In Vivo Lymphocyte Tropism of Feline Immunodeficiency Virus
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
JOURNAL OF VIROLOGY, Sept. 1993, p. 5175-5186 Vol. 67, No. 9
0022-538X/93/095175-12$02.00/0
Copyright © 1993, American Society for Microbiology
In Vivo Lymphocyte Tropism of Feline
Immunodeficiency Virus
ROBERT V. ENGLISH, CALVIN M. JOHNSON, DOUGLAS H. GEBHARD,
AND MARY B. TOMPKINS*
Department ofMicrobiology, Pathology, and Parasitology, College of Veterinary Medicine,
North Carolina State University, Raleigh, North Carolina 27606
Received 8 February 1993/Accepted 26 May 1993
Feline immunodeficiency virus (FlY) infection in the cat is similar to human immunodeficiency virus type 1
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
infection in causing a selective reduction in CD4+ cell numbers, leading to inversion of the CD4+/CD8+ ratio. To
determine whether FIV, similar to human immunodeficiency virus type 1, has a tropism for CD4+ cells, we
examined the in vitro and in vivo susceptibilities of feline lymphocyte subpopulations to FIV infection. Infection
of interleukin-2-dependent CD4+ or CD8+ lymphocyte cultures with the NCSU1 isolate of FIV (FIV-NCSU1)
resulted in syncytium formation, cell death, and Mg2e-dependent reverse transcriptase (RT) activity in both
cases. Monoclonal antibodies to feline lymphocyte subsets were used to sort peripheral blood mononuclear cells
from FIV-infected cats into highly (>95%) purified CD4+ cell, CD8+ cell, immunoglobulin-positive (Ig+) cell,
and monocyte subpopulations. The mononuclear cell subpopulations were analyzed for FlV provirus by
polymerase chain reaction and Southern blot analysis and for virus expression by RT activity. All 16 cats infected
with F1V-NCSU1 demonstrated FIV provirus in CD4+ cell-, CD8+ cell-, and Ig+ cell-enriched lymphocyte
populations. Southern blot detection of amplified gag gene sequences and limiting-cell-dilution polymerase chain
reaction analysis indicated that Ig+ cells carried a higher FlV provirus burden in chronically (21-year) infected
cats than either CD4+ or CD8+ cells. In contrast, CD4+ cells carried the greatest provirus burden in acutely (2-
to 4-week) infected cats. FIV provirus was detected in monocytes from only 1 of 10 cats with asymptomatic
infection. Addition of culture supernatants from enriched CD4+, CD8+, and Ig+ cells from FIV-infected cats to
an FlV-susceptible CD4+ lymphocyte culture resulted in syncytium formation, cell death, and RT activity.
Infection of Ig+ cells is not unique to FIV-NCSU,, as lymphocyte subpopulations from other cats with natural
infections and cats infected with the Petaluma or Mount Airy isolate of FIV demonstrated a similar distribution
of FIV provirus and RT activity. These data suggest that FlY possesses a broad tropism for peripheral blood
mononuclear cells and that an Ig+ cell may serve as a major reservoir for the virus in chronically infected cats.
