|Home | About | Journals | Submit | Contact Us | Français|
Tumor necrosis factor alpha (TNF-α) has been shown to have a protective role in the eyes and brains of herpes simplex virus type 1 (HSV-1)-infected mice. To determine whether overexpression of TNF-α affected the course of virus infection following uniocular anterior chamber inoculation, a recombinant of HSV-1 that produces TNF-α constitutively (KOSTNF) was constructed. BALB/c mice were injected with the TNF-α recombinant, a recombinant containing the pCI plasmid, a recombinant rescue virus, or the parental virus. Flow cytometry and immunohistochemistry were used to identify virus-infected cells and to determine the numbers and types of infiltrating inflammatory cells in the uninjected eyes. Virus titers were determined by plaque assay. There were no differences among the groups in virus titers or the route and timing of virus spread in the injected eyes or in the suprachiasmatic nuclei. However, in the uninjected eyes of KOSTNF-infected mice, TNF-α expression was increased and there were more viral antigen-positive cells and immune inflammatory cells. There was earlier microscopic evidence of retinal infection and destruction in these mice, and the titers of virus in the uninjected eyes were significantly increased in KOSTNF-infected mice on day 7 postinfection compared with those of KOSpCI-, KOS6βrescue-, or KOS6β-infected mice. The results suggest that instead of moderating infection and reducing virus spread, overexpression of TNF-α has deleterious effects due to increased inflammation and virus infection that result in earlier destruction of the retina of the uninoculated eye.
Acute retinal necrosis (ARN) in human patients was initially described in Japan in 1971 as Kirisawa's uveitis (35). ARN is a potentially blinding disease that is characterized by necrotizing retinitis, optic neuropathy, retinal arteritis, vitritis, and retinal vasculitis (11, 15, 33). Several members of the human herpesvirus family, including herpes simplex virus type 1 (HSV-1), HSV-2, and varicella-zoster virus, have been implicated as causes of ARN (8, 12, 13, 18, 22, 23).
A mouse model of ARN has previously been described and studied (36, 38, 40). After inoculation of HSV-1 (KOS) into the anterior chamber (AC) of one eye, the virus spreads from the anterior segment of the injected eye via synaptically connected neurons and enters the optic nerve and retina of the uninjected contralateral eye by spreading from the ipsilateral suprachiasmatic nucleus (SCN) (25, 36). HSV-1 infection of the retinas of the contralateral eyes of BALB/c mice mimics several features observed in human patients with ARN (for example, necrotizing retinitis, retinal arteritis, and vitritis) (38). In mice, although the virus infects the contralateral SCN, it does not spread from there to infect the optic nerve and retina of the injected (ipsilateral) eye and the retina of the injected eye is not infected (36, 38).
The interaction between the virus and specific immunomodulators in patients with ARN is a complex process (40). The traffic of immune effector cells and effector modalities, such as cytokines, plays both a protective and a pathogenic role in the progression of ARN (2, 3, 40). The immune response to HSV infection involves both innate and adaptive immune responses. A primary infection is normally associated with a vigorous host response, including the infiltration of immune cells, cytokine production, and chemokine upregulation. The innate response is characterized by the infiltration of effector cells of the immune system, such as macrophages, neutrophils, and natural killer cells (39). These effector cells have the capacity to produce cytokines such as tumor necrosis factor alpha (TNF-α) and the type 1 interferons (39).
TNF-α is a cytokine produced by many cell types, including macrophages, T cells, astrocytes, and glial cells (1, 20, 24, 31, 34, 40). TNF-α is a pleiotropic proinflammatory cytokine that has a wide range of biological activities, such as initiation of the immune response and induction of cell death (7). TNF-α is a 26-kDa type II transmembrane glycoprotein, with a C terminus that is external to the cell and an intracellular cytoplasmic domain (19). The mature active form of TNF-α exists as a 17-kDa protein that is cleaved by matrix metalloproteinase enzymes (19). TNF-α has been implicated in the pathogenesis of many autoimmune and inflammatory diseases, such as rheumatoid arthritis and uveoretinitis (6, 29). Induction of TNF-α also induces the upregulation of adhesion molecules such as vascular cell adhesion molecule-1 on the vascular endothelium, thereby upregulating and promoting the infiltration of inflammatory cells such as neutrophils and macrophages (4, 27, 29). Results from studies using a mouse model of ARN suggest that cytokines such as TNF-α are active participants in the development of HSV-1 retinitis (40). In contrast, TNF-α has also been shown to play a protective role in the eyes and brains of HSV-1-infected mice (14, 26). The purpose of this study was to determine whether the expression of TNF-α by the HSV-1 (KOS) strain would affect virus location and spread in the eyes and brain following uniocular AC inoculation of BALB/c mice.
