Delayed Spread and Reduction in Virus Titer after Anterior Chamber Inoculation of a Recombinant of HSV-1 Expressing IL-16

Abstract

purpose. The timing of T-cell infiltration of the hypothalamus is crucial in the prevention of bilateral retinitis in mice inoculated with HSV-1 through the anterior chamber (AC). In H129-infected mice, T-cells are recruited to the suprachiasmatic nuclei of the hypothalamus too late to protect infected mice from development of bilateral retinitis. The purpose of these studies was to determine whether alteration of T-cell recruitment to the hypothalamus would affect the timing and pattern of virus spread after AC inoculation.

methods. A recombinant of the H129 strain of HSV-1 expressing IL-16, a cytokine with lymphocytic and monocytic chemoattractant properties, was constructed, and mice were inoculated in the AC with H129wt, H129wt and H129/IL-16, or H129wt and H129/pGal10 (a recombinant virus containing vector only).

results. AC inoculation of BALB/c mice with H129wt and H129/IL-16 resulted in a delay of virus spread to the hypothalamus and the contralateral retina, and this delay correlated with decreased virus titers in infected tissues, compared with mice infected with H129wt or mice infected with H129wt and H129/pGal10. Although the number of infiltrating T-cells in the brains of mice infected with H129wt, H129wt and H129/IL-16, or H129wt and H129/pGal10 was similar, more Mac-1-positive cells were detected early (postinoculation day 2) in the injected eyes of mice infected with H129wt and H129/IL-16 than in mice infected with H129wt and/or H129wt and H129/pGal10.

conclusions. These results suggest that early recruitment of Mac-1-positive cells to the injected eye may play a role in delaying virus spread in mice infected with H129wt and the IL-16-expressing recombinant virus. IL-16 delivery vectors could be exploited to prevent or delay HSV-1 infection of the hypothalamus, allowing development of the antiviral immune response and subsequent inhibition of virus spread into the optic nerve and retina.

Acute retinal necrosis (ARN) syndrome was first described several decades ago. ¹ Although rare, the disease has since been recognized to be a cause of blindness in otherwise healthy individuals throughout the world. ^{2 3 4 5 6 7 8} ARN is characterized by necrotizing retinitis, vitritis, retinal arteritis, and, in some cases, optic neuropathy. ^{9 10 11} Untreated, the disease is rapidly progressive and often leads to vision loss in the affected eye. Herpes simplex virus (HSV) types 1 and 2, varicella-zoster virus, and, to a lesser extent, cytomegalovirus have all been implicated as causes of ARN in humans. ^{2 3 4 5 6 7 8}  

In a mouse model of ARN, anterior chamber (AC) inoculation of BALB/c mice with the KOS strain of HSV-1 results in retinal necrosis in the uninoculated eye only. ^{12 13 14} Although the AC of the injected eye is infected, the retina is spared. By contrast, AC inoculation of BALB/c mice with H129, a neuroinvasive and neurovirulent strain of HSV-1, results in bilateral retinal necrosis. ¹⁵ After AC inoculation, both strains of virus spread retrogradely through the central nervous system (CNS) to infect the suprachiasmatic nuclei (SCN) of the hypothalamus ^{13 15} from which virus can then spread to the retina through the optic nerve. However, although both H129 and KOS are able to spread from the ipsilateral SCN to the retina of the contralateral eye, only H129 is also able to spread from the contralateral SCN to infect the retina of the injected, ipsilateral eye. ^{13 15}  

Recent investigations have shown that the timing of virus entry and T-lymphocyte infiltration of the SCN differ for the two strains of virus and that these differences appear to determine whether bilateral or unilateral retinitis develops in the infected mice. ^{16 17} H129 is detected in both the ipsilateral and contralateral SCN at postinoculation (PI) day 4, whereas KOS is not detected in the ipsilateral SCN until PI day 5 nor in the contralateral SCN until PI day 7. ^{13 15} In H129-infected mice, infiltration of the SCN by T-cells and cytokine production are not observed until PI day 5, by which time virus has already spread to the retina of both eyes. ¹⁶ By contrast, in KOS-infected mice, T-cells are detected in the ipsilateral SCN concomitant with virus and in the contralateral SCN 2 days before virus. ¹⁷ Several studies support the idea that T-cells in the contralateral SCN of KOS-infected mice limit the spread of virus into the optic nerve of the injected eye thereby sparing the retina of that eye from virus infection. Bilateral retinitis develops in thymectomized, CD4⁺ and CD8⁺ T-cell-depleted mice after AC inoculation of KOS, whereas it does not develop in mice with depletion of only CD4⁺ or CD8⁺ T cells. ¹⁸ Furthermore, in thymectomized, CD4⁺ and CD8⁺ T-cell-depleted mice, virus is detected in the ipsilateral SCN a day earlier than in normal, nondepleted mice and in the contralateral SCN 2 days earlier. ¹⁹ Adoptive transfer of KOS-specific immune effector T cells protects mice from retinitis after AC inoculation by preventing virus spread into the SCN. ²⁰  

Because H129-infected mice have a delay in T-cell recruitment to the SCN and develop bilateral retinitis after AC inoculation, it was hypothesized that if T-cell recruitment to the hypothalamus was altered, the timing and/or pattern of virus spread would be affected. To test this hypothesis, a recombinant of H129 expressing IL-16 was constructed. IL-16, formerly lymphocyte chemoattractant factor or LCF, ^{21 22} is a proinflammatory cytokine that is chemotactic for all CD4⁺ lymphoid cells, including CD4⁺ T-lymphocytes, monocytes, eosinophils, and dendritic cells. ^{23 24 25 26} IL-16 induces expression of IL-2 receptor and major histocompatibility class (MHC) class II in human CD4⁺ T cells and increases intracellular Ca²⁺ and inositol-1,4,5-triphosphate in these cells. ^{24 27} Sources of IL-16 include, but are not limited to, lymphocytes, macrophages, eosinophils, dendritic cells, and mast cells. ²⁸ IL-16 is synthesized as a precursor molecule (pro-IL-16) composed of approximately 631 amino acid residues. ²⁹ In CD8⁺ T-cells, pro-IL-16 is cleaved by caspase-3 to produce functional IL-16. ³⁰  