Information obtained to date on the pathogenesis of feline vivo. We recently showed that cats naturally infected with
immunodeficiency virus (FIV) infection suggests that it is a FIV have a T-cell lymphopenia, characterized by a decrease
valuable animal model for human immunodeficiency virus in CD4+ cells, and an inversion of the CD4+/CD8+ ratio
type 1 (HIV-1)-induced AIDS. HIV-1 and FIV belong to the (21). In addition, we (30) and others (1, 33) have reported
lentivirus subfamily of retroviruses and have similar mor- that cats experimentally infected with FIV develop an in-
phologies, protein compositions, and Mg2+-dependencies of verted CD4+/CD8+ ratio characterized by a decrease in
their reverse transcriptases (RT) (23, 24). Both viruses infect CD4+ cell numbers and, in some cases, an increase in CD8+
T lymphocytes and macrophages and are capable of inducing cell numbers (1, 30). B-cell numbers, on the other hand, are
these cells to form syncytia (3, 10, 24). HIV-1 displays a unaffected in natural (21) or experimental (1, 30) FIV infec-
particular tropism for CD4+ lymphocytes, which leads to tions. To further address the issue of FIV cell tropism and
their gradual depletion and an inversion of the CD4+/CD8+ how it may relate to the observed in vivo changes in
ratio (5, 17). The pathogenesis of HIV-1 infection has been lymphocyte populations, we examined the susceptibilities of
attributed to virus-induced reduction of CD4+ lymphocyte feline T-cell subsets to in vitro infection with the NCSU1
numbers and functions, resulting in decreased immune respon- strain of FIV (FIV-NCSU1) and analyzed B and T cells from
siveness and subsequent severe secondary infections (17). FIV-infected cats for the presence of FIV provirus. We
There have been limited studies reported on the cell report herein that both CD4+ and CD8+ cells are susceptible
tropism of FIV. Brown et al. (2) recently reported that both to FIV infection in vitro. Similar to the in vitro susceptibil-
CD4+ and CD8+ lymphocytes can be infected in vitro with ity, both T-cell subsets are infected in vivo in naturally
the Petaluma and the PPR isolates of FIV. In vivo, FIV has infected cats and in cats experimentally infected with FIV.
been detected in lymphocytes (23), peritoneal macrophages In addition, immunoglobulin-positive (Ig+) cells from all the
(3), and central nervous system tissue (7). However, there cats examined are also infected with FIV. Moreover, in the
have been no reports addressing in vivo peripheral blood case of cats with chronic, asymptomatic infections, the
mononuclear cell (PBMC) subset tropism of this virus. provirus burden in peripheral blood lymphocytes appears to
Lymphocyte subset alterations similar to those seen in be greater in the Ig+ cell population.
HIV infections have been reported to occur in FIV infec-
tions, suggesting that there may be a selective cell tropism in
MATERIALS AND METHODS
Virus and in vitro infection. FIV-NCSU1, isolated from a
*
Corresponding author. naturally infected, feline leukemia virus-seronegative cat,
51755176 ENGLISH ET AL. J. VIROL.
was used for the in vitro infection studies and has been (Perfix; Fischer Scientific, Orangeburg, N.Y.) before indi-
described previously (30). An NCSU1 virus stock for in vitro rect fluorescent-antibody staining or flow cytometry.
studies was prepared by coculturing PBMCs from this cat Separation of lymphocyte subsets from whole blood. Per-
with PBMCs from a normal specific-pathogen-free (SPF) cat. coll-purified PBMCs (31) (107) were incubated with M-450
Coculture supernatant with high RT activity (2 x 105 cpm/ beads (Dynal, Great Neck, N.Y.) covalently coated with
ml) was filtered (0.2 ,um) and stored at -135°C. For in vitro polyclonal goat anti-cat IgG heavy- and light-chain-specific
infection, 200 ,ul of virus stock was added to 106 cells in 1 ml antibody (KPL, Gaithersburg, Md.) for 30 min at 4°C at a
of medium and incubated for 24 h at 37°C. Cells were then bead-to-cell ratio of 10:1. Bead-bound Ig+ cells were re-
washed twice and resuspended in fresh culture medium. moved magnetically, resuspended in medium, gently agi-
Cell cultures. An interleukin 2 (IL-2)-dependent, CD4+ tated to release any trapped nonbound cells, and then
lymphocyte culture (FCD4-D) has been established in our magnetically reisolated. The remaining PBMCs were then
laboratory through long-term culture of PBMCs from an SPF incubated for 30 min at 4°C with goat anti-mouse M-450
cat in the presence of recombinant human IL-2 (kindly beads coated with the anti-cat CD8 MAb at a bead-to-cell
provided by Hoffman-LaRoche, Nutley, N.J.). These cells ratio of 3:1, and the bead-bound cells were removed mag-
are 100% positive for the feline pan-T-cell marker 1.572 (29), netically as described for Ig+ cells. Highly purified CD4+
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
60 to 65% positive for the feline CD4 homolog as recognized cells were obtained from the remaining CD8 cell- and Ig cell-
by antibody CAT30A (29), and negative for the CD8 ho- depleted population by a two-color fluorescence activated cell
molog recognized by antibody 3.357 (29). We have also sorting (FACS) procedure. Cells were stained with biotiny-
established a CD8+ (FCD8-CH) lymphocyte culture. The lated anti-cat CD4 MAb, FITC-conjugated goat anti-cat IgG
CD8+ culture is 100% positive for the pan-T-cell marker, 95 heavy- and light-chain-specific antibody, and FITC-conju-
to 98% positive for the CD8 homolog, and negative for the gated anti-cat CD8. After the addition of phycoerythrin-
CD4 homolog. Cells were maintained in RPMI supplemented conjugated strepavidin, cell sorting for phycoerythrin-posi-
with 10% fetal bovine serum and either 40 U (FCD4-D cells) tive, FITC-negative cells was performed. In cases when
or 100 U (FCD8-CH cells) of recombinant human IL-2 per monocytes were purified as well, these cells were removed
first by using goat anti-mouse beads coated with anti-cat
ml. monocyte MAb. The remaining cells were then purified as
FIV-infected cats. Lymphocyte subsets from both experi- described above.
