Molecularly Cloned Feline Immunodeficiency Virus NCSU 1 JSY3 Induces Immunodeficiency in Specific-Pathogen-Free Cats

JOURNAL OF VIROLOGY, May 1996, p. 3011 3017 Vol. 70, No. 5 0022-538X/96/$04.00 0 Copyright 1996, American Society for Microbiology Molecularly Cloned Feline Immunodeficiency Virus NCSU 1 JSY3 Induces Immunodeficiency in Specific-Pathogen-Free Cats JOO-SUNG YANG, 1 ROBERT V. ENGLISH, 1 JERRY W. RITCHEY, 1 MICHAEL G. DAVIDSON, 2 TERRI WASMOEN, 3 JULIE K. LEVY, 1 DOUGLAS H. GEBHARD, 1 MARY B. TOMPKINS, 1 AND WAYNE A. F. TOMPKINS 1 * Department of Microbiology, Parasitology, and Pathology 1 and Department of Companion Animal and Special Species Medicine, 2 College of Veterinary Medicine, North Carolina State University, Raleigh, North Carolina, and Research and Development, Fort Dodge Laboratories, Fort Dodge, Iowa 3 Received 21 August 1995/Accepted 26 January 1996 A full-length feline immunodeficiency virus NCSU 1 (FIV-NCSU 1 ) genome (JSY3) was cloned directly from FIV-NCSU 1 -infected feline CD4 lymphocyte (FCD4E) genomic DNA and identified by PCR amplification with 5 long terminal repeat, gag, env, and 3 long terminal repeat primer sets. Supernatant from FCD4E cells cocultured with JSY3-transfected Crandell feline kidney (CrFK) cells was used as an inoculum. Cell-free JSY3 virus was cytopathogenic for FCD4E lymphocytes but did not infect CrFK cells in vitro. To determine in vivo infectivity and pathogenesis, six young adult specific-pathogen-free cats were inoculated with cell-free JSY3 virus. Provirus was detected at 2 weeks postinfection (p.i.) and was still detectable at 25 weeks p.i. as determined by gag region PCR-Southern blot analysis of peripheral blood mononuclear cell lysates. Infectious virus was recovered from peripheral blood mononuclear cells at 6 and 25 weeks p.i., and an antibody response to FIV was detected by 4 weeks. In the acute phase of infection, JSY3 provirus was found only in the CD4 lymphocyte subset; however, by 14 weeks p.i., the greatest provirus burden was detected in B lymphocytes. All six cats were panlymphopenic at 2 weeks p.i., CD4 /CD8 ratios were inverted by 6 weeks p.i., and five of the six cats developed lymphadenopathy by 10 weeks p.i. To determine if the JSY3 molecular clone caused immunodeficiency similar to that of the parental wild-type FIV-NCSU 1, the cats were challenged with the low-virulence ME49 strain of Toxoplasma gondii at 29 weeks p.i. Five of six cats developed clinical signs consistent with generalized toxoplasmosis, and three of six cats developed acute respiratory distress and required euthanasia. Histopathologic examination of the severely affected cats revealed generalized inflammatory reactions and the presence of T. gondii tachyzoites in multiple tissues. None of the six age- and sex-matched specific-pathogen-free cats inoculated with only T. gondii developed clinical disease. Our results suggest that the pathogenesis of the molecularly cloned NCSU 1 JSY3 is similar to that of wild-type FIV-NCSU 1. Feline immunodeficiency virus (FIV), a lentivirus of cats, causes an AIDS-like disease similar to that seen in humans with human immunodeficiency virus (HIV) infection. FIV shares features with HIV, such as structural morphology, Mg 2 -dependent reverse transcriptase (RT) activity (24, 35), and genome organization (23, 29). In vivo, FIV causes an early plasma- and cell-associated viremia, followed by a strong antibody response that correlates with a marked reduction in plasma viremia (8, 22, 35). Cats develop an acute infection syndrome similar to that seen in HIV type 1, including lowgrade fever and transient generalized lymphadenopathy, followed by a long asymptomatic period in which the CD4 / CD8 ratio declines because of a progressive decrease in CD4 cells (11, 31, 35). The asymptomatic period progresses into a stage characterized by a variety of AIDS-associated disorders, such as opportunistic infections, chronic gingivitis or stomatitis, chronic upper respiratory infections, superficial and systemic fungal infections, and ocular disease (9, 11, 25, 35). The development of neoplasms, particularly B-cell lymphomas, neurologic disease, and a chronic wasting syndrome marks the terminal stages of FIV infection (4, 11). The similarities between HIV and FIV in genome organization, biological characteristics, and disease progression toward AIDS suggest that * Corresponding author. Mailing address: Department of Microbiology, Parasitology, and Pathology, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough St., Raleigh, NC 27606. Phone: (919) 515-7400. Fax: (919) 515-4237. FIV infection should provide an excellent animal model for HIV infection. A major limitation of the FIV-feline model for the study of human AIDS is the unavailability of molecular clones that retain the pathogenic characteristics of the wildtype viruses. To more fully understand the molecular basis for determinants of FIV pathogenesis and its relevance to HIV infection, it is necessary to obtain genetically homogeneous molecular clones of FIV that retain the biological and diseasecausing properties of the pathogenic wild-type populations. At this time, five FIV molecular clones, FIV-14 (23), FIVpF34 of FIV-Petaluma (28), FIV-pPPR of FIV-PPR (28), pftm191cg of FIV-TM1 (20), and 19K1 of FIV Amsterdam-19 (27), have been reported to be infectious in vivo