HSV-1 (KOS6β), constructed by the insertion of a cassette containing the early ICP6 promoter regulating the expression of β-galactosidase into the KOS strain of HSV-1 (a gift from David A. Leib, Washington University School of Medicine, St. Louis, MO), was used as the parental strain in this study (9). HSV-1 (KOS) wild type (KOS wt), without β-galactosidase inserted, was also used in some experiments. The stocks of all viruses used in this study were propagated on Vero cells (American Type Culture Collection [ATCC], Manassas, VA) grown in complete Dulbecco's modified Eagle's medium (DMEM) containing 5% fetal bovine serum (FBS) and antibiotics. The titers of the virus stocks were determined by standard plaque assays on Vero cells. The stocks were stored at −80°C in 0.5-ml aliquots, and a fresh aliquot of virus stock was thawed and diluted for each experiment. The titers of virus stocks used in these studies were 2.20 × 108 PFU/ml (KOS6β), 2.75 × 108 PFU/ml (KOSTNF), 4.04 × 108 PFU/ml (KOSpCI), 2.02 × 108 PFU/ml (KOS6βrescue), and 9.3 × 108 PFU/ml (KOS wt).
pMuTNF, a plasmid that contains a 1.1-kb sequence for mouse TNF-α (ATCC), was used in the study. Mouse TNF-α DNA was excised from pMuTNF by EcoR1 restriction digestion and purified. The gene for TNF-α was then cloned into a pCI mammalian expression vector (Promega, Madison, WI). This vector contains the human cytomegalovirus (CMV) major immediate-early gene enhancer/promoter region. The mouse TNF-α DNA was cloned into the pCI vector at the EcoR1 site to generate the expression vector pCI/TNF-α. In this vector, the mouse TNF-α gene is located downstream of the CMV major immediate-early gene enhancer/promoter region. The plasmid pUIC contains a unique BglII site in the intergenic region between the UL49 and UL50 genes of HSV-1 (9). The CMV IE promoter::TNF-α cassette was digested with BamHI and BglII and ligated into pUIC at the BglII site to create pUIC/TNF-α. The pUIC/lacZ plasmid contains an ICP6 promoter::lacZ cassette at the unique BglII site in the intergenic region between the UL49 and UL50 genes of HSV-1 (KOS) (9).
The TNF-α recombinant virus HSV-1 (KOSTNF) was generated from HSV-1 (KOS6β), as shown in Fig. Fig.1A.1A. HSV-1 (KOS6β) was grown to a high titer on Vero cells, and viral DNA was extracted from KOS6β virions that had been isolated by centrifugation. pUIC/TNF-α DNA was amplified by transforming Escherichia coli HB101 cells (Promega), and plasmid DNA was extracted and purified by using a S.N.A.P. MidiPrep kit (Invitrogen, Carlsbad, CA). An isolate containing the TNF-α gene was selected, grown in a large culture, and purified by using a S.N.A.P. MidiPrep kit. pUIC/TNF-α plasmid DNA was extracted and purified by using a S.N.A.P MidiPrep kit and digested with FspI (New England Biolabs, Beverly, MA). BHK-21 (hamster kidney) cells (ATCC) were cotransfected with KOS6β and digested plasmid DNA, using a lipophilic transfection agent, Lipofectamine 2000 (Invitrogen). To screen for recombinants, transfected cell supernatants were incubated with complete DMEM containing 5% FBS, antibiotics, 1% agarose, and 250 μg/ml β-galactosidase chromogenic substrate, X-GAL (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside; Promega), at 37°C in 5% CO2 for several days on cultured Vero cells. Clear plaques (indicative of TNF-α-containing recombinant virus) were selected and plaque purified by three passages through a cell culture before being grown to a high titer. To control for the effects of cloning the TNF-α cassette into the intergenic region between the UL49 and UL50 genes of the pUIC plasmid, a control virus containing a 1,344-bp fragment of pCI plasmid DNA alone, without TNF-α DNA, was also made, using the process described above (Fig. (Fig.1A1A).