mentally and naturally FIV-infected cats were analyzed in Genomic DNA from a fraction of the sorted lymphocytes
this study. The group of experimentally infected cats con- was immediately analyzed for FIV provirus by polymerase
sisted of 16 cats infected with FIV-NCSU1 (13 SPF and 3 chain reaction (PCR) and Southern blotting techniques. The
random-source cats). Lymphocyte subsets from two cats remaining cells were placed in culture (see below) to allow
infected with the Petaluma isolate and one cat infected with time for release of magnetic beads. By 6 days of culture, all
the Mount Airy isolate of FIV were also analyzed. All of the sorted lymphocyte populations had released magnetic beads
experimentally infected cats were clinically normal at the and reexpressed their surface markers as determined by flow
time of analysis. Two naturally infected cats were also cytometry.
analyzed, one of which was the original source of the Culture and virus detection in sorted lymphocytes from
NCSU1 isolate (cat TH). The other naturally infected cat FWY-infected cats. Sorted lymphocytes were cultured in
was the only cat with clinical disorders, consisting of chronic 24-well plates (Costar, Cambridge, Mass.) at a concentration
gingivitis, vitritis, and focal retinal degeneration. FIV infec- of 2 x 106 cells per ml of RPMI medium supplemented with
tion was confirmed in all cats by Western immunoblot 30% supernatant from concanavalin A-stimulated feline sple-
analysis of antibody to FIV proteins (21). All cats were nocytes, 30% conditioned supernatant from FCD4-D cells,
negative for FeLV by enzyme-linked immunosorbent assay 20% fetal bovine serum, and 100 U of recombinant human
(ELISA) (IDEXX, Portland, Maine). IL-2 per ml. This medium promoted survival of B cells in
MAbs and flow cytometry. Monoclonal antibodies (MAbs) culture for 7 days. After 48 h of culture, a fraction of the
to the feline CD4 homolog (CAT 30A), CD8 homolog (3.357), nonadherent cells was collected for analysis of population
and pan-T-cell marker (1.572) have been previously de- purity (by flow cytometry) and provirus detection (by PCR
scribed (29). A fluorescein isothiocyante (FITC)-conjugated and Southern blotting). The remaining nonadherent, sorted
F(ab')2 fragment of goat anti-cat IgG heavy- and light-chain- cells (2 x 106/ml) were placed in fresh supernatant-enriched
specific antibody (KPL, Gaithersburg, Md.) was used to RPMI medium and cultured for an additional 4 days, at
identify feline B cells. An anti-canine B-cell MAb (B5), which time flow cytometric analysis was repeated. All cul-
developed in the Flow/Hybridoma Laboratory at North tured cell populations retained their high degree of purity
Carolina State University and shown to cross-react with (>95%). Culture supernatants from these 4-day cultures
feline B cells, was also used. Monocytes were quantitated by were harvested to assay for virus production, and nonadher-
using a neutrophil/monocyte-specific MAb (CM277) de- ent cells were harvested for provirus detection.
