also screened for the presence of replicating virus by plaque assays of tissues on permissive MEFs. MCMV plaques were detected by plaque assay of sonicated extracts of organs from acutely infected animals after 6 days in culture but were not detected in extracts from uninfected animals or from mice infected 6 months previously, even after 3 weeks in culture. Mice that had been infected 6 months previously and were seropositive for antibodies against MCMV and negative for viral replication in plaque assays were considered to be latently infected.

Acutely infected, uninfected, and latently infected mice were then screened for the presence of viral DNA and RNA by PCR analysis of nucleic acid extracted from various tissues. MCMV-specific DNA from the IE-1 region was detected in the heart, spleen, kidney, liver, lung, and bone marrow of all of six latently infected animals but was not detected in tissue from uninfected animals. Figure 1 shows a representative gel of MCMV-specific PCR products amplified from various tissues of one latently infected and one uninfected mouse. The specificity of the amplification product was confirmed by Southern blot hybridization (Fig. 1B). MCMV DNA was not detected in the peripheral blood lymphocytes of either uninfected or latently infected animals, although β-actin sequences could be amplified from all samples from both latently infected and uninfected mice (Fig. 1C).

Transcripts from the IE-1 region of the genome are expressed during the early stages of productive infection and, in animals surviving acute infection, are suggestive of persistent rather than latent infection. To determine whether IE-1 transcripts were present in tissues from these mice, RNA extracted from the liver, lung, heart, kidney, and spleen was screened by RT-PCR with oligonucleotide primers specific for the IE-1 region. MCMV IE-1 transcripts were detected in the liver, lung, spleen, kidney, heart, and bone marrow of acutely infected animals. Southern blot hybridization confirmed the specificity of the amplified products. No MCMV IE-1 transcripts were detectable in any organ from six of six latently infected mice or three of three uninfected mice, although β-actin transcripts could be amplified. Figure 2 shows a Southern blot of a representative PCR analysis of DNA and RNA from tissues of an acutely and a latently infected mouse. In this experiment, approximately 5-fold more amplified latent DNA and 10-fold more latent RNA were loaded compared to amplified acute samples (based on the ethidium bromide fluorescence of amplified β-actin bands). In general, the latent samples had 1- to 2-log-fold less IE-1 DNA than the corresponding acute samples.

Latent MCMV can be reactivated in vitro by culturing spleen explants in the presence of permissive MEFs. This virus can be recovered from the media and detected by PCR. To demonstrate viral reactivation from our latently infected mice, spleen explants from five latently infected mice were cultured in the presence of MEFs. Cultures were observed for the presence of plaques, and culture medium was harvested every fifth day and screened for the presence of viral DNA by PCR with primers specific for the IE-1 region. No virus was detected in the media at 11 days after explant, but after 16 days in culture, plaques were visible and viral DNA was detected in the media by PCR and Southern blot hybridization (Fig. 3).

These data demonstrate that we have established latent MCMV infection in our mice. Despite the presence of viral DNA in many organs in these mice, transcripts from the IE-1 region were not observed, indicating that these mice are in fact latently, rather than persistently, infected. Consistent with these observations, shedding of virus into the media was not detected until 16 days after spleen explant culture. The presence of reactivated virus 16 days after explant indicates that the viral DNA remains biologically viable in cells harboring latent virus.

Cellular localization of MCMV DNA in latently infected tissues.

To identify the cell types harboring latent MCMV DNA, we analyzed tissues by standard in situ hybridization without PCR amplification. While MCMV DNA was readily detected in tissues from acutely infected animals, we were unable to detect MCMV DNA in tissues from latently infected animals (Fig. 4), suggesting that the copy number of viral DNA in latently infected cells is below the limits of detection by in situ hybridization. To increase the sensitivity of detection of MCMV DNA, we used PISH analysis on fixed tissue from each organ after having verified the presence of latent MCMV DNA by standard PCR analysis. We have previously demonstrated that this technique is sensitive enough to detect a single copy of intracellular proviral DNA in situ (29, 39, 52).

Using PISH, we were able to detect MCMV DNA in kidney sections of both latently and acutely infected mice (Fig. 5B to E). Kidney sections of uninfected control mice were negative (Fig. 5A). The location and morphology of the PCR-positive cells strongly suggested that these cells were peritubular ECs. To demonstrate the presence of ECs surrounding the kidney tubule, immunofluorescence with anti-PECAM-1 antibody was performed on the tissue section immediately adjacent (4 μm) to the section assayed by PISH. PECAM-1 is constitutively expressed at high levels on all continuous endothelial linings in vivo (55). Staining with PECAM-1 antibody clearly shows the presence of ECs in the location of the PCR-positive cells in the kidney and strongly suggests that these PCR-positive cells are ECs (Fig. 6A). A survey of other organs showed that PISH-positive cells were also present in the liver, spleen, and heart (Fig. 6B to D). PECAM-1 staining of adjacent sections was consistent with their identity as ECs. These observations were reproduced in six of six latently infected animals. PCR-positive cells were also detected in cells of the lung alveolar septum of latently infected mice (Fig. 7). However, cells in the adjacent section in the area of the PCR-positive cells were negative for PECAM-1 expression (data not shown). Immunofluorescent staining with Mac-3, a marker for macrophages (19, 44), indicated that the cells in the region of the PCR-positive cells in the lung were macrophages, and thus macrophages are likely to be the cell type harboring latent MCMV in the lung (Fig. 7).

MCMV DNA was also detected in bone marrow samples from latently infected mice. PISH analysis of both cytospin (Fig. 8A) and axial bone marrow (Fig. 8B) sections revealed positive cells, the identities of which have not been determined.

There was a 100% correlation between PCR organ screening and the detection MCMV DNA by PISH in all tissue samples. Detection of nonspecific product DNA, formed as a result of mispriming and/or polymerase-associated DNA repair, was avoided by the detection of hybridized, target-specific probe. Nonspecific hybridization of the biotinylated oligonucleotide probe was not observed, as shown by lack of staining in uninfected organ sections and sections where either the DNA polymerase was omitted or nonspecific primers were used in the PCR mixture. Digoxigenin-11-dUTP was incorporated into the PCR protocol to prevent diffusion of amplicon DNA into neighboring uninfected cells, and MCMV amplicon sequences were not detected in either the postamplification PCR buffer or the hybridization buffer.