Although the role of IL-16 in HSV-1 infection is not known, recombinant IL-16 has been shown to repress HIV-1 replication in peripheral blood mononuclear cells (PBMCs), macrophages, and dendritic cells. ^{31 32 33} In addition, human CD4⁺ cells transfected with IL-16 cDNA are resistant to HIV infection. ³⁴ Resistance is believed to occur at the level of mRNA expression and most likely involves inhibition of HIV-1 promoter activity. ^{34 35} The sequence and general structure of IL-16 are conserved across species. Both the human and mouse IL-16 genes comprise seven exons and six introns. ³⁶ Furthermore, human and mouse IL-16 display 82.1% homology at the C-terminal region, thus exhibiting conservation in both structure and function. ^{37 38}  

The purpose of these studies was to determine whether expression of IL-16 by HSV-1 strain H129 would affect the route of spread, timing, and/or extent of infection after uniocular AC inoculation of BALB/c mice. Studies were also conducted to determine the pattern and timing of inflammatory cell infiltration in animals infected with H129wt, with H129wt, and an IL-16 expressing recombinant virus (H129/IL-16) or with H129wt and a recombinant virus containing the vector only (H129/pGal10). 

Materials and Methods

Viruses

H129, a neurovirulent strain of HSV-1 (a gift from Richard D. Dix, PhD, University of Miami School of Medicine, Miami, FL) was originally isolated from a patient with encephalitis. ³⁹ Stocks of all viruses used in this study were propagated on Vero cells (ATCC, Manassas, VA) grown in complete Dulbecco’s modified Eagle’s medium (DMEM) containing 5% fetal bovine serum (FBS) and antibiotics. The titer of virus stocks was determined by standard plaque assay on Vero cells and titers were expressed as plaque-forming units (PFU) per milliliter. Stocks were stored at −70°C in 1-mL aliquots, and a fresh aliquot of stock virus was thawed and used for each experiment. The titers of virus stocks used in these studies were 1 × 10⁸ PFU/mL (H129/IL-16), 2.25 × 10⁸ PFU/mL (H129wt), and 3.87 × 10⁸ PFU/mL (H129/pGal10). 

Plasmids

pRc/CMV-IL16, a plasmid that contains an open reading frame of 390 base pairs and that codes for the C-terminal (functional) region of the human IL-16 gene, was a gift from Paul Zhou, PhD (Southwest Foundation for Biomedical Research, San Antonio, TX). The IL-16 cDNA was excised from pRc/CMV-IL16 by restriction enzyme digestion. HindIII and SalI restriction sites were added to the cDNA by using PCR methods and the primers that are listed at the end of the paragraph. The cDNA was then cloned into the vector pGal10 (a gift from Jerry Weir, PhD, U.S. Food and Drug Administration, Bethesda, MD) at the HindIII and SalI site to generate the vector pGal10/IL-16. In this vector, the IL-16 gene is located downstream from the HSV immediate early 110K promoter. The pGal10 vector was designed specifically for the generation of recombinant HSVs. ⁴⁰ It contains the lacZ gene under the HSV gC promoter and provides a marker for selecting recombinants. The vector also contains flanking thymidine kinase (tk) sequences to direct homologous recombination. However, because recombination disrupts the HSV genomic tk locus, all recombinants generated with this vector are TK deficient (tk ⁻). As a consequence, recombinants cannot replicate in cells that have no cellular tk. Primers (Genosys, The Woodlands, TX) used for generating pGal10/IL-16 were as follows: IL16-forward, 5′-GTCGACCATGCCTGACCTCAACTCCT-3′, which has a SalI site; and IL16-reverse, 5′-AAGCTTCTAGGAGTCTCCAGCAGCTG-3′, which has a HindIII site. 

Generation of H129/IL-16

Recombinant viruses were generated from H129wt. H129 virus was propagated in Vero cells. H129 virions were isolated by centrifugation through a sucrose gradient, and DNA was extracted from purified virions. pGal10/IL-16 DNA was amplified by transforming Escherichia coli HB109 cells (Life Technologies, Rockville, MD). Plasmid DNA was purified using a plasmid extraction kit (Wizard; Promega, Madison, WI) and digested with EcoRI (Promega). Vero cells were then cotransfected with H129 and digested plasmid DNA by calcium phosphate precipitation (Invitrogen, Carlsbad, CA) or by using a lipophilic transfection agent (Lipofectin; Life Technologies). To screen for recombinants, transfectants were incubated with complete DMEM containing 5% fetal bovine serum, antibiotics, 1% methylcellulose (Sigma-Aldrich, St. Louis, MO), 18 μg/mL acyclovir (Sigma-Aldrich), and 250 μg/mL X-gal (Life Technologies) at 37°C in 5% CO₂ for several days. Blue plaques indicative of recombinant viruses were selected and plaque purified by three or four passages through cell culture before growing to high titer. Recombinants were then assayed for the presence of the IL-16 insert by Southern blot hybridization and for production of IL-16 by ELISA. In the H129/IL-16 recombinant, IL-16 was expressed as an immediate early gene product. 

Generation of H129/pGal10

pGal10 plasmid DNA alone, without IL-16 cDNA, was amplified in HB109 cells by the process described earlier. Vero cells were then transfected with H129 and EcoRI-digested pGal10 DNA, and recombinants were screened and purified using the procedures described in the prior section. 