scribed by Whitehurst et al. (34). Cells for flow cytometric Virus production was measured by using FCD4-D cells as
analysis were incubated for 30 min at 4°C with either FITC- targets of infection. FCD4-D cells (2 x 105) were treated
or biotin-conjugated antibodies, washed, and analyzed on a with DEAE-dextran (25 ,ug/ml) for 30 min and then washed
FACScan (Becton Dickinson, Mountain View, Calif.). When and incubated with 100 ,ul of filtered (0.45 ,um), 4-day
biotinylated MAbs were used, cells were incubated with supernatant from sorted lymphocyte cultures. After 24 h, the
phycoerythrin-conjugated strepavidin (Serotec, Oxford, En- cells were washed and resuspended in 100 ,ul of fresh culture
gland) for 15 min at 4°C, washed, and then analyzed. Plasma medium, and 10 pl of supernatant was collected every 3 days
from FIV-infected SPF cats was preadsorbed to normal to assay for RT activity.
feline splenocytes and used at a dilution of 1:500 to detect RT assay. The assay for Mg2 -dependent RT activity was
FIV antigens in infected cells by flow cytometry or indirect performed as described previously (21) and is a modification
fluorescent-antibody staining. Cytoplasmic FIV antigens of the procedure of Goff et al. (11).
were detected by first fixing cells in 2% paraformaldehyde Detection of FIV provirus by PCR and Southern blotVOL. 67, 1993 CELL TROPISM OF FIV 5177
analysis. PCR amplification and Southern blot hybridization 20
using primers and probes specific for the p26 region of the
gag open reading frame were used to detect FIV provirus.
Initial studies used purified DNA extracted from cells as
described by Psallidopoulos et al. (25). Cells (1 x 106 to 2 x 16
106) from each sorted lymphocyte population were digested.
Because these cats had CD4+ lymphopenias, 106 CD4+ cells
were not always available for digestion after PBMC sorting.
To decrease the loss of DNA during extraction, uninfected
FCD4-D cells were added to sorted CD4+ cells as needed to
achieve a total of 106 cells for digestion. When this was
necessary, the other sorted populations were equally diluted
with FCD4-D cells. In later studies, the loss of DNA from
the digestion of low cell numbers was avoided by using cell
lysates. Cells (2 x 105) were digested in 100 ,ul of digestion
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
buffer (1 x PCR buffer with 60 ,ug of proteinase K per 100 ,ul) 0.
at 56°C for 2 h and then boiled for 10 min (8). Equal amounts
of DNA or cell lysate from each sorted lymphocyte popula-
tion from a single cat were always amplified, gel electro-
phoresed, blotted, hybridized, and developed simulta-
neously. Because of extremely low monocyte numbers
(usually c5 x 104 total cells recovered), when monocytes
were analyzed the amount of purified DNA was usually half
of that from the lymphocytes.
Two sets of primers for the PCR were selected from a o2
1 6 12 is is
published FIV sequence (28). A 583-bp segment was ampli-
fied from the gag p26 region by using a minus-strand primer DAYS POST INOCLLATION
(FIV-1, 5'-CCACAATATGTAGCACTTGACC) and a plus- FIG. 1. Mg2+-dependent RT activity in cultured CD4+ and
strand primer (FIV-2, 5'-GGGTACTTlCTGGCITAAG CD8+ T cells infected in vitro with FIV-NCSU,. FCD4-D and
GTG). To allow double amplification with nested primers, FCD8-CH cells were infected with FIV-NCSU, as described in
we used primers FIV-7 (5'-TGACGGTGTCTACTGCTGCT) Materials and Methods, and culture supernatants were assayed for
and FIV-8 (5'-CACACTGGTCCTGATCC`I"TTT), which RT activity at various times postinfection. Numbers represent the
amplify an 838-bp gag segment which includes the region means and standard deviations of three replicates. RT activity in
amplified by FIV-1 and FIV-2. PCR amplification was per- uninfected FCD4-D and FCD8-CH cells was < 130 -+ 26 cpm.