A rescue recombinant virus was generated from HSV-1 (KOSTNF) by the removal of the TNF-α and reinsertion of the lacZ gene, as shown in Fig. Fig.1A.1A. HSV-1 (KOSTNF) was grown to a high titer on Vero cells, and viral DNA was extracted from KOSTNF virions that had been isolated by centrifugation. pUIC/lacZ DNA (a gift from David A. Leib) was amplified by transforming E. coli HB101 cells, and plasmid DNA was extracted and purified by using a S.N.A.P. Midiprep kit. BHK-21 cells were cotransfected with KOSTNF and digested plasmid DNA, using Lipofectamine 2000. To screen for recombinants, transfected cell supernatants were incubated with complete DMEM containing 5% FBS, antibiotics, 1% agarose, and 250 μg/ml β-galactosidase chromogenic substrate, X-GAL (Promega), at 37°C in 5% CO2 on cultured Vero cells. Blue plaques (indicative of recombinant rescue virus) were selected and plaque purified by three passages through cell cultures prior to preparation of the virus stock.
The viruses used in this study were genotyped by PCR. Primers were designed from publicly available GenBank sequences to ensure the specificity of the primers for the genes of interest. Samples were mixed with 100 pmol of primers specific for mouse TNF-α DNA, (sense) 5′-TCCAGAACATCCTGGAAATAGCTC-3′ and (antisense) 5′-AGAGGCCCACAGTCCAGGTCACTG-3′; lacZ DNA, (sense) 5′-GATGCGCCCATCTACACCAACGTG-3′ and (antisense) 5′-CAGCGCGGATCATCGGTCAGACGA-3′; or the HSV-1 VP16 gene, (sense) 5′-CGGTACCTGCGCGCCAGCGTC-3′ and (antisense) 5′-CAGCGGGAGGTTAAGGTGCTC-3′.
PCR was done, using a PCR reagent system (Invitrogen) and a Biometra T3000 thermocycler (Biometra, Goettingen, Germany). The PCR was performed according to the manufacturer's PCR protocol with 35 cycles at 94°C for 45 seconds, 55°C for 30 seconds, and 72°C for 90 seconds. After 35 cycles, the reaction mixture was maintained at 72°C for an additional 90 seconds. The products were subjected to electrophoresis on a 1% agarose gel and visualized under UV light.
Vero (African green monkey kidney) cells and SH-SY5Y (human brain [neuroblastoma]) cells (ATCC) were infected with HSV-1 (KOSTNF), HSV-1 (KOS6β), HSV-1 (KOSpCI), HSV-1 (KOS6βrescue), or HSV-1 (KOS wt) at a multiplicity of infection of 5/PFU per cell and incubated at 37°C in 5% CO2 for 1 h to allow the viruses to attach. The plates were then washed with phosphate-buffered saline (PBS) to remove any unattached viruses, overlaid with 2 ml DMEM containing 5% FBS and antibiotics, and incubated at 37°C in 5% CO2. At 0, 2, 4, 6, 12, and 24 h after infection, the cell supernatants were collected and stored at −80°C. To determine if TNF-α was produced, the samples were thawed and centrifuged at 300 × g at 4°C for 15 min. The protein concentrations were determined and normalized, using a Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). The supernatants were then assayed for TNF-α using a Quantikine mouse TNF-α enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems, Inc., Minneapolis, MN). The results were read by a microplate reader (AT 400; SLT, Hillsborough, NC) set to 450 nm and analyzed using DeltaSoft plate reader software (DeltaSoft, Princeton, NJ). Uninfected-cell supernatant was used as the negative control. The results were plotted graphically (DeltaGraph version 5.0; RockWare, Inc., Golden, CO). For virus titration, samples were freeze-thawed three times and centrifuged. The titers of virus in the supernatants were determined by plaque assays on duplicate cultures of Vero cells.
The cytotoxic activity of the recombinant TNF-α was assayed using WEHI-13VAR (mouse fibrosarcoma) cells (ATCC). The cells were grown in RPMI 1640 medium (ATCC) containing 10% FBS and 1% antibiotics-antimycotics (Gibco). The cells were seeded in a 96-microwell plate at 2 × 104 cells per well in 0.1 milliliters of RPMI medium containing 3% FBS and 1 μg/ml actinomycin D (Sigma). Recombinant TNF-α was concentrated from HSV-1 (KOSTNF)-infected Vero cell supernatants, using a Centricon Plus-70 centrifugal filter device (Millipore, Bedford, MA), and quantified using a Quantikine mouse TNF-α ELISA kit. Tenfold dilutions of recombinant TNF-α from HSV-1 (KOSTNF) and recombinant mouse TNF-α (R&D Systems, Inc.) were made in RPMI 1640 medium, and 0.1 ml of the dilutions were added to WEHI cells (in triplicate). After a 6-h incubation, cytotoxicity was determined, using a CytoTox 96 nonradioactive cytotoxicity assay (Promega). Absorbance was recorded at 490 nm, using a microplate reader and DeltaSoft plate reader software.