Southern Blot Hybridization

DNA was isolated from recombinant and wild-type viruses, as described herein. Viral DNA and pGal10/IL-16 plasmid DNA were digested with SalI and HindIII. DNA from pRc-CMV/IL-16 was digested with HindIII. Digested DNAs were then electrophoretically separated on a 1% agarose gel containing ethidium bromide and visualized under a UV lamp. The DNA was then transferred to a nylon membrane (Nytran Supercharge; Schleicher & Schuell, Keene, NH), using a rapid downward transfer system (Schleicher & Schuell) under neutral conditions. The blot was hybridized with IL-16 DNA probe that had been labeled with digoxigenin (DIG) using a kit (DIG-High Prime; Roche Applied Science, Indianapolis, IN). The membrane was then reacted with anti-Dig alkaline phosphatase antibody (Roche Applied Science) and visualized colorimetrically with BCP/NBP (Roche Applied Science). 

Production of IL-16 in Cultured Cells

Vero cells were infected with H129/IL-16, H129wt, or H129/pGal10 at a multiplicity of infection (MOI) of 0.01 PFU and incubated at 37°C in 5% CO₂. At 4, 8, 12, 18, 24, and 48 hours after infection. Infected cell supernatant was collected and stored at −80°C. The supernatant was then assayed for IL-16 using a anti-human IL-16 ELISA kit (Cytoscreen; Biosource, Camarillo, CA). The results were read by plate reader (AT 400; SLT, Hillsborough, NC) and plate reader software (DeltaSoft, Princeton, NJ). Uninfected cell supernatant was used as the negative control. The results were plotted graphically on computer (DeltaGraph; DeltaPoint, Inc., Monterey, CA). 

In Vitro Single-Cycle Growth Kinetics of Viruses

Cell-culture dishes (35 mm) were seeded with 2 × 10⁶ Vero cells/dish and incubated overnight at 37°C under 5% CO₂. Cells were then inoculated with H129wt, H129/IL-16, or H129/pGal10 at an MOI of 5 PFU/cell and incubated at 37°C under 5% CO₂ for 1 hour to allow virus to adsorb. Plates were then washed with PBS to remove any unadsorbed virus, overlaid with 2 mL DMEM containing 5% FBS and antibiotics, and incubated as described earlier. Samples were collected at 0, 2, 4,6, 12, and 24 hours postinfection and stored at −80°C. Samples were then subjected to several cycles of freeze-thaw and centrifuged. The titer of virus in the supernatants was determined by plaque assay on duplicate cultures of Vero cells. 

Animals

Adult female BALB/c mice 8 to 12 weeks old (Taconic, Germantown, NY) were used in all experiments. The mice were housed in accordance with National Institutes of Health guidelines, and all study procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. All mice were maintained on a 12-hour light-dark cycle and were given unrestricted access to food and water. For all intraocular injections and perfusions, mice were anesthetized intramuscularly with a mixture of 0.08 mg xylazine and 0.9 mg ketamine per 25 g body mass. 

Inoculations

Mice were anesthetized and inoculated through the AC route as follows: The right eye was proptosed, aqueous humor was removed by paracentesis, and 2 μL containing 5 × 10³ PFU of H129wt virus or 2 μL containing a mixture of 5 × 10³ PFU of H129wt virus and 1 × 10⁵ PFU of either H129/IL-16 or H129/pGal10 was injected into the AC with a 30-gauge needle attached to a 100-μL microsyringe (Hamilton, Reno, NV). The inoculum was prepared by diluting virus stock in DMEM containing antibiotics. Control mice were inoculated in the AC with 2 μL of either phosphate-buffered saline (PBS) or DMEM from mock-infected cells. 

Perfusions

Mice were deeply anesthetized and perfused transcardially with PBS for approximately 3 minutes. After perfusion, the brain, both eyes, and both trigeminal ganglia were quickly removed, snap frozen on dry ice, and stored at −80°C. 

Plaque Assays

The brain, eyes, and trigeminal ganglia from mice infected with H129wt only, H129wt and H129/IL-16, or H129wt and H129/pGal10 were homogenized separately in 500 μL of DMEM containing antibiotics. Each homogenate was serially diluted and plated on Vero cells that were 80% confluent. Adsorption of virus was performed for 1 hour at 37°C in a CO₂ incubator. After adsorption, the cells were overlaid with a 1:1 solution of 2× DMEM containing 10% serum and antibiotics and 1% low-melting-point agarose (Life Technologies). After 5 days at 37°C, the cells were fixed with 0.5% glutaraldehyde (Sigma-Aldrich) and incubated overnight with X-gal (Life Technologies). Blue and clear plaques were counted, and the titer of virus (total PFU per sample) was calculated and plotted on computer (DeltaGraph; DeltaPoint, Inc.). 

Immunohistochemistry

Virus, T Cells, and IL-16 Single Labeling.

Serial frozen 8- to 10-μm sections were collected on positively charged slides (Fisher Scientific, Pittsburgh, PA). Sections for immunohistochemistry were fixed with acetone and blocked with normal goat serum (Vector Laboratories, Burlingame, CA). Sections to be stained for T cells or IL-16 were also blocked against endogenous biotin, with an avidin-biotin blocking kit (ABC; Vector Laboratories) used according to the manufacturer’s instructions. Sections were then incubated with (1) a rabbit anti-HSV-1 polyclonal antibody (Accurate Chemicals, Westbury, NY); (2) a hamster anti-mouse α/β TCR monoclonal antibody (Clone H57-597; BD PharMingen, San Diego, CA); (3) a biotinylated rat anti-mouse CD4 monoclonal antibody (clone RM4-5; BD PharMingen); (4) a biotinylated rat, anti-mouse CD8 monoclonal antibody (Clone 53-6.7; BD PharMingen); (5) biotinylated rabbit, anti-human IL-16 polyclonal antibody (Biosource); or (6) FITC-labeled anti-CD11b (clone M1/70, Mac-1 Ab; BD PharMingen). After incubation with the CD11b primary antibody, labeled sections were washed, mounted with anti-fade medium containing 4′,6′-diamino-2-phenylindole (DAPI; VectaShield; Vector Laboratories) and examined under a fluorescent microscope for CD11-positive, fluorescent cells. After incubation with all other primary antibodies, sections were washed with PBS several times and incubated with the appropriate biotinylated secondary antibody, if necessary. Sections were then reacted with 0.3% hydrogen peroxide to eliminate endogenous peroxidase, washed several times, and reacted with avidin-biotin solution (Vector Laboratories). Diaminobenzidine (DAB; Sigma-Aldrich) was used as the chromogen for sections stained for HSV-1. DAB reaction was intensified by adding 0.04% nickel chloride (Sigma-Aldrich) to the DAB solution. Sections stained for T cells and IL-16 were reacted with purple dye and peroxidase substrates, respectively (Vector VIP Purple and Vector SG, respectively; Vector Laboratories). All sections (except those stained for IL-16) were then counterstained with methyl green (Sigma-Aldrich), dehydrated in a graded ethanol series, cleared with xylene, coverslipped, and examined microscopically for the purple-black-stained cells indicative of virus infection or purple-stained T lymphocytes. IL-16 sections were counterstained with nuclear red dye (Nuclear Fast Red; Vector Laboratories), processed as described earlier, and examined microscopically for gray-stained IL-16-expressing cells. 