formed with either 1.0 tg of sample DNA or 50 ,ul of cell
lysate (105 cells) and the Gene Amp kit under previously
determined optimal conditions for these primers and tem- cells by day 3 in both cultures and in 80% of the cells by day
plate sequence (0.1 ,um each primer, 1.5 mM MgCl2, and 2.0 7 in both cultures (data not shown). Syncytia developed by 7
U of Taq polymerase per 100-pl reaction mixture). An days and were followed by progressive cell lysis leading to
amplification cycle of 95°C for 45 s, 59°C for 45 s, and 72°C almost complete cell death in both cultures by 24 days
for 45 s was repeated 35 times and was followed by a 7-min, postinfection. Because the FCD4-D culture contained both
72°C extension cycle. Ten microliters of amplified product CD4+ and CD4- (T-null) T cells, two-color flow cytometric
was run on a 1.2% agarose gel, which was then transferred to analysis was performed to determine which population was
a nitrocellulose membrane (Bio-Rad). Membranes were hy- infected. With both fixed and unfixed cell preparations, FIV
bridized for 12 to 18 h at 420C with two 32P-terminally antigens were detected in both the CD4+ and CD4- cells in
labelled, internally located oligonucleotides (5'-TACTGAC the FCD4-D culture at 4 days postinoculation. While HIV-1
ATGGCCACATT and 5'-GCTCCCAGACCATTACCCTA). infection has been shown to down regulate CD4 expression
After three washes in lx SSC (0.15 M NaCl plus 0.015 M in T cells (5, 13), the percentage and intensity of CD4
sodium citrate)-1% sodium dodecyl sulfate at 370C, blots expression were similar in infected and uninfected FCD4-D
were exposed to radiographic film for 2 to 24 h. These cultures. Thus, FIV-NCSU, is able to productively infect
primers and probes did not recognize any sequences from CD4-/CD8- T-null cells as well as CD4+ and CD8+ T
normal cat PBMCs, from FCD4-D or FCD8-CH cells, or lymphocytes, leading to syncytium formation and cell death.
from feline leukemia virus-infected FL-74 cells. By limiting- The experiment depicted in Fig. 1 was repeated five times
cell-dilution PCR and Southern analysis, these primers are with similar results.
able to detect provirus in as few as 100 PBMCs from Purification of peripheral blood lymphocyte subsets from
FIV-infected, asymptomatic cats in a background of 105 FIV-infected cats. To explore in vivo virus tropism, we
uninfected PBMCs. wished to obtain highly purified subpopulations of lympho-
cytes. Because FIV-infected cats may have lower lympho-
RESULTS cyte counts than normal cats (24), we used magnetic bead
sorting whenever possible to maximize our yields. This
In vitro infection of lymphocyte cultures. FCD4-D (CD4+) method of sorting yielded subpopulations of lymphocytes of
cells and FCD8-CH (CD8+) cells inoculated with FIV- >95% purity for CD8+ and Ig+ cells. The purity of CD4+
NCSU1 revealed magnesium-dependent RT activity by day cells by magnetic bead sorting was extremely low. There-
4, with peak RT activity seen between days 7 and 10 fore, we obtained CD4+ cells by FACS, which produced a
postinfection (Fig. 1). FIV antigen was detected by indirect pure population (> 95 %) but lower total cell numbers. Purity
fluorescent-antibody staining in approximately 50% of the of the sorted populations was determined by flow cytometric5178 ENGLISH ET AL. J. VIROL.