as determined by seroconversion, cell-associated virus, and the presence of FIV provirus. However, no clone to date has been reported as pathogenic to the extent that it causes immunodeficiency and increased susceptibility to secondary opportunistic infections. An isolate of FIV (FIV-NCSU 1 ) that is pathogenic in vivo, as measured by a severe loss of CD4 cells and development of secondary infections, severe wasting, neurological disease, and B-cell lymphomas, has been described recently (11). Davidson et al. (7) were able to demonstrate that FIV-NCSU 1 causes a relatively early and profound state of immunodeficiency, as measured by loss of resistance to challenge with a Toxoplasma gondii strain with a low level of virulence. This dual FIV-T. gondii infection provides an excellent model with which to determine the ability of FIV isolates as well as molecular 3011

3012 YANG ET AL. J. VIROL. TABLE 1. Criteria for scoring of clinical status after T. gondii inoculation Score Attitude Appetite Respiratory signs 0 Normal, alert, and active Normal appetite Normal rate and effort 1 Alert but inactive Slight (less than half) decrease in food intake Mild increase in rate (30 50/min) 2 Depressed and inactive Moderate (more than half) decrease in food intake Moderate increase in rate ( 50/min) 3 Profoundly depressed and inactive Anorexic Increased rate and effort 4 Moribund Profound tachypnea and dyspnea clones of FIV to cause immunodeficiency. In this paper, we describe the in vitro and in vivo biological characteristics of a molecular clone (JSY3) of the FIV-NCSU 1 isolate and utilize the FIV-T. gondii infection model to establish that the JSY3 molecular clone causes immunodeficiency in cats. MATERIALS AND METHODS Viruses. The biological parent virus isolate FIV-NCSU 1 was obtained from the peripheral blood mononuclear cells (PBMCs) of a cat naturally infected with FIV and has been described elsewhere (7, 10, 11, 31). The FIV-NCSU 1 molecular clone JSY3 inoculum was collected from an FCD4E feline lymphocyte culture which had been cocultured with transfected Crandell feline kidney (CrFK) cells (see below). Molecular cloning of the FIV proviral genome. Genomic DNA was isolated by equilibrium centrifugation in CsCl-ethidium bromide gradients (18) from 5 10 7 FCD4E cells (interleukin-2-dependent, FIV-NCSU 1 -infected feline CD4 lymphocytes) inoculated with FIV-NCSU 1 obtained from the original source cat. FCD4E genomic DNA which had been partially digested with Sau3AI and size fractionated was cloned into the EMBL3 lambda vector arm. Genomic libraries were screened primarily by plaque hybridization with a gag region PCR product probe (838 bp) as described elsewhere (18). A full-length clone was identified by PCR of phage suspension with six primer sets designed from FIV-14 sequences (GenBank accession no. M25381). These primer sets amplified 5 long terminal repeat, gag, env, and 3 long terminal repeat regions under the PCR conditions described below. The following primers were used for identification of the fulllength lambda clone JSY3: 3U (U3), 5 -GGA TGA GTA TTG GAA CCC TGA A-3 ; 337L (U5), 5 -GAT TCC GAG ACC TCA CAG GTA A-3 ; 447U, 5 -AAT AGG GAA GCA GTA GCA GAC-3 ; 829L, 5 -GTA AAT CGC AAA TAA CCA ACC-3 ; 919U (FIV7), 5 -TGA CGG TGT CTA CTG CTG CT-3 ; 1756L (FIV8), 5 -CAC ACT GGT CCT GAT CCT TTT-3 ; 1057U, 5 -CCA CAA TAT GTA GCA CTT GAC C-3 ; 1639L, 5 -GGG TAC TTT CTG GCT TAA GGT G-3 ; 6938U, 5 -GGG GGA CCT ACC TTG GGG AAT TGG GCT-3 ; 7252L, 5 -GGT GAT CAT GAT CAG TGG GAT TTG TAA TGG GTC TG-3 ; 8859U, 5 -ATA AGG GAG ATA CTG TGC TGA-3 ; and 9029L, 5 -GCG ATC TTC TAA CTC TGT CAT-3. Each primer was designated by the 5 nucleotide of the complete FIV-14 sequence. DNA transfection. Ten micrograms of lambda clone DNA was transfected into CrFK and AH927 (a feline embryonic fibroblast cell line) cells by using the cationic liposome DOTAP (Boehringer Mannheim, Indianapolis, Ind.) according to the manufacturer s protocol. Twenty-four hours after transfection, these cells were cocultured for 72 h with FCD4E or concanavalin A (10 g/ml)- stimulated normal cat PBMCs. FCD4E (or PBMCs) and CrFK (or AH927) cells were then cultured separately. Culture supernatant was collected at 3- to 4-day intervals and assayed for RT activity. Pooled samples for in vivo infection were titrated in FCD4E cells by the 50% tissue culture infective dose (TCID 50 ) method. In vitro infections with JSY3 clone. Cultures of FCD4E or DEAE-dextrantreated CrFK cells were inoculated with cell-free FIV-NCSU 1 JSY3 clone containing 2 10 4 cpm of RT activity. The culture supernatant was collected twice weekly and assayed for RT activity. In vivo FIV infection. Six 6-month-old female cats were inoculated intravenously with 10 6 TCID 50 s of the JSY3 clone. Nine age- and sex-matched specificpathogen-free (SPF) cats were inoculated with wild-type FIV-NCSU 1, and nine mock-infected SPF cats were used as controls. The wild-type-fiv-ncsu 1 -infected group was examined up to 18 weeks postinfection (p.i.) in parallel with the JSY3-infected cats and then utilized in other experiments. Blood sampling. Whole blood was collected by jugular venipuncture into sodium citrate anticoagulant tubes. Aliquots were removed for complete blood counts and flow cytometry, and plasma was collected for anti-fiv antibody assays. PBMCs were purified over Percoll as described