Adult female BALB/c mice 6 to 8 weeks old (Taconic, Germantown, NY) were used in all in vivo experiments. The mice were housed in accordance with National Institutes of Health guidelines, and all study procedures conformed to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee of the Medical College of Georgia. The mice were maintained on a 12-h light/dark cycle and were given unrestricted access to food and water. The mice were anesthetized by intramuscular injection of a mixture of 42.9 mg/ml ketamine, 8.57 mg/ml xylazine, and 1.43 mg/ml acepromazine (0.5 to 0.7 ml/kg) before all experimental manipulations.
Mice were anesthetized and inoculated through the AC route. The right eye was proptosed, aqueous humor was removed by paracentesis, and 2 μl of inoculum containing 2 × 104 PFU of HSV-1 (KOS6β), HSV-1 (KOSTNF), HSV-1 (KOS6βrescue), or HSV-1 (KOSpCI) was injected into the AC with a 30-gauge needle attached to a 100-μl microsyringe (Hamilton, Reno, NV). The inocula were prepared by diluting the virus stocks in DMEM with antibiotics.
The mice were deeply anesthetized and perfused transcardially with PBS for approximately 3 min. After perfusion, the brains and both eyes were removed and immediately stored at −80°C.
The brains and eyes from mice infected with HSV-1 (KOS6β), HSV-1 (KOSTNF), HSV-1 (KOS6βrescue), or HSV-1 (KOSpCI) were removed. One millimeter of coronal slices was prepared from the brains, using a brain matrix (ASI Instruments, Inc., Houston, TX), and slices containing the SCN were collected and homogenized in 500 μl of DMEM containing antibiotics. Both eyes from each mouse were homogenized individually in 500 μl of DMEM containing antibiotics. Each homogenate was serially diluted and plated on Vero cells (ATCC) that were 80% confluent. After adsorption of the virus for 1 h at 37°C in a CO2 incubator, the cells were washed to remove any unattached viruses and overlaid with a 1:1 solution of 2× DMEM (containing 10% serum and antibiotics) and 1% low-melt agarose (Life Technologies, Rockville, MD). After 5 days at 37°C, the cells were fixed with 10% buffered formalin and stained with 0.13% crystal violet (Sigma). The plaques were counted, and the titers of virus were calculated and analyzed using DeltaGraph (DeltaPoint, Inc., Monterey, CA).
The uninjected eyes from the mice infected with HSV-1 (KOS6β), HSV-1 (KOSTNF), or HSV-1 (KOSpCI) were removed and homogenized in 500 μl of DMEM containing antibiotics. The samples were then centrifuged at 300 × g at 4°C for 15 min. The protein concentrations were determined and normalized, using Bio-Rad protein assays. The supernatants were then assayed for TNF-α using Quantikine mouse TNF-α ELISA kits. The results were read by a microplate reader (AT 400; SLT) set to 450 nm and analyzed using DeltaSoft plate reader software. The results were plotted graphically, using DeltaGraph version 5.0.
Frozen sections were fixed with 4% paraformaldehyde, washed in PBS, and blocked with normal goat serum (Vector Laboratories) for 30 min. For virus detection, sections were incubated with an anti-HSV-1/fluorescein isothiocyanate (FITC) polyclonal antibody (DAKO, Carpinteria, CA), washed with PBS, and mounted with VectorShield containing DAPI (4′-6′-diamidino-2-phenylindole; Vector Laboratories). Slides were examined using a fluorescence microscope connected to the computer program Spot Advanced (Diagnostic Instruments, Inc., Sterling Heights, MI).
Frozen sections were fixed with acetone, air dried, and placed in hematoxylin for 2 min. The sections were washed in water, placed in 0.14N HCl in 70% ethyl alcohol, washed again in water, and stained with eosin for 30 seconds. The sections were then washed in water, dehydrated in a graded ethanol series, cleared with xylene, coverslipped, and examined microscopically.