Virus, T Cells, and IL-16-Double Labeling.

Frozen sections were processed as described in the prior section and incubated with either biotinylated or nonbiotinylated rabbit anti-human IL-16 (Biosource). Sections were washed with PBS and incubated with avidin D-Texas red or 7-amino-4-methylcoumarin-3-acetic acid (AMCA)-labeled-anti-rabbit IgG (Vector Laboratories) and washed again. Sections were then incubated overnight with one of the following second primary antibodies: (1) FITC-labeled anti-CD11b, (2) FITC-labeled anti-CD8a, (3) FITC-labeled anti-α/β TCR (all from BD PharMingen), or (4) FITC-labeled anti-HSV-1 (Dako, Carpinteria, CA) for 1 hour at 37°C. Sections were washed, mounted with antifade medium containing DAPI (VectaShield; Vector Laboratories), and examined with a fluorescence microscope for red IL-16-expressing cells; green CD11b-, CD8-, α/β TCR-, or HSV-1-positive cells, and orange or blue-green cells coexpressing IL-16 and any of the listed antigens. 

β-Galactosidase Staining

Frozen sections were prepared as described earlier. Sections adjacent to those used for immunohistochemistry were fixed with glutaraldehyde and stained with the β-galactosidase substrate X-gal, using previously reported methods. ¹³ Sections were then counterstained with Safranin-O (Sigma-Aldrich), dehydrated, cleared with xylene, coverslipped, and examined microscopically for blue-stained areas indicating sites of recombinant virus infection. 

Results

Construction of Recombinant Viruses

Recombinant HSV-1 (H129/IL-16) was constructed by first inserting the cDNA coding for functional IL-16 into pGal10, a vector that has previously been used to generate HSV-1 recombinants expressing the cytokine IFN-α. ⁴⁰ The resultant plasmid, pGal10-IL-16, was then used to generate H129/IL-16 (Fig. 1A) , as described in the Methods section. To control for the effect of the pGal10 vector in these studies, a recombinant H129 virus containing the pGal10 vector alone, H129/pGal10 (Fig. 1B) , was also constructed. 

To confirm the presence of IL-16 cDNA in H129/IL-16, viral DNAs were isolated and digested with HindIII and SalI and evaluated by Southern blot hybridization. As expected, IL-16 DNA sequences were detected only in the vectors in which IL-16 cDNA had been inserted (i.e., pRc-CMV-IL-16 and pGal10-IL-16) and in the recombinant virus, H129/IL-16 (Fig. 2) . IL-16 cDNA was not detected in wild-type virus or in recombinant virus containing vector alone (H129/pGal10). 

To determine whether cells infected with H129/IL-16 expressed IL-16, ELISA was performed on infected Vero cell supernatants, by using an anti-human IL-16 ELISA kit (Biosource). IL-16 was detected as early as 8 hours after infection. The highest level of IL-16 was 15 ng per 5 × 10⁶ PFU/mL at 48 hours after infection (Fig. 3) . IL-16 was not detected in supernatants of Vero cells infected with wild-type virus or H129/pGal10. 

In Vitro Growth Kinetics of Viruses

To assay the in vitro replication of H129wt, H129/IL-16, and H129/pGal10, single-cycle growth experiments were performed. As shown in Figure 4 , the kinetics of viral replication were similar for all viruses. 

Complementation of Recombinant Viruses with Wild-Type Virus

Construction of HSV-1 recombinants using the pGal10 vector involved homologous recombination at the HSV tk locus. All recombinants generated using this vector were tk ⁻ and to replicate in cells without tk, such as neurons, H129/IL-16 and H129/pGal10 were complemented with wild-type virus, as previously described. ^{13 41}  

To determine the ratio of wild-type to recombinant virus with the highest complementation, mixtures of H129/IL-16 and H129wt or H129/pGal10 and H129wt were prepared at ratios of 1:1 to 20:1 (recombinant/wild type). Two groups of BALB/c mice were then inoculated with a mixture of wild-type and recombinant virus in one AC. Control mice were inoculated with 5 × 10³ PFU of H129wt, 1 × 10⁵ PFU of H129/IL-16, or 1 × 10⁵ PFU of H129/pgal10. On postinfection (PI) days 6 and 8, the contralateral eye was removed, and a standard plaque assay was performed on Vero cells, using homogenates of the contralateral eyes. Plaques were stained with X-gal to differentiate blue (recombinant) from clear (wild type) plaques. Although the AC of the inoculated eye of mice injected with H129/IL-16 or H129/pGal10 was infected, virus was not recovered from the contralateral eye of these mice. In contrast, high titers of virus (>1 × 10⁶ PFU/mL at PI day 8) were recovered from the contralateral eye of mice infected with H129wt alone. Virus was also recovered from the contralateral eye of mice infected with a mixture of H129wt and H129/IL-16 or H129wt and H129/pGal10. The ratio of recombinant to wild-type virus that resulted in the highest titer of both viruses was 20:1. At this ratio, an average of 1.43 × 10⁶ PFU/mL of H129/IL-16 and 6.40 × 10⁶ PFU/mL of H129wt was recovered from the uninoculated eye at PI day 8 The ratio of 20:1 (recombinant/wild type) was therefore used for all studies with recombinant viruses. 