previously (32). PBMCs were then cocultured with FCD4E cells for infectious-virus recovery, lysed for provirus detection by PCR, or sorted for lymphocyte subset tropism studies. Lymphocyte subset analysis by flow cytometry. Lymphocyte subsets were determined by two-color flow cytometric analysis as described elsewhere (7) using a panel of monoclonal antibodies (MAbs) developed in our laboratory (30). Briefly, plasma was removed, the cells were washed twice in phosphate-buffered saline (PBS), and MAbs were added in a combination of fluorescein isothiocyanate-labeled anti-cat immunoglobulin (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md.) and biotin-labeled anti-pan T-cell antibodies or fluorescein isothiocyanate-labeled anti-cd8 and biotin-labeled anti-cd4 antibodies. Biotinlabeled antibodies were developed with phycoerythrin. Erythrocytes were lysed with fluorescence-activated cell sorter (FACS) lysing solution (Becton Dickinson Immunocytometry Systems, San Jose, Calif.), and the percent positively stained lymphocytes was determined by flow cytometric analysis using a Becton Dickinson FACScan. The absolute numbers for each lymphocyte subset were calculated by multiplying the percent positive cells by the total number of lymphocytes, determined by a complete blood count and differential performed on the blood sample. PCR-Southern blot analysis for FIV provirus detection. Percoll-purified PB- MCs were washed with PBS, and cell pellets were stored at 70 C until assayed. Cells (10 6 ) were lysed in 200 l of1 PCR buffer and digested with 600 g of proteinase K per ml. An 838-bp length of the FIV gag region was amplified with the primer set 919U-1756L. Amplification was performed as described previously (10), with minor modifications. Briefly, 2 l of cell lysate (equivalent to 10 4 cells) was amplified in a 100- l PCR mixture (1 PCR buffer, 1.5 mm MgCl 2, 200 M each deoxynucleoside triphosphate, 0.5 M each primer, and 2.5 U of Taq DNA polymerase [Promega, Madison, Wis.]) over 40 cycles (one cycle was 94 C for 1 min, 59 C for 2 min, and 72 C for 1 min; final extension was done at 72 C for 10 min). Amplified products were resolved on a 1.2% agarose gel, blotted, and hybridized with radiolabeled internal oligonucleotides probe. Western blot analysis for plasma antibody to FIV. The Western blot (immunoblot) assay was performed as previously described (21). RT activity assay. The Mg 2 -dependent RT activity assay was performed as previously described (21) and is a modification of the procedure of Goff et al. (13). Lymphocyte subset sorting of feline PBMCs. The JSY3 clone-infected cat PBMCs were sorted into CD4, CD8, and B lymphocyte subsets by using MiniMACS (Miltenyi Biotec, Sunnyvale, Calif.) magnetic beads. Percoll-enriched PBMCs were divided among three tubes and incubated at 4 C for 30 min with biotin-labeled anti-cd4 or anti-cd8 or anti-canine B-cell MAb (B5) for a non-immunoglobulin-positive B-cell epitope (10). Streptavidin-conjugated Mini- MACS beads were then added, and the cells were incubated for an additional 20 min at 4 C and then positively sorted. A fraction of each sorted subset was analyzed for purity by two-color flow cytometry. Cells were stained with biotinlabeled MAbs, developed with phycoerythrin-conjugated streptavidin, and analyzed on the FACScan. The remaining sorted lymphocytes were stored at 70 C until they were assayed for the presence of FIV provirus by PCR-Southern blotting. T. gondii infection. Twenty-nine weeks after infection with the JSY3 clone, cats were inoculated via the carotid artery with 10,000 tachyzoites of the ME49 strain of T. gondii as described previously (7). Six age- and sex-matched SPF cats were also inoculated with T. gondii as controls. The cats were examined daily for clinical signs of illness by using scoring criteria developed by Davidson et al. (7) (Table 1). Cats showing severe clinical signs indicative of generalized toxoplasmosis were euthanized by barbiturate overdose. Postmortem examination. Following euthanasia, a gross necropsy was performed and tissues were sampled for microscopic examination. Tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin stain. RESULTS Molecular cloning of the FIV-NCSU 1 proviral genome. A total of 5 10 7 FCD4E cells were infected with wild-type FIV-NCSU 1 from the FIV-NCSU 1 source cat. Genomic DNA from this culture was cloned into the EMBL3 lambda vector arm. Primary hybridization-positive clones, determined by plaque hybridization with a randomly labeled 838-bp FIV gag PCR product probe, were screened further by PCR as described in Materials and Methods. Five microliters of phage plaque suspensions of each hybridization-positive clone was directly amplified with six different primer sets, and a fulllength proviral clone was identified (designated JSY3). The