TABLE 1. Purity of sorted lymphocyte subpopulations from
three different cats 0 st
Q 2
cc
o 0 +
- CL C) _3 u X
fori:
6
Cat Sorted % of cells staining positive
population' Ig Pan T CD4 CD8
1 Unsortedc
Ig+
29.4
95.3
66.2
NDd
21.9
0.2
43.5
0.8 0 0 0 0
AL 4-W'
CD8+ 0.0 ND 1.5 95.1
CD4+ 0.0 ND 98.4 0.0
2 Unsortedc 35.1 64.6 33.6 26.0 FIG. 2. Southern blots of PCR-amplified gag genes from FIV
Ig+ 96.0 2.0 0.0 2.2 provirus lymphocyte subsets from two cats 15 months after exper-
CD8+ 1.5 98.0 3.2 95.3 imental infection with FIV-NCSU1. Lymphocytes were cultured for
CD4+ 2.4 90.5 95.0 1.3 48 h, and nonadherent cells were collected for analysis. For all
samples, PCR was performed with gag primers FIV-1 and FIV-2
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
TH Unsortedc 54.5 33.9 9.4 23.3 and 1 ,ug of extracted DNA. Lanes contained DNA from CD8+ cells
Ig+ 97.5 ND 0.0 0.0 (CD8), Ig+ cells (IG), CD4+ cells (CD4), unsorted PBMCs (PBM),
CD8+ 0.0 ND 3.0 96.4
CD4+ 0.0 ND 98.0 0.0 FIV-NCSU,-infected FCD4-D cells (+), and PBMCs from an unin-
fected control cat (-). Bands represent DNA segments of approxi-
a Lymphocytes were sorted into Ig+, CD8+, and CD4+ populations as mately 580 bp as determined by molecular weight standards on the
described in Materials and Methods. The sorted populations were then original gels.
cultured for 6 days prior to flow cytometric analysis for purity.
b The percentage of positive cells was determined by flow cytometric
analysis with MAbs as described in Materials and Methods.
c These percentages are typical of those for FIV-asymptomatic SPF cats.
The distribution of PBMCs in normal SPF cats is 30% Ig+, 64% pan-T, 39% analysis of lymphocyte populations immediately after sort-
CD4+, and 17% CD8+ cells (means calculated from 36 age-matched controls). ing (see Fig. 3 and 7) and after 6 days of culture (not shown)
Monocytes comprise 0 to 6% of the PBMCs of normal cats (12). yielded similar results, indicating that the relative levels of
d ND, not determined. provirus in different lymphocyte subpopulations were not an
artifact of culture. PCR amplification and Southern blot
analysis of sorted PBMCs from normal cats were negative
analysis, which necessitated that the CD8+ and Ig+ cells be for FIV provirus (data not shown).
free of magnetic beads. The CD8+ population released their To further evaluate the level of provirus in the Ig+ cell
magnetic beads after 48 to 72 h in culture, and sequential population, digestion, PCR amplification, and Southern blot-
flow cytometric analysis revealed complete reexpression of ting were carried out with serial 10-fold dilutions of sorted
the CD8 molecule by 5 days. The Ig+ cell population was lymphocytes from two of the long-term-infected SPF cats.
devoid of surface Ig or attached beads after 24 h in culture The total number of cells digested at each dilution was kept
and had reexpressed Ig by 48 h in culture. Flow cytometric constant by adding an appropriate number of uninfected
analysis of the CD4+ population was performed immediately FCD4-D cells. One microgram of DNA was amplified from
after FACS. The purity of sorted populations from all cats each dilution. Provirus was detectable in as few as 103 Ig+
used in this study was determined prior to PCR and Southern cells, whereas no provirus was detected with fewer than 106
blot analysis and was always greater than 95%. Table 1 lists CD4+ or CD8+ cells (data not shown).
the results of three typical lymphocyte subpopulation puri- Detection of FIV provirus in lymphocyte subsets from
fications from three FIV-infected cats. acutely infected cats. Long-term-FIV-infected cats produce
Detection of FIV provirus in lymphocyte subsets from high levels of antibody to FIV antigens (22). Because antif-
chronically infected cats. To determine the in vivo tropism of eline Ig was used to purify the Ig+ cell population, it is
FIV, we analyzed lymphocyte subsets from seven SPF and possible that infected T cells coated with anti-FIV antibody
three random-source asymptomatic cats chronically infected were sorted as Ig+ cells and are responsible for the positive
with FIV-NCSU1 (1 year postinfection) for the presence of signal in the Ig+ population. Therefore, lymphocyte sorting
provirus by PCR and Southern blot hybridization. At the was also performed for six acutely infected cats between 2
time of analysis, all 10 cats had markedly low CD4+/CD8+ and 4 weeks postinfection. At this time they were negative,
ratios because of decreased CD4+ cell numbers and in- by ELISA, for antibodies to FIV, and their CD4+ cell
creased CD8+ cell numbers (30). Ig+ cell numbers were numbers had not decreased significantly compared with
similar to those of age-matched, uninfected controls. Lym- controls. As with the chronically infected cats, FIV provirus
phocytes were separated into CD8+, CD4+, and Ig+ cell was detected in CD4+, CD8+, and Ig+ lymphocytes (Fig.