Expression of Viral Antigen and Recovery of Virus in the Brain

After AC inoculation with H129wt, virus spreads through the brain to infect the retina of both the inoculated and uninoculated eye. ¹⁵ Therefore, the titer of virus in the brain of mice infected with recombinant viruses was compared with the titer of virus recovered from mice infected with H129wt to determine whether infection with H129/IL-16 affected the spread of virus into the brain. As shown in Figure 5A , the titer of virus recovered from the ipsilateral half of the brain of mice infected with H129wt and H129/IL-16 at PI day 4 was significantly lower (P < 0.05) than that recovered from mice infected with either H129wt or with H129wt and H129/pGal10. Although less virus was also recovered from the contralateral half of the brain of mice infected with H129wt and H129/IL-16 than from mice infected with H129wt, the difference between the titer of virus recovered from the contralateral brain of H129/IL-16- and H129/pGal10-infected mice was not significant (not shown). 

Compared with mice infected with H129wt or with H129wt and H129/pGal10, virus infection of the SCN was delayed in mice infected with H129wt and H129/IL-16. At PI day 4, virus was observed in the ipsilateral SCN of three of nine H129/IL-16-infected mice, compared with four of six mice infected with H129/pGal10 and six of six mice infected with H129wt (Fig. 5B) . In the contralateral SCN, at PI day 4, virus was detected in one of nine mice infected with H129/IL-16 compared with three of six pGal10-infected mice and five of six of H129wt-infected mice (not shown). At PI day 5, less viral antigen was detected in both the ipsilateral and contralateral SCN of mice infected with H129/IL-16 compared with mice infected with H129wt or H129/pGal10 (Fig. 6) . 

Expression of Viral Antigen and Recovery of Virus in the Eye

To determine whether delayed virus infection of the SCN and the decrease of virus titer in the brain affects the spread of virus into the eyes, virus recovery studies were performed on the eyes of mice inoculated with H129wt, H129wt, and H129/IL-16, or H129wt and H129/pGal10. Although the titer of virus recovered from the injected eye was similar for all three viruses at all time points (not shown), at PI day 6, the titer of virus recovered from the contralateral eye of H129/IL-16-infected mice was significantly lower (P < 0.05, Student-Newman-Keuls test) than from mice infected with H129wt (Fig. 7A) . This difference appeared to correlate with the presence of IL-16; the titer of virus recovered from the contralateral eye of mice infected with H129/IL-16 was also significantly lower (P < 0.05) than the titer recovered from mice infected with H129/pGal10 (Fig. 7A) . 

Irrespective of the infecting virus, the ipsilateral retina of mice became infected at PI day 5 (not shown). However, virus infection of the contralateral retina was delayed in H129/IL-16-infected mice. Virus was observed in the contralateral retina of mice infected with either H129wt or H129/pGal10 at PI day 5 (Figs. 7B 8B 8C) , whereas in mice infected with H129/IL-16, virus was not detected in the contralateral retina until PI day 6 (Figs. 7B 8A) . Although the contralateral retina of H129/pGal10-infected mice was infected at PI day 5, less viral antigen was detected than in the retina of mice infected with wild-type virus (compare Figs. 8B and 8C ). 

Expression of IL-16 in the Eyes and Brain

In the injected eyes of H129/IL-16-infected mice, IL-16 was observed mainly in the anterior chamber along the iris and ciliary body at PI day 2 (Fig. 9A) A few IL-16-positive cells were also observed in the iris and ciliary body of H129wt- and H129/pGal10-infected mice at PI day 2. However, the expression of IL-16 in the injected eyes of H129wt- and H129/pGal10-infected mice was less than that observed in H129/IL-16-infected mice (compare Fig. 9A with Figs. 9B 9C ). 

To determine whether IL-16 was expressed in the brain of H129/IL-16-infected mice, brain sections adjacent to sections that contained viral antigen were stained for IL-16. The staining pattern for IL-16 correlated with the pattern of viral antigen expression (Figs. 9D 9E) , suggesting that some or all of the IL-16 was produced by virus-infected neurons. To confirm this observation, H129/IL-16-infected brain sections were stained for IL-16 and HSV-1, for IL-16 and α/β T cells, for IL-16 and CD8⁺ T cells, or for IL-16- and Mac-1-positive cells. Although a few CD8⁺ T cells, α/β T cells, and Mac-1-positive cells colabeled with IL-16 antigen (not shown), the majority of cells expressing IL-16 in the brain appeared to be infected neurons (Figs. 9D 9E 9F) . However, not all infected neuronal cells expressed IL-16. 

Brain sections from H129/pGal10- and H129wt-infected mice were also immunolabeled for IL-16. IL-16 was also detected in the brain of these mice, and the staining pattern for IL-16 correlated with the pattern of viral antigen expression as well. However, IL-16 expression in H129/pGal10- and H129wt-infected mice was less extensive than that observed in the brain of H129/IL-16-infected mice (not shown). Double immunofluorescent labeling of brain tissues from H129/pGal10- and H129wt-infected mice showed that infected neurons as well as T cells and Mac-1-positive cells were sources of IL-16 in the brain of these mice (not shown). IL-16 was not detected in the brain of mock-infected mice. 