Single-cell suspensions were prepared from the eyes of HSV-1 (KOSTNF)-, HSV-1 (KOSpCI)-, and HSV-1 (KOS6β)-infected mice on different days postinfection. Enucleated whole eyes were incubated with collagenase IV for 60 min at 37°C in a CO2 incubator. After being incubated, cells were removed by pressing the digested eye through a 70-μm nylon cell strainer (BD Falcon). The cells were suspended in Hank's balanced salt solution (Cellgro; Mediatech, Herndon, VA), centrifuged at 400 × g at 4°C for 5 min, and resuspended in PBS containing 0.5% FBS, 1 mM EDTA, and 25 mM HEPES. FITC anti-mouse CD11b (integrin αM chain, Mac-1 α chain; Pharmingen) and FITC anti-mouse Gr-1 (Ly-6G and Ly-6C) (RB6-8C5; Pharmingen), antibodies recognizing macrophages and neutrophils, respectively, were used to determine the extent of cellular infiltration. Flow cytometry of stained cell samples was performed (FACSCalibur; Becton, Dickinson, and Co., Franklin Lakes, NJ), and the flow cytometry results were analyzed using CellQuest software (Becton, Dickinson, and Co.).
To confirm the presence of the mouse TNF-α sequence in the HSV-1 (KOSTNF) viral genome, viral DNA was isolated and PCR was done with the appropriate primers. All plasmid DNA was sequenced for verification of the proper insertion of genes of interest into the intergenic region of the HSV-1 genome between the UL49 and UL50 genes. As shown in Fig. Fig.1B,1B, the sequence for mouse TNF-α was amplified from the HSV-1 (KOSTNF) recombinant virus but not from HSV-1 (KOS6β), HSV-1 (KOS6βrescue), or HSV-1 (KOSpCI) (vector-alone recombinant). The LacZ gene sequence was amplified from the HSV-1 (KOS6β) parental strain and the HSV-1 (KOS6βrescue) recombinant but not from HSV-1 (KOSTNF) or HSV-1 (KOSpCI) (Fig. (Fig.1B).1B). As expected, the sequence for VP16 was amplified from all parental and recombinant viruses (Fig. (Fig.1B1B).
To ensure that insertion of the TNF-α sequence did not affect the replication kinetics of HSV-1 (KOS6β), HSV-1 (KOS6βrescue), HSV-1 (KOSTNF), and HSV-1 (KOSpCI), single-cycle growth experiments were performed in Vero and SH-SY5Y cells. The kinetics of viral replication were similar for all viruses in Vero cells (Fig. (Fig.2A)2A) and in SH-SY5Y cells (Fig. (Fig.2B2B).
To verify production of TNF-α by HSV-1 (KOSTNF), ELISA was performed on supernatants collected from infected Vero cells and SH-SY5Y cells. As shown in Fig. 3A and B, TNF-α was detected as early as 4 h after HSV-1 (KOSTNF) infection of Vero cells and SH-SY5Y cells. TNF-α was not detected in the supernatants of Vero cells (Fig. (Fig.3A)3A) or of SH-SY5Y cells (Fig. (Fig.3B)3B) infected with HSV-1 (KOS wt), HSV-1 (KOS6β), HSV-1 (KOS6βrescue), or HSV-1 (KOSpCI).
To verify that the virally produced TNF-α was biologically active, TNF-α was concentrated from the supernatants of HSV-1 (KOSTNF)-infected cells. A recombinant TNF-α protein was used as the positive control. As shown in Fig. Fig.4,4, the cytotoxic activity of TNF-α protein produced by HSV-1 (KOSTNF) was comparable to that of recombinant TNF-α.
The above results show that the replication of HSV-1 (KOSTNF) is comparable to that of the wild-type virus, that TNF-α is produced by the virus, and that the virally produced TNF-α is biologically active. To determine whether the insertion of TNF-α affected virus spread, the eyes and brains of mice infected with HSV-1 (KOSTNF), HSV-1 (KOSpCI) (vector alone), or HSV-1 (KOS6β) (parental strain) were removed, frozen, sectioned, and stained for HSV-1 or homogenized for the recovery of replicating virus.
Virus infection was observed in the AC of the injected eyes of mice inoculated with HSV-1 (KOSTNF), HSV-1 (KOSpCI), or HSV-1 (KOS6β) beginning at day 1 postinfection (p.i.) and continuing until day 9 p.i. (data not shown). Irrespective of the infecting virus, infection was not observed in the retina of the injected eye of any mouse at any time (data not shown). Moreover, there were no significant differences in the amounts of virus recovered from the injected eyes of mice infected with HSV-1 (KOSTNF), HSV-1 (KOSpCI), or HSV-1 (KOS6β) at any time (data not shown).