VOL. 70, 1996 PATHOGENESIS OF FIV MOLECULAR CLONE 3013 FIG. 1. Mg 2 -dependent RT activity in culture supernatant after transfection with JSY3 lambda clone. Ten micrograms of lambda DNA was transfected into CrFK and AH927 cells, which were then cocultured with either FCD4E cells or concanavalin A-stimulated PBMCs. After 72 h, all cells were cultured separately. No detectable RT activity was present in the CrFK or AH927 cultures. Data are the means and standard deviations of triplicates. specificity of each FIV PCR product was established by comparing it with the FIV-pPPR plasmid clone (26). Biological activity of clone JSY3. To determine the biological activity of the JSY3 clone, lambda DNA was transfected into CrFK, AH927, and FCD4E cells, which were then cocultured with FCD4E cells or PBMCs. While no RT activity was detected in culture supernatants of JSY3-transfected CrFK or AH927 cells when cultured alone, RT activity was detected when the transfected cells were cocultured with either PBMCs or FCD4E cells (Fig. 1). The replication kinetics of FIV in FCD4E cells is more rapid than in PBMCs because of the greater percentage of CD4 cells in the FCD4E culture. Supernatants collected at 15 and 19 days of culture from FCD4E cells were filtered (0.2- m pore size) and stored in aliquots for use as in vitro and in vivo inocula. These inocula were designated the FIV-NCSU 1 JSY3 clone. No RT activity was detected in the FCD4E cultures directly transfected with JSY3, suggesting that the transfection was unsuccessful (data not shown). To determine in vitro infectivity of the JSY3 clone, FCD4E and CrFK cells were inoculated with cell-free JSY3 clone. Similarly to the FIV-NCSU 1 wild-type virus (10), the JSY3 clone replicated efficiently in FCD4E cells, resulting in syncytium formation and cell death (data not shown). However, the JSY3 clone was unable to infect CrFK cells. In vivo infectivity of molecular clone JSY3. To determine the in vivo infectivity of the JSY3 molecular clone, six SPF cats were inoculated intravenously with 10 6 TCID 50 s of JSY clone. Nine age-matched SPF cats were inoculated with 10 6 TCID 50 s of the biological parent FIV-NCSU 1, also produced in FCD4E cells. Plasma and PBMCs were collected at various times p.i. and tested for antibodies to FIV by Western blotting and tested for cell-associated FIV provirus by PCR. As previously reported (11, 31), cats infected with the FIV-NCSU 1 parent virus were anti-fiv antibody positive by 4 weeks p.i. and were provirus positive by PCR by 2 weeks p.i. (data not shown). The response of the cats infected with the JSY3 clone was similar to that of the cats infected with the wild type. By 4 weeks p.i., all six cats had antibody to the FIV gag proteins p17 and p24, and they were still antibody positive at 25 weeks p.i. (data not shown). The presence of FIV provirus in PBMCs from six cats infected with the JSY3 clone was determined by PCR and Southern analysis. A PBMC lysate (equivalent to 10 4 cells) was amplified with the gag region primer set 919U- 1756L, resolved on an agarose gel, and subjected to Southern blot analysis with a 5 -end-labeled internal probe. Provirus was detected in PBMCs from all cats by 2 weeks p.i. (Fig. 2), and all cats remained provirus positive for the remainder of the study. Cat CCH6 was provirus positive when the amount of cell lysate in the PCR mixture was increased (data not shown). To establish the presence of infectious virus in PBMCs from the JSY3-infected cats, PBMCs collected at 6 and 25 weeks p.i. were cocultured with FCD4E cells and the supernatants were assayed for RT activity. Syncytium formation and cell death were observed in cocultures from all six cats at both 6 and 25 weeks p.i. RT activity was detectable in all cocultures by 8 to 10 days and peaked by 16 to 18 days of culture (data not shown). Lymphocyte subset changes in JSY3-infected cats. Lymphocyte profiles in naturally and experimentally FIV-infected cats are well documented (1, 11, 15, 21, 31) and similar to those for HIV. To determine whether the JSY3 clone causes hematologic and immunologic abnormalities similar to those of the biological parent FIV-NCSU 1, lymphocyte subset profiles were analyzed by two-color flow cytometry. As reported previously for the NCSU 1 biological isolate (11, 31), both the biological virus and the JSY3 clone caused a panlymphopenia 2 to 4 weeks p.i. The parent FIV-NCSU 1 and the JSY3 molecular clone caused parallel alterations in the CD4 /CD8 ratio (Fig. 3). At 6 weeks p.i., the mean CD4 /CD8 cell ratios ( standard errors) decreased from 3.48 0.50 to 1.30 0.21 for JSY3-infected cats and from 3.23 0.19 to 1.07 0.20 for the parent virus-infected cats. By using total cell counts and flow cytometric analysis of lymphocyte subsets, the decrease in the CD4 /CD8 ratio was determined to be the result of a decrease in CD4 lymphocytes and an increase in CD8 lymphocytes (Fig. 4, data collected at 21 weeks p.i.). These results indicate that the JSY3 clone-infected cats have hematologic and immunologic abnormalities, including CD4 and CD8 lymphocyte changes similar to those of cats infected with the biological parent virus. In vivo lymphocyte tropism. The in vivo hematopoietic target cells of FIV isolates, including NCSU 1, have been reported to be CD4, CD8, monocytes, and B lymphocytes (2, 3, 10). To examine whether the JSY3 molecular clone has a similar panlymphotropism in vivo, PBMCs from JSY3 clone-infected FIG. 2. PCR of unsorted PBMCs from six JSY3-infected cats (collected at the indicated times p.i.). A lysate from each sample representing 10 4 cells was amplified, resolved on an agarose gel, and analyzed by Southern blotting. The positive control ( ) is cell lysate from FIV-Black-infected CrFK cells, and the negative control ( ) is lysate from normal, uninfected FCD4E cells. The image was generated on a Macintosh computer using Adobe Photoshop software.