Inflammatory Cell Infiltration

Because IL-16 is a chemoattractant for several types of immune cells, including CD4⁺ T cells, monocytes, dendritic cells and eosinophils, sections from the brain and inoculated eyes of H129wt-, H129wt- and H129/IL-16-, or H129wt- and H129/pGal10-infected mice were immunolabeled for CD4⁺ T cells, α/β T cells, CD8⁺ T cells, and Mac-1-positive cells to determine whether IL-16 affects recruitment of these cells to infected tissues. Results from H129wt- and H129/IL-16-infected mice were compared with results from mice infected with H129wt or with H129wt and H129/pGal10. At PI days 2 and 3, only a few T cells (three to five mice) were detected in the AC, and the vitreous of the injected eyes of mice inoculated with H129wt or with either of the two recombinant viruses (not shown). However, more Mac-1-positive cells were detected in the injected eyes of mice infected with H129/IL-16 at PI day 2 compared with the injected eyes of H129/pGal10- or H129wt-infected mice (Figs. 10A 10B 10C) . Mac-1-positive infiltrates were located primarily in the AC and the vitreous and the ganglion cell layer. 

Early after inoculation (PI days 2–4), T cells were detected in the brain mainly in the root entry zone of the oculomotor nerve and the ventricles, irrespective of the virus type. At this time, the number of infiltrating T lymphocytes in the brain of mice infected with either wild-type or recombinant viruses was similar (not shown). At PI day 6, many infiltrating T-cells were observed in the brain of H129/IL-16, especially the mesencephalon, when the titer of recombinant virus in the brain was also the highest and many β-gal-positive cells were observed (not shown). Because there was an abundance of viral antigen and T-cells at PI day 6, ¹⁶ it was not possible to observe a difference between mice infected with wild-type virus and mice infected with recombinant virus. However, a difference was observed in the number of Mac-1-positive cells in the SCN at PI day 6. There were more Mac-1-positive cells in the SCN of H129/IL-16-infected mice than in the SCN of mice infected with H129/pGal10 or with H129wt (Figs. 10D 10E 10F) . Because the Mac-1 antigen is expressed on microglia, it was not surprising that Mac-1-positive cells were also observed in the brain of mock-infected mice. However, the number of Mac-1-positive cells in the brain of mock-infected mice was lower than was observed in the brain of virus-infected mice (not shown). 

Discussion

In this study, a recombinant H129 virus expressing IL-16, a cytokine with lymphocytic and monocytic chemoattractant properties, was constructed to determine whether production of IL-16 would affect the timing or route of virus spread after AC inoculation and whether mice infected with IL-16-expressing recombinant virus would be protected from bilateral retinitis. Several key differences were observed in mice infected with H129/IL-16 compared with mice infected with H129wt alone or with mice infected with H129/pGal10. Infection of the SCN was delayed in mice infected with H129/IL-16 compared with mice infected with H129/pGal10 or H129wt. Between PI days 3 and 5, less viral antigen was detected in the brains of H129/IL-16-infected mice than in mice inoculated with H129wt or H129/pGal10. The lower amount of viral antigen correlated with recovery of significantly lower titers of virus from the brains of H129/IL-16-infected mice. Although H129/IL-16-infected mice developed bilateral retinitis, reduction in the amount of virus in the brain and delayed infection of the ipsilateral SCN correlated with a delay in infection of the contralateral retina and with recovery of significantly lower titers of virus in the uninoculated eye. Finally, more Mac-1-positive cells were observed in the injected eyes of H129/IL-16-infected mice early after infection compared with H129wt- or H129/pGal10-infected mice. 

The results in mice infected with H129/pGal10 were intermediate between those of mice infected with H129wt and mice infected with H129/IL-16 virus. For example, although virus was observed in the SCN of H129wt- and H129/pGal10-infected mice on the same day, less viral antigen was observed in the H129/pGal10-infected mice. Similarly, although virus was observed in the contralateral eye of H129/pGal10- and H129wt-infected mice at the same time, less viral antigen and lower titers of virus were recovered from H129/pGal10-infected mice. These results suggest that in vivo complementation and/or the vector used in the construction of the recombinant viruses influenced the spread and extent of viral infection. However, the observations that the titers of complemented virus were essentially the same for both H129/pGal10 and H129/IL-16 and that H129/IL-16 and H129/pGal10 replicated to similar titers in vitro suggest that the effect of the vector was similar in the recombinant viruses. Therefore, the presence of the vector alone does not account for the differences in the timing of viral infection in H129/IL-16-infected mice compared with H129wt- or H129/pGal10-infected mice and supports the idea that production of IL-16 in H129/IL-16-infected mice affects the timing of virus infection. Nevertheless, the results presented in this study demonstrated a decrease in virulence of H129/pGal10 compared with H129wt. Factors that may influence the virulence and therefore the spread of H129/pGal10 include a reduced efficiency of replication in certain cell types such as neurons, the ability to induce immunomodulatory molecules such as IFN-γ and TNF-α, and/or the capacity for the induction of inflammatory cells, all of which may limit the spread of H129/pGal10 in vivo. 

The mechanism by which IL-16 delays spread of virus in the brain and to the contralateral retina in H129/IL-16-infected mice is not known. However, delay of virus spread into the brain and contralateral retina most likely begins in the injected eye. At PI day 2, more Mac-1-positive cells were detected in the injected eyes of H129/IL-16-inoculated mice than in mice inoculated with H129wt or H129/pGal10. Although at PI day 2, H129/IL-16 virus and IL-16 were detected only in AC structures, such as the iris and ciliary body, many Mac-1-positive cells were also detected in the vitreous and along the ganglion cell layer of the retina. The mechanism by which production of IL-16 in the AC might induce infiltration of Mac-1 positive cells in the posterior segment of the eye is not clear. The Mac-1 antigen is expressed on macrophages, microglia, dendritic cells, and natural killer (NK) cells at various levels and is rapidly upregulated on neutrophils after activation. ^{42 43 44} Mac-1, also known as complement receptor 3 (CR3), is a member of the β2-intergrin family, which plays an important role in lymphocyte and leukocyte migration into tissues. Mac-1 mediates adhesion to c3bi and intracellular adhesion molecule (ICAM)-1. ^{43 45 46} The Mac-1-expressing cells infiltrating the injected eye probably include cells of monocytic lineage as well as dendritic cells. The ciliary body and iris contain resident macrophages and dendritic cells, ^{47 48} which may be induced to migrate by IL-16 present in these structures. Similarly, retinal microglia located in the ganglion cell layer, the inner plexiform layer, and the subretinal space ^{49 50} could migrate as a result of IL-16 in the injected eye. The precise role of Mac-1-positive cells in H129/IL-16 infection remains to be deciphered, but it may involve nonspecific antiviral mechanisms, such as secretion of type 1 interferons or other nonspecific immunomodulatory factors. 