populations, and equal amounts of extracted DNA from each 3A). However, the relative level of provirus in the sorted
sorted cell population were analyzed for the presence of FIV populations was different. CD4+ cells contained the highest
provirus. PCR amplification and Southern blot analyses of level of provirus, on the basis of blot intensity, while proviral
cells from two of these cats (cat 1 and cat 2) are shown in levels in the Ig+ population were proportionally less than
Fig. 2. In support of the in vitro results, FIV provirus was that seen in chronically infected cats.
detected in both the CD4+ and CD8+ cells. Interestingly, Lymphocytes from these cats were collected and sorted
provirus was also detected in the Ig+ cell population. In- again between 11 and 15 weeks postinfection. At this time
deed, on the basis of signal intensity, the greatest amount of the cats had developed an antibody response to FIV, and
provirus appears to be in the Ig+ cell population. This their CD4+/CD8+ ratios were less than 1 because of both a
pattern of FIV provirus distribution in the three lymphocyte significant decrease in CD4+ cells and a significant increase
subpopulations was the same in all 10 chronically infected in CD8+ cells. As shown in Fig. 3B, the proviral burden had
cats that were analyzed. The data shown in Fig. 2 were shifted from the CD4+ cells to the Ig+ cells, presenting the
obtained from cells after 2 days in culture. However, PCR same picture as in the chronically infected cats. In spite ofVOL. 67, 1993 CELL TROPISM OF FIV 5179
A CAT Al CAT K4
antibody CM277 (34) and flow cytometry to evaluate mono-
cyte contamination of sorted lymphocyte populations from
2co D I the FIV-infected cats used for in Fig. 2. While monocytes
M co IV
represented 0 to 6% of total, unsorted PBMCs, which is
; a) I+ X 0) -
within the normal range for adult cats (12), all three postsort,
nonadherent populations were negative for monocytes.
We also used the CM277 MAb (34) and magnetic beads to
m0 mo purify monocytes from PBMCs and analyze the DNA for
provirus. The percent monocytes in the peripheral blood of
normal adult cats ranges from 0 to 5%, with a mean of 2% of
. U,.
total white cells (12). This converts to a mean of 6% of the
total unsorted PBMCs (30% of total white cells are lympho-
cytes [12]). Therefore, the yield of purified monocytes, even
when magnetic beads were used to sort, was extremely low.
With 5 x 104 purified monocytes (the maximum number
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
B CAT Al CAT K4
obtained), FIV provirus was detected in only 1 of 10 cats
examined. However, in all 10 cats, the remaining lympho-
C Oin0 cyte subpopulations (purified after the monocytes had been
co a (.7-C)
XL (.
c I + X O c) I + removed) retained provirus signals. A Southern blot from
one of these cats is shown in Fig. 4A. These data indicate
that it is unlikely that monocytes are a source of provirus
signal in any of the enriched lymphocyte populations, includ-
ing the Ig+ cell population.
The Ig+, provirus-positive cell is recognized by a MAb that
binds a subpopulation of feline B cells. Although two-color
analysis indicates that the Ig+ population does not contain T
cells, it is possible that there is T-cell or monocyte contam-
ination that is below the detection level of flow cytometry. It
FIG. 3. Southern blots showing provirus burden in lymphocyte is also possible that FIV infection down regulates T-cell
subsets from cats acutely infected with FIV-NCSU1. (A) Blot from and/or monocyte markers such that infected cells with
two cats acutely infected (3 weeks postinfection) with FIV-NCSU