After uniocular AC inoculation of HSV-1 (KOS), the virus spread from the injected eye and infected the ipsilateral SCN by day 5 p.i. and the contralateral SCN by day 7 p.i. (36). Viral antigen was detected in the ipsilateral SCN at day 5 p.i. (data not shown) and in both the ipsilateral and contralateral SCN at day 7 p.i. (data not shown) in mice infected with HSV-1 (KOSTNF), HSV-1 (KOSpCI), or HSV-1 (KOS6β). The titers of virus from the ipsilateral SCNs and contralateral SCNs of mice infected with HSV-1 (KOSTNF) were not significantly different from those recovered from HSV-1 (KOSpCI)- and HSV-1 (KOS6β)-infected mice on days 5, 6, and 7 p.i. (data not shown).
As shown in Fig. Fig.5,5, at day 7 p.i., the time when the virus was first observed in the retina of the uninoculated eye (36), more virus-infected cells were observed in the retinas of the uninoculated contralateral eyes of mice infected with HSV-1 (KOSTNF) than in the contralateral retinas of mice infected with HSV-1 (KOSpCI) or HSV-1 (KOS6β). At days 8 and 9 p.i., more viral antigen continued to be observed in the retinas of mice infected with HSV-1 (KOSTNF) than in mice infected with HSV-1 (KOSpCI) or HSV-1 (KOS6β) (Fig. (Fig.55).
To determine whether increased levels of viral antigen in the contralateral eyes of mice infected with HSV-1 (KOSTNF) correlated with a comparable increase in virus titers in these mice, the titers of virus in the contralateral eyes of mice infected with HSV-1 (KOSTNF) were compared with the titers of virus recovered from mice infected with HSV-1 (KOSpCI) or HSV-1 (KOS6β). As shown in Fig. Fig.6,6, the titers of virus recovered from the contralateral eyes of mice infected with HSV-1 (KOSTNF) were significantly higher (P < 0.05) than those for mice infected with HSV-1 (KOSpCI) or HSV-1 (KOS6β) on days 7, 8, and 9 p.i.
To determine whether the detection of more viral antigen and increased viral titers correlated with increased retinal pathology, hematoxylin- and eosin-stained sections of the contralateral eye were studied. As shown in Fig. Fig.7,7, the extent of retinal damage, amount of inflammatory cell infiltrate, and disruption of the retinal architecture/retinal necrosis was greater in the contralateral eyes of HSV-1 (KOSTNF)-infected mice than in the contralateral eyes of mice infected with HSV-1 (KOSpCI) or HSV-1 (KOS6β) at all time points.
ELISA was used to determine whether TNF-α was being produced in the uninoculated eyes of mice injected with HSV-1 (KOSTNF). The average amounts of TNF-α were higher in the contralateral eyes of KOSTNF mice than in those of HSV-1 (KOSpCI)- and HSV-1 (KOS6β)-infected mice on day 7 (77.34 pg/ml versus 12.41 pg/ml and 9.33 pg/ml), day 8 (92.46 pg/ml versus 8.26 pg/ml and 10.76 pg/ml), and day 9 (86.42 pg/ml versus 8.87 pg/ml and 16.33 pg/ml) p.i. At all time points, the amounts of TNF-α in the KOSTNF-infected uninoculated eyes were significantly higher than in the uninoculated eyes of mice infected with KOSpCI or KOS6β (P < 0.05).
Because TNF-α is a chemoattractant for cells such as macrophages and neutrophils, the contralateral eyes from mice infected with HSV-1 (KOSTNF), HSV-1 (KOSpCI), or HSV-1 (KOS6β) were analyzed by flow cytometry to determine whether overproduction of TNF-α affects recruitment of these cells into infected tissues. Flow cytometry revealed that there were more Mac-1-positive (Mac-1+) (Fig. (Fig.8A)8A) and Gr-1+ (Fig. (Fig.8B)8B) cells in the contralateral eyes of mice infected with the HSV-1 (KOSTNF) recombinant on days 7, 8, and 9 p.i. than in the contralateral eyes of mice injected with HSV-1 (KOSpCI) or HSV-1 (KOS6β).
To eliminate the possibility of an inadvertent genetic change or second-site mutation within the viral genome upon generation of the HSV-1 (KOSTNF) recombinant virus affecting the results, a revertant virus, HSV-1 (KOS6βrescue), was generated. Euthymic BALB/c mice injected with 2 × 104 of HSV-1 (KOSTNF) or HSV-1 (KOS6βrescue) were sacrificed on day 7, 8, or 9 to determine the timing and amount of viral spread in the uninoculated eyes. As shown in Fig. Fig.9,9, more virus-infected cells were observed in the contralateral eyes of mice infected with HSV-1 (KOSTNF) than in those of mice infected with HSV-1 (KOS6βrescue) at day 7 p.i. Moreover, the amounts of virus recovered from the contralateral eyes were significantly higher in mice infected with HSV-1 (KOSTNF) than in mice infected with HSV-1 (KOS6βrescue) at day 7 p.i. (Fig. (Fig.1010).