In these studies, although virus was not detected in the contralateral SCN of most H129/IL-16-infected mice until PI day 5, the ipsilateral retina of all mice contained virus. In H129wt-infected mice, virus normally spreads from the contralateral SCN, which is infected at PI day 4, to infect the retina of the ipsilateral eye at PI day 5. ¹⁵ Thus, the results in H129/IL-16-infected mice suggest that in addition to spreading through the optic nerve, virus may also spread directly from the anterior to the posterior segment of the injected eye. Direct anterior-to-posterior spread of virus in H129/IL-16-infected mice may explain the observation that the ipsilateral retina is infected concomitant with the infection of the contralateral SCN. Because early direct spread of virus from the anterior to the posterior segment of the eye is controlled by NK cells, ⁵¹ early production of IL-16 may downregulate IFN-γ and other molecules required for NK cell activation. ⁵²  

Finally, IL-16 was detected in the brain of mice infected with wild-type or recombinant viruses. However, in mice infected with H129/IL-16, expression of IL-16 was more extensive than that observed in mice infected with either H129wt or H129/pGal10. This observation suggests that some of the IL-16 detected in H129/IL-16-infected mice was of recombinant origin. However, because of the limitations of the methods used to detect IL-16, additional studies are needed to distinguish endogenous IL-16 from IL-16 made by recombinant virus. 

Although the initial goal of these studies was to delay and/or prevent the spread of virus into the eyes by promoting early T-cell infiltration of the hypothalamus, the number of T cells observed in infected tissues of H129/IL-16-injected mice early (days 2–4) during infection was not significantly different from that observed in mice infected with H129wt or H129/pGal10. The lack of difference in T-cell recruitment most likely reflects delayed infection of the hypothalamus in H129/IL-16-infected mice and a lower amount of IL-16 available to participate in the chemotaxis of T-cells. Indeed, many infiltrating T-cells were observed in the hypothalamus of H129-IL-16-infected mice at PI day 6, when the highest titer of recombinant virus and the greatest number of recombinant virus-infected cells were also observed. More Mac-1-positive cells were also observed in the SCN of H129/IL-16-infected mice at PI day 6 than in mice infected with either wild-type virus or H129/pGal10. However, these cells did not appear to play a role in delaying virus spread to the contralateral retina, because virus had already spread to the contralateral retina at PI day 6. 

In conclusion, AC inoculation of BALB/c mice with H129/IL-16 resulted in a decrease of virus in infected tissues and a delay in virus infection of the SCN and of the contralateral retina. Although the mechanism by which virus infection was delayed is not clear, the results presented herein suggest that early recruitment of Mac-1-positive cells to the injected eye affects virus spread. Additional studies are needed to differentiate recombinant virus-produced IL-16 from endogenously produced IL-16 and to define the mechanism by which IL-16 retards virus spread after uniocular AC inoculation of H129/IL-16. 

 Supported by the National Eye Institute Grants EY06753 (NMA) and National Institutes of Health Grant EY06012 (SSA).

 Submitted for publication October 18, 2002; revised January 25, 2003; accepted February 7, 2003.

 Disclosure: N.M. Archin, None; L. van den Boom, None; L. Perelygina, None; J.M. Hilliard, None; S.S. Atherton, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Construction of H129/IL-16 and H129/pGal10. (A) Diagrams of HSV-1 (H129), pGal10/IL-16, and H129/IL-16. a, b, and c denote the inverted repeat sequences in the HSV genome. Plasmid pGal10/IL-16 was constructed by inserting the cDNA coding for functional IL-16 into the SalI and HindIII site of plasmid pGal10. IL-16 was placed under the control of the 110k HSV promoter (k). pGal10 contains the lacZ gene under the control of the HSV gC promoter (g) and is flanked by tk sequences to direct homologous recombination by disrupting the HSV tk locus. H129/IL-16 was produced after cotransfection of Vero cells with H129 and EcoRI-digested pGal10/IL-16 DNA. IL-16+, tk ⁻, β-gal⁺ virus was screened. (B) Diagrams of HSV-1 (H129), pGal10, and H129/pGal10. H129/pGal10 was produced by cotransfection of Vero cells with H129 and EcoRI-linearized pGal10 DNA. β-gal, tk ⁻ virus was screened. E, EcoRI; H, HindIII; S, SalI; X, XbaI.

Figure 2.

 

Southern blot hybridization confirming the presence of IL-16 in H129/IL-16. Viral DNAs were isolated and digested with HindIII and SalI, plasmid pGal10 was also digested with HindIII and SalI, and plasmid pRc/CMV/Il-16 was digested with HindIII. DNAs were electrophoretically separated, and Southern blot hybridization was performed. (A) Photomicrograph showing separation of digested viral DNA. (B) Blot of transferred DNA after probing with DIG-labeled IL-16 probe. The predicted IL-16 fragment size is indicated by the arrow and was detected only in H129/IL-16 and plasmid-digested DNAs. Ladder: DNA Marker VIII ladder (Roche Applied Science, Indianapolis, IN).

Figure 3.

 

In vitro production of IL-16 by H129/IL-16. Supernatant from Vero cells infected with H129/IL-16 was evaluated for production of IL-16 by ELISA. Supernatants from H129wt- and H129/pGal10-infected Vero cells were also studied. The highest level of IL-16 by H129/IL-16 was detected at 48 hours per 5 × 10⁶ PFU/mL of virus. IL-16 was not detected in the supernatant of either H129wt- or H129/pGal10-infected cells. Bkgrd, background.

Figure 4.

 

Single-cycle growth kinetics of H129wt, H129/IL-16, and H129/pGal10. Vero cells were infected at an MOI of 5 PFU/cell. Samples were collected at different time points, and the titer of virus in each sample was determined by standard plaque assays on Vero cells. The kinetics of viral growth was similar for all three viruses. The limit of the detection was 0.7 log₁₀ PFU/2 × 10⁶ cells.

Figure 5.

 

(A) Titers (average PFU ± SEM) of virus recovered from the ipsilateral brain of mice inoculated with H129wt, H129wt and H129/IL-16 (H129/IL-16), or H129wt and H129/pGal10 (H129/pGal10). Mice were inoculated in the AC of the right eye with 5 × 10³ PFU of H129wt alone or a mixture of 5 × 10³ PFU of H129wt and 1 × 10⁵ PFU of either H129/IL-16 or H129/pGal10. Mice were killed on the days indicated, and the brain from each mouse was removed and snap frozen. Brains were separated into right (ipsilateral) and left (contralateral) halves. A standard plaque assay was performed on tissue homogenates. Plaques were stained with X-gal. Clear and blue plaques were counted separately and the data combined and analyzed for statistical significance. Virus titers were determined in duplicate. Results are combined from two separate experiments: n = 3 and n = 5 mice per day, per group. *Significantly different from H129wt and H129/pGal10 (P < 0.05; Student-Newman-Keuls test). The limit of the detection was 0.7 log₁₀ PFU/tissue. (B) Delayed infection of the ipsilateral SCN of H129/IL-16-infected mice. Mice were inoculated in the AC with H129wt, H129wt and H129/IL-16 (H129/IL-16), or H129wt and H129/pGal10 (H129/pGal10). The brain from each mouse was removed and processed for immunohistochemistry to detect viral antigen.

Figure 6.

 

Photomicrograph of viral infection of the SCN of mice infected with H129wt and H129/IL-16 (A), H129wt and H129/pGal10 (B), or H129wt (C) at PI day 5. At this time, less viral antigen was observed in the SCN of H129/IL-16-infected mice than in H129wt- or H129/pGal10-infected mice. (⋆) Midline of the brain; OC; optic chiasma. Original magnification, ×99.

Figure 7.

 

(A) Titers (average PFU ± SEM) of virus recovered from the contralateral retina of mice inoculated with H129wt, H129wt and H129/IL-16 (H129/IL-16), or H129wt and H129/pGal10 (H129/pGal10). Mice were inoculated in the AC of the right eye with 5 × 10³ PFU of H129wt alone, or a mixture of 5 × 10³ PFU of H129wt plus 1 × 10⁵ PFU of either H129/IL-16 or H129/pGal10. Mice were killed on the days indicated, and the eyes of each mouse were removed and snap frozen. A standard plaque assay was performed on tissue homogenates. Plaques were stained with X-gal. Clear and blue plaques were counted separately and the data combined and analyzed for statistical significance. Virus titers were determined in duplicate. Results are combined from two separate experiments: n = 3 and n = 5 mice per day, per group. *Significantly different from H129wt and H129/pGal10 (P < 0.05 by Student-Newman-Keuls test). The limit of the detection is 0.7 log₁₀ PFU/ tissue. (B) Delayed infection of the contralateral retina of H129/IL-16-infected mice. Mice were inoculated in the AC with H129wt, H129wt and H129/IL-16 (H129/IL-16), or H129wt and H129/pGal10 (H129/pGal10). The eyes of each mouse was removed and processed for immunohistochemistry to detect viral antigen.

Figure 8.

 

Photomicrograph of viral infection of the contralateral retina of mice infected with H129wt and H129/IL-16 (A), H129wt and H129/pGal10 (B), or H129wt (C) at PI day 5. Viral antigen was not detected in the retina of mice infected with H129/IL-16 until PI day 6, whereas viral antigen (arrows) was detected in the contralateral retina of H129wt and H129/pGal10-infected mice at PI day 5. Original magnification, ×99.

Figure 9.

 

Photomicrograph of IL-16 expression in the iris of the injected eyes of mice infected with H129wt and H129/IL-16 (A), H129wt and H129/pGal10 (B), or H129wt (C) at PI day 2. IL-16 expression in the injected eyes of H129/IL-16-infected mice was more extensive (arrows) than that observed in H129/pGal10- or H129wt-infected mice. Viral antigen (D) and IL-16 (E) in the oculomotor nucleus of H129/IL-16-infected mice at PI day 6. Colabeling of viral antigen and IL-16 in the same area (F) showed that many infected neurons also expressed IL-16 (blue-green stained cells). Arrows: virus infected IL-16-expressing cells and/or cells expressing both antigens. Original magnification, ×198.

Figure 10.

 

Mac-1-positive cells in the retina and vitreous of the injected eyes of mice inoculated with H129wt and H129/IL-16 (A), H129wt and H129/pGal10 (B), or H129wt (C) at PI day 2. More Mac-1-positive cells were detected in the injected eyes of H129/IL-16-infected mice than in H129wt- or H129/pGal10-infected mice at PI day 2. Mac-1-positive cells were located in the vitreous (⋆⋆) and the ganglion cell layer (arrows) of the retina. A few Mac-1-positive cells were also observed in the inner plexiform layer (double arrows). A few Mac-1-positive cells were observed in the vitreous of H129wt-infected mice (arrowheads). Mac-1-positive cells were also detected in the AC (not shown). Mac-1-positive cells in the SCN of H129/IL-16 (D), H129/pGal10 (E), and H129wt (F) at PI day 6. More Mac-1-positive cells (arrows) were detected in the SCN of H129/IL-16-infected mice than in H129wt- or H129/pGal10-infected mice at PI day 6. (⋆) Midline of the brain. Original magnification, ×198.