3014 YANG ET AL. J. VIROL. FIG. 3. Mean CD4 /CD8 ratios of JSY3 clone-infected (n 6), NCSU 1 wild-type-infected (n 9), and mock-infected (n 6) cats. Ratios were calculated from the percent CD4 and CD8 cells as determined by flow cytometry. Data are the means of the individual ratios, and error bars represent standard errors of the means. FIG. 4. Mean CD4 and CD8 cell numbers of six SPF cats before and 21 weeks after infection with the JSY3 clone and of six age-matched controls before T. gondii infection. Cell numbers were determined by multiplying the percent positive CD4 or CD8 cells, determined by flow cytometry, by the total lymphocyte count from a complete blood count. The CD4 /CD8 ratios are the means of the individual ratios. Error bars represent standard errors of the means. cats were sorted into CD4, CD8, and B lymphocyte populations by using antibody-coated magnetic beads. Each cell subset was lysed, PCR amplified with the gag region 919U- 1756L primer set, and analyzed by Southern blotting. As previously reported for the NCSU 1 parent virus, FIV provirus was first detected in CD4 lymphocytes during the acute-stage infection with JSY3 (2 to 4 weeks p.i.) (Fig. 5). However, at a later stage of infection (as early as 14 weeks p.i.), FIV provirus was found in CD8 and B lymphocytes in addition to CD4 lymphocytes, as reported for the parent FIV-NCSU 1 isolate (10). While data are presented for only two cats, all six FIV JSY3-infected cats showed similar shifts in provirus burden from predominantly CD4 cells during the acute-stage infection to predominantly B cells during the asymptomatic stage. While CD4 and CD8 cells were not always positive for provirus under the PCR conditions described in Materials and Methods, we were always able to detect provirus in these cells during the asymptomatic-stage infection by increasing cell numbers or using nested primers as described by English et al. (10). Therefore, the JSY3 molecular clone, similarly to the parent biological isolate, appears to have a CD4 tropism during the acute-stage infection that then shifts to a panlymphotropism as the infection progresses. Acute-stage disease in JSY3-infected cats. In the primary phase of infection (2 to 16 weeks p.i.), both the JSY3- and the parent isolate-infected cats developed low-grade fevers, panlymphopenia, neutropenia, and generalized lymphadenopathy (data not shown), as has been reported for a number of biological isolates of FIV (35), including NCSU 1 (11). Clinical response of JSY3-infected cats to T. gondii challenge. Davidson et al. (7) reported that FIV-NCSU 1 causes immune system impairment in cats as early as 18 weeks after infection and enhances susceptibility to a primary T. gondii infection. To further explore the pathogenicity of molecularly cloned FIV-NCSU 1 JSY3 virus and determine if the molecular clone caused immune impairment early in the asymptomatic stage of infection, the cats were parenterally inoculated with the ME49 strain of T. gondii 29 weeks after JSY3 infection. Six age-matched SPF control cats were similarly infected with T. gondii. At the time of T. gondii inoculation, all six FIV-infected cats were clinically normal; however, they had a marked decrease in their CD4 /CD8 ratios in comparison with preinfection ratios and those of the control cats (Fig. 4). Only one of six T. gondii-infected cats in the non-fiv-inoculated group had positive clinical scores, as a result of anorexia and lethargy on days 8 to 11 after inoculation (Fig. 6). Cats in this group also developed multifocal chorioretinitis beginning on days 7 to 10 after inoculation, which resolved over a 3-week course. The infection was otherwise subclinical in these cats. This clinical response is similar to that previously reported for healthy cats challenged with the mildly virulent ME49 strain of T. gondii (6, 7). Five of the six FIV-positive cats challenged with T. gondii had positive clinical scores in all three categories (attitude, appetite, and respiratory signs), and the total scores were higher than those of the T. gondii control group (Fig. 6). Beginning on days 6 to 9 after inoculation, three FIV-infected cats challenged with T. gondii developed high fevers, depression, and moderate to severe ocular lesions, including chorioretinitis with subretinal granuloma formation, localized retinal detachment, and fibrinous anterior uveitis. Severe and progressive tachypnea, dyspnea, tachycardia, and icterus were noted, and interstitial and consolidated lung sounds were auscultated. These three cats were euthanized when moribund on day 9 or 10 after inoculation. Two of the three remaining cats developed mild to moderate clinical toxoplasmosis but recovered. This clinical course of T. gondii infection in the JSY3 clone-infected cats, including the high morbidity, was remarkably similar to that reported by Davidson et al. (7) for cats infected with the parent FIV-NCSU 1 biological isolate. Postmortem findings. Postmortem exams were performed on the three FIV-T. gondii-infected cats that were euthanized to confirm that their clinical disease was due to toxoplasmosis. Of the three cats necropsied, only one had gross evidence of interstitial pneumonia characterized by mottled red-tan-brown discoloration and firm lung parenchyma. The livers of all necropsied animals had few, scattered, pale tan, 1- to 3-mm-diameter foci of discoloration consistent with hepatic necrosis. In addition, the hearts of affected animals contained occasional