To determine whether the detection of more viral antigen and increased viral titers correlated with an increased retinal pathology, hematoxylin and eosin staining was performed on frozen sections of the contralateral eyes. As shown in Fig. Fig.11,11, the extent of retinal damage, retinal inflammation, and retinal necrosis was greater in the contralateral eyes of HSV-1 (KOSTNF)-injected mice than in the contralateral eyes of mice injected with HSV-1 (KOS6βrescue) day 8 and 9 p.i.
In this study, a recombinant HSV-1 (KOS) virus expressing TNF-α was used to determine whether production of TNF-α would affect virus spread after uniocular AC inoculation of BALB/c mice. Virus replication in vitro was essentially the same for all viruses in epithelial and neuronal cell lines, and in vivo, the pattern and timing of virus infection in the inoculated eye and SCN were not affected by the increased expression of TNF-α. However, several key differences were observed in the uninoculated eyes of mice infected with the KOS TNF-α-expressing recombinant and those of mice infected with the HSV-1 (KOS6β) parental strain, the HSV-1 (KOSpCI) recombinant control, or the HSV-1 (KOS6βrescue) revertant virus.
Previous studies from our laboratory have suggested that TNF-α plays a role in the limitation of virus spread in the SCN (14), so it was somewhat surprising that no differences in virus titers or the timing of spread were observed in the SCNs of the mice in any group. The lack of difference in the SCN could be due to the relatively low titer of virus (typical for the amount of virus in the hypothalamus following uniocular AC inoculation of the KOS strain of HSV-1) in the SCN. In KOSTNF-infected mice, the titer of virus in the SCN correlated with the amount of TNF-α, which in turn affected the extent of the effects of the recombinant TNF-α protein. All recombinant viruses, including the TNF-α-expressing recombinant HSV-1, followed the pattern and timing of spread from the injected eye that has been described previously for the spread of HSV-1 following uniocular AC inoculation (36).
In the present study, although infection with HSV-1 (KOSTNF) did not affect the location or the titer of virus in the ipsilateral eye or SCN, significant differences were observed in the uninoculated contralateral eye. In mice infected with HSV-1 (KOSTNF), the spread of the virus to the retina at day 7 p.i. was increased compared to that for HSV-1 (KOS6β)- and HSV-1 (KOSpCI)-infected mice. This trend continued until the peak of the disease at day 9 p.i. In accordance with the observation of increased viral spread in the retinas of HSV-1 (KOSTNF)-infected mice, the titers of virus in the uninoculated eyes were also significantly increased. Hematoxylin and eosin staining revealed that the extent of retinitis and tissue damage was increased at days 7, 8, and 9 p.i. in mice infected with HSV-1 (KOSTNF) compared with that in the HSV-1 (KOS6β)- and HSV-1 (KOSpCI)-infected mice, which correlated with the amount of recombinant TNF-α produced in the inoculated eye. Flow cytometry revealed more Mac-1 (CD11b)+ and Gr-1+ inflammatory cells in the uninoculated eyes of HSV-1 (KOSTNF)-injected mice.
To demonstrate that these results were not due to some inadvertent genetic change or second-site mutation created during the insertion of the gene for TNF-α, a revertant virus was generated from the HSV-1 (KOSTNF) recombinant. Comparisons of HSV-1 (KOSTNF)- and HSV-1 (KOS6βrescue)-infected mice showed that both groups of mice developed retinitis by day 7 p.i.; however, the retinitis in mice infected with HSV-1 (KOSTNF) was more severe, and infection was observed throughout the retina. In accordance with this observation, more viral antigen and increased viral titers were observed at day 7 p.i. in mice infected with HSV-1 (KOSTNF). Although no significant differences in virus spread or viral titers were observed at days 8 and 9 p.i. in mice infected with HSV-1 (KOSTNF) compared to those infected with HSV-1 (KOS6βrescue), the extent of retinitis and retinal destruction was increased in mice infected with HSV-1 (KOSTNF). While we cannot completely rule out that the lack of differences at day 8 and 9 p.i. may be caused by some unanticipated mutation or second-site mutation, use of the rescue recombinant virus suggests that overexpression of TNF-α exacerbates retinal pathology rather than limiting virus spread or replication.