VOL. 70, 1996 PATHOGENESIS OF FIV MOLECULAR CLONE 3015 FIG. 5. PCR of positively sorted lymphocyte subsets (CD4 [4 ], CD8 [8 ], and B cells) from JSY3-infected cats sampled at 2 and 14 weeks p.i. A lysate representing 10 4 cells ( 95% purity) was amplified and analyzed by Southern blotting. The positive control ( ) is cell lysate from FIV-Black-infected CrFK cells, and the negative control ( ) is lysate from normal, uninfected FCD4E cells. The image was generated on a Macintosh computer using Adobe Photoshop software. foci of myocardial necrosis characterized by pale white streaks on epicardial and cut surfaces. Histologically, lesions were present in the lungs, livers, hearts, and brains of the three cats and were similar to those seen in cats with dual FIV-NCSU 1 T. gondii infection as described by Davidson et al. (7). The lesions were multifocal and characterized by an infiltration of macrophages and lymphocytes and lesser numbers of neutrophils. In the lungs, the alveoli were partially to completely filled with an admixture of macrophages, lymphocytes, intact and degenerate neutrophils, and abundant amounts of fibrin. In all the tissues, the most severe regions of inflammation contained a centralized region of necrosis and replacement by degenerate cells and cell debris. Except for the heart, T. gondii tachyzoites were seen in all tissues examined. The tachyzoites were never numerous but most conspicuous as clusters inside macrophages in the regions of severe inflammation and necrosis in the brain, lung, and liver. More-inconspicuous individualized, extracellular tachyzoites could be identified as well. When intracellular, tachyzoites were seen only in macrophages. DISCUSSION FIG. 6. Clinical signs of illness after T. gondii infection. Ten thousand tachyzoites were injected into JSY3-infected cats at 29 weeks after FIV infection (A), and six age- and sex-matched SPF cats were injected with only T. gondii as a control (B). Clinical status was examined daily and scored by an observer unaware of the FIV infection status of the animals. Each bar represents the total clinical disease score (as determined from Table 1) of an individual cat. Three cats (asterisks in panel A) were euthanized because of the severity of their clinical disease at 9 or 10 days post-t. gondii challenge. Clinical scores recorded after 10 days refer to the remaining three animals. None of the cats infected with only T. gondii had to be euthanized. Here, we describe the isolation of a full-length provirus molecular clone (JSY3) of the NCSU 1 isolate of FIV that retains the essential in vitro and in vivo biological characteristics of the parent virus. This clone was obtained from an EMBL3 lambda phage library made from FCD4E cells, and the intact genomic structure was confirmed by PCR comparison with the FIV-pPPR molecular clone. The JSY3 molecular clone recovered from FCD4E cells was highly infectious for PBMCs and FCD4E cells but failed to infect CrFK cells, thus retaining the tropism of the parent FIV-NCSU 1 virus. Miyazawa et al. (20) and Siebelink et al. (27) similarly reported that CD4 lymphoblastoid cell line MYA-1 cell-derived or bone marrow-derived molecular clones of FIV recovered from transfected CrFK cells failed to reinfect CrFK but retained their tropism for PBMC and CD4 cell cultures. Similarly, the PBMC-derived molecular clone FIV-pPPR replicated efficiently in PBMCs but did not infect adherent cells such as CrFK or G355-5 cells (26), whereas the FIV-p34 clone, derived from the CrFK-adapted Petaluma isolate, replicated efficiently in feline adherent cells, including CrFK cells, but inefficiently in PBMCs (28). The in vivo biological characteristics of the JSY3 molecular clone also are remarkably similar to those of the parent virus. Both viruses caused a significant inversion of the CD4 /CD8 ratio by 6 weeks p.i. As reported previously for a number of biological isolates of FIV (1, 33, 34), including the NCSU 1

3016 YANG ET AL. J. VIROL. isolate (11, 31), the inverted CD4 /CD8 ratio caused by the JSY3 clone was the result of a loss of CD4 lymphocytes and an increase in CD8 lymphocytes. Consistent with the NCSU 1 biological isolate, the JSY3 molecular clone caused a strong antibody response to gag and env antigens, and PBMCs had a high burden of FIV provirus during the acute-stage infection. As the alteration in CD4 /CD8 ratio and virus persistence despite a strong antibody response are hallmarks of the pathogenesis of both HIV and FIV infections, the JSY3 molecular clone fulfills an important criterion for an AIDS model. Interestingly, English et al. (10) showed that while the NCSU 1 isolate of FIV could be recovered from CD4, CD8, and B cells during the clinically asymptomatic stage of infection, the virus was found predominantly in CD4 lymphocytes during the acute-stage disease. The JSY3 molecular clone demonstrated an identical pattern of high provirus burden in CD4 cells during acute-stage infection, followed by a gradual shift to a panlymphotropic pattern during the transition from the acute to the asymptomatic stage of infection. The significance of the panlymphotropism of FIV in the pathogenesis of the virus is not known. However, it is important that despite being a reservoir for FIV infection, CD8 and B cell numbers remain normal or are elevated in FIV-infected cats, while there is a marked and progressive loss of CD4 cells. Perhaps the FIV infection in CD8 and B cells in vivo is latent and not cytolytic. Experiments are in progress to explore this possibility. Derivation of molecular clones of viruses from in vitro culture systems poses the risk of selection of some viral genotypes over others, as has been shown for HIV (5, 12, 19), or introduction of modifications in cultured virus, as has been demonstrated for SIVmac-adapted isolates (14, 16). For FIV, Sparger et al. (28) reported that the pf34 clone derived from the