TNF-α has multiple functions in the immune response, such as containment of local infections and participation in both neuroprotective and neurodestructive processes in chronic and acute neurodegenerative disorders (16, 26). TNF-α is involved in leukocyte migration into tissues and in the upregulation of intracellular cell adhesion molecule-1 and is also a major player in the induction of apoptosis (17, 30, 37). Many of the functions associated with TNF-α are a direct result of its proinflammatory properties such as leukocyte adhesion, vascular leakage, and apoptotic cell death (21). These events play an important role in the immune response to pathogens, but they can also play significant roles in the evolution of the pathology seen in many diseases. For example, during endotoxin-induced uveitis, the production of TNF-α correlates with the pathology, such as apoptotic cell death and vascular leakage, seen in this model (21). After treatment with a TNF-α inhibitor, a significant reduction in leukocyte rolling, adhesion, activation, and apoptosis was observed, thereby limiting the rate of inflammatory uveitis (21). Moreover, in models of experimental autoimmune uveoretinitis and experimental allergic encephalomyelitis, treatment with antibodies against TNF-α and sTNFr-IgG, a soluble TNF-α receptor, delays the onset of pathology and tissue damage in the retina and the central nervous system (5, 10, 29). Inflammation and apoptosis induced by cytokines such as TNF-α are regulated by factors such as NF-κB during immune and inflammatory responses in many pathological conditions (28, 32). TNF-α has also been implicated in the pathogenesis of HSV-1 retinitis and acute retinal necrosis (40). The interaction of this cytokine with other immunomodulatory factors is thought to be involved in the development or evolution of HSV retinitis (40). However, the response to overproduction of TNF-α in a mouse model of ARN has not been defined.
In this study, the overproduction of TNF-α correlated with increased inflammation in the contralateral retinas of mice infected with HSV-1 (KOSTNF). The amount of tissue damage in the contralateral retinas of mice infected with HSV-1 (KOSTNF) was also increased compared to that for mice infected with the vector-alone control, the parental strain, or the rescue control virus, suggesting that there is a direct relationship between the overproduction of TNF-α by the recombinant virus and the evolution of retinitis in the contralateral eye. One possible explanation for these observations is that the increase in inflammation may result in more-rapid retinal destruction due to increased infiltration of leukocytes to the infected retina. Increased retinal destruction and tissue degradation may, in turn, facilitate viral spread, resulting in an increase in virus titers at earlier time points, corresponding to earlier destruction of the retina of the uninoculated contralateral eye. Comparisons of TNF-α recombinant virus-infected mice with mice infected with the rescue virus showed no significant differences in virus titers in the contralateral eyes at days 8 and 9 p.i. even though there were striking differences in the degrees of retinitis (characterized by the amount of inflammatory cell infiltrate and disruption of the retinal architecture/retinal necrosis). This result further supports the idea that in addition to virus infection, inflammatory mediators such as TNF-α play a role in the pathology of ARN. In HSV-1 (KOS)-infected Gr-1-depleted mice, the extent of retinitis in the uninoculated contralateral eyes, as well as the virus titer, was reduced compared to that for the control immunoglobulin G-treated mice (M. Zheng, M. Fields, Y. Liu, H. Cathcart, E. Richter, and S. S. Atherton, submitted for publication). These observations suggest that the infiltration of inflammatory cells such as Gr-1 plays a role in the exacerbation of the pathology seen in ARN.
The results of this study indicate that instead of moderating infection and reducing virus spread, overexpression of TNF-α has deleterious effects due to increased inflammation and the presence of more virus-infected cells, which lead to earlier destruction of the retina of the uninoculated eye. These results support the idea that the exact relationship between inflammatory modulators and the virus in the evolution of HSV-1-induced retinitis is a complex process. Taken together, the results of this study underscore the challenge of trying to understand how to exploit cytokine gene-delivery systems to combat disease progression in disorders such as ARN, since in some sites (e.g., the SCN) related to virus spread TNF-α is beneficial, whereas in other sites (e.g., the retina of the uninoculated eye) it exacerbates disease pathology. Additional studies are needed to decipher the mechanisms by which the virus and inflammatory cells and their products interact in the retina during the development of ARN in the mouse model and in human patients.
This work was supported by Public Health Service grant EY006012 (S.S.A.) and a fellowship grant EY015392 (M.A.F.).
The authors gratefully acknowledge David A. Leib, Washington University School of Medicine, St. Louis, MO, for his assistance in the preparation of the recombinant herpes simplex viruses and Penny Roon, Medical College of Georgia, for her assistance with hematoxylin and eosin staining.
Published ahead of print on 5 March 2008.