CrFK-adapted Petaluma isolate is less pathogenic than the parent Petaluma virus isolated from PBMCs. In contrast, the FIV-pPPR molecular clone derived from PPR-infected PB- MCs and the biological parent PPR isolate show similar pathogenicities, including virus burden in PBMCs and reduced CD4 /CD8 ratios (28). The JSY3 molecular clone described here also appears to have retained the essential biological characteristics of the parent isolate. This may be largely because we minimized the risk of culture-related artifacts by isolating FIV-NCSU 1 genomic DNA from FIV-inoculated CD4 lymphocytes (FCD4E cells). While the FCD4E cells have been cultured for several years in our laboratory, they are still interleukin-2 dependent and appear to express a normal rather than a transformed phenotype and thus represent as near as possible in vitro the primary in vivo target of FIV. These cloning conditions thus appear to be satisfactory for deriving a molecular clone representative of the virus population present in infected animals. The value of a molecular clone for studies of pathogenesis ultimately depends upon its ability to replicate the disease caused by its biological parent virus. Similarly to HIV (17), the NCSU 1 isolate of FIV causes an acute-stage clinical disease characterized by fever and lymphadenopathy that is transient in nature and resolves as the infection progresses to the clinically asymptomatic stage of infection (11). The JSY3 acutestage infection was also characterized by a fever and lymphadenopathy that was followed by a clinically asymptomatic stage. Davidson et al. (7) reported that cats infected with FIV- NCSU 1 become highly susceptible to a normally avirulent strain of T. gondii as early as 18 weeks post-fiv infection. We utilized this FIV-T. gondii dual infection to determine if infection with the NCSU 1 molecular clone JSY3 also caused an immunodeficiency early in the asymptomatic stage of infection. As reported previously for the biological parent FIV-NCSU 1 isolate, T. gondii infection of cats with prior JSY3 infection resulted in severe clinical disease. Three cats euthanized because of the severity of their disease were subsequently shown to have generalized toxoplasmosis by histopathology. Tissue lesions characterized by inflammatory infiltrates and intracellular T. gondii tachyzoites were similar to those described for cats infected with the biological parent NCSU 1 and challenged with T. gondii (7). We interpret these observations as indicating that the JSY3 molecular clone causes a major impairment in the immune response, resulting in enhanced susceptibility to secondary infection by T. gondii. Thus, the JSY3 molecular clone of FIV possesses all of the biological characteristics of the parent NCSU 1 isolate, including induction of immunodeficiency. The availability of the pathogenic JSY3 clone will make it possible to address in an animal model important questions regarding the determinants of pathogenesis. When the JSY3 clone is sequenced (in progress), it will be possible to analyze the determinants for viral virulence, to understand the molecular basis of biological variation and tissue tropism, and to develop effective vaccines. As FIV and HIV share many structural and biological characteristics, knowledge gained with the FIV model may prove useful to our understanding of HIV pathogenesis and to the development of effective treatments. ACKNOWLEDGMENTS We thank Deborah Anderson, Tedd Childers, Andy Bruce, James Rottman, and Rance Sellon for assistance and Fred Fuller, Mark Conkling (Department of Genetics, North Carolina State University), and Stan Lilleberg (Fort Dodge Laboratories, Fort Dodge, Iowa) for helpful advice. We also thank Tom Phillips (Scripps Research Institute, La Jolla, Calif.) for the FIV-PPR plasmid clone and Mark Tompkins (Emory University, Atlanta, Ga.) for advice. This work was supported by Public Health Service grant NO1 AI 35515 from the NIAID-DAIDS and by a grant from Fort Dodge Laboratories. REFERENCES 1. Ackley, C. D., J. K. Yamamoto, N. Levy, N. C. Pedersen, and M. D. Cooper. 1990. Immunologic abnormalities in pathogen-free cats experimentally infected with feline immunodeficiency virus. J. Virol. 64:5652 5655. 2. Beebe, A. M., N. Dua, T. G. Faith, P. F. Moore, N. C. Pedersen, and S. Dandekar. 1994. Primary stage of feline immunodeficiency virus infection: viral dissemination and cellular targets. J. Virol. 68:3080 3091. 3. Brown, W. C., L. Bissey, D. S. Logan, N. C. Pedersen, J. H. Elder, and E. W. Collisson. 1991. Feline immunodeficiency virus infects both CD4 and CD8 T lymphocytes. J. Virol. 65:3359 3364. 4. Callanan, J. J., I. A. McCandlish, B. O Neil, C. E. Lawrence, M. Rigby, A. M. Pacitti, and O. Jarrett. 1992. Lymphosarcoma in experimentally induced feline immunodeficiency virus infection. Vet. Rec. 130:293 295. 5. Dahl, K., K. Martin, and G. Miller. 1987. Differences among human immunodeficiency virus strains in their capacities to induce cytolysis or persistent infection of a lymphoblastoid cell line immortalized by Epstein-Barr virus. J. Virol. 61:1602 1608. 6. Davidson, M. G., M. R. Lappin, R. V. English, and M. B. Tompkins. 1993. A feline model of ocular toxoplasmosis. Invest. Ophthalmol. Visual Sci. 34:3653 3660. 7. Davidson, M. G., J. Rottman, R. V. English, M. R. Lappin, and M. B. Tompkins. 1993. Feline immunodeficiency virus predisposes cats to acute generalized toxoplasmosis. Am. J. Pathol. 143:1486 1497. 8. Diehl, L. J., C. K. Mathiason-DuBard, L. L. O Neil, and E. A. Hoover. 1995. Longitudinal assessment of feline immunodeficiency virus kinetics in plasma by use of a quantitative competitive reverse transcriptase PCR. J. Virol. 69:2328 2332. 9. English, R. V., M. B. Davidson, M. P. Nasisse, V. E. Jamieson, and M. R. Lappin. 1990. Intraocular disease associated with feline immunodeficiency virus infection in cats. J. Am. Vet. Med. Assoc. 196:1116 1119. 10. English, R. V., C. M. Johnson, D. H. Gebhard, and M. B. Tompkins. 1993. In vivo lymphocyte tropism of feline immunodeficiency virus. J. Virol. 67: 5175 5186. 11. English, R. V., P. Nelson, C. M. Johnson, M. Nasisse, W. A. Tompkins, and M. B. Tompkins. 1994. Development of clinical disease in cats experimen-

VOL. 70, 1996 PATHOGENESIS OF FIV MOLECULAR CLONE 3017 tally infected with feline immunodeficiency virus. J. Infect. Dis. 170:543 552. 12. Evans, L. A., T. M. McHugh, D. P. Stites, and J. A. Levy. 1987. Differential ability of human immunodeficiency virus isolates to productively infect human cells. J. Immunol. 138:3415 3418. 13. Goff, S., P. Traktman, and D. Baltimore. 1981. Isolation and properties of Moloney murine leukemia virus mutants: use of a rapid assay for release of virion reverse transcriptase. J. Virol. 38:239 248. 14. Hirsch, V. M., P. Edmondson, M. Murphey-Corb, B. Arbeille, P. R. Johnson, and J. I. Mullins. 1989. SIV adaption to human cells. Nature (London) 341:767 768. 15. Hoffmann-Fezer, G., J. Thum, C. Ackley, M. Herbold, J. Mysliwietz, S. Thefeld, K. Hartmann, and W. Kraft. 1992. Decline in CD4 cell numbers in cats with naturally acquired feline immunodeficiency virus infection. J. Virol. 66:1484 1488. 16. Kodama, T., D. P. Wooley, Y. M. Naidu, H. W. Kestler III, M. D. Daniel, Y. Li, and R. C. Desrosiers. 1989. Significance of premature stop codons in env of simian immunodeficiency virus. J. Virol. 63:4709 4714. 17. Levy, J. A. 1993. Pathogenesis of human immunodeficiency virus infection. Microbiol. Rev. 57:183 289. 18. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 19. Meyerhans, A., R. Cheynier, J. Albert, M. Seth, S. Kwok, J. Sninsky, L. Morfeldt-Manson, B. Asjo, and S. Wain-Hobson. 1989. Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolation. Cell 58:901 910. 20. Miyazawa, T., M. Fukasawa, A. Hasegawa, N. Maki, K. Ikuta, E. Takahashi, M. Hayami, and T. Mikami. 1991. Molecular cloning of a novel isolate of feline immunodeficiency virus biologically and genetically different from the original U.S. isolate. J. Virol. 65:1572 1577. 21. Novotney, C., R. English, J. Housman, M. Davidson, M. Nasisse, C. R. Jeng, W. Davis, and M. Tompkins. 1990. Lymphocyte population changes in cats naturally infected with feline immunodeficiency virus. AIDS 4:1213 1218. 22. O Connor, T. J., Jr., S. Tanguay, R. Steinman, R. Smith, M. C. Barr, J. K. Yamamoto, N. C. Pedersen, P. R. Andersen, and Q. J. Tonelli. 1989. Development and evaluation of immunoassay for detection of antibodies to the feline T-lymphotropic lentivirus (feline immunodeficiency virus). J. Clin. Microbiol. 27:474 479. 23. Olmsted, R. A., A. K. Barnes, J. K. Yamamoto, V. M. Hirsch, R. H. Purcell, and P. R. Johnson. 1989. Molecular cloning of feline immunodeficiency virus. Proc. Natl. Acad. Sci. USA 86:2448 2452. 24. Pedersen, N. C., E. Ho, M. L. Brown, and J. K. Yamamoto. 1987. Isolation of a T-lymphotropic virus from domestic cats with an immunodeficiency-like syndrome. Science 235:790 793. 25. Pedersen, N. C., J. K. Yamamoto, T. Ishida, and H. Hansen. 1989. Feline immunodeficiency virus infection. Vet. Immunol. Immunopathol. 21:111 129. 26. Phillips, T. R., R. L. Talbott, C. Lamont, S. Muir, K. Lovelace, and J. H. Elder. 1990. Comparison of two host cell range variants of feline immunodeficiency virus. J. Virol. 64:4605 4613. 27. Siebelink, K. H. J., I.-H. Chu, G. F. Rimmelzwaan, K. Weijer, A. D. M. E. Osterhaus, and M. L. Bosch. 1992. Isolation and partial characterization of infectious molecular clones of feline immunodeficiency virus obtained directly from bone marrow DNA of a naturally infected cat. J. Virol. 66:1091 1097. 28. Sparger, E. E., A. M. Beebe, N. Dua, S. Himathongkam, J. H. Elder, M. Torten, and J. Higgins. 1994. Infection of cats with molecularly cloned and biological isolates of the feline immunodeficiency virus. Virology 205:546 553. 29. Talbott, R. L., E. E. Sparger, K. M. Lovelace, W. M. Fitch, N. C. Pedersen, P. A. Luciw, and J. H. Elder. 1989. Nucleotide sequence and genomic organization of feline immunodeficiency virus. Proc. Natl. Acad. Sci. USA 86:5743 5747. 30. Tompkins, M. B., D. H. Gebhard, H. R. Bingham, M. J. Hamilton, W. C. Davis, and W. A. F. Tompkins. 1990. Characterization of monoclonal antibodies to feline T lymphocytes and their use in the analysis of lymphocyte tissue distribution in the cat. Vet. Immunol. Immunopathol. 26:305 317. 31. Tompkins, M. B., P. D. Nelson, R. V. English, and C. Novotney. 1991. Early events in the immunopathogenesis of feline retrovirus. J. Am. Vet. Med. Assoc. 199:1311 1315. 32. Tompkins, M. B., G. K. Ogilvie, R. A. Franklin, K. W. Kelley, and W. A. Tompkins. 1987. Induction of IL-2 and lymphokine activated killer cells in the cat. Vet. Immunol. Immunopathol. 16:1 10. 33. Torten, M., M. Franchini, J. E. Barlough, J. W. George, E. Mozes, H. Lutz, and N. C. Pedersen. 1991. Progressive immune dysfunction in cats experimentally infected with feline immunodeficiency virus. J. Virol. 65:2225 2230. 34. Willett, B. J., M. J. Hosie, J. J. Callanan, J. C. Neil, and O. Jarrett. 1993. Infection with feline immunodeficiency virus is followed by the rapid expansion of a CD8 lymphocyte subset. Immunology 78:1 6. 35. Yamamoto, J. K., E. Sparger, E. W. Ho, P. R. Andersen, T. P. O Connor, C. P. Mandell, L. Lowenstine, R. Munn, and N. C. Pedersen. 1988. Pathogenesis of experimentally induced feline immunodeficiency virus infection in cats. Am. J. Vet. Res. 49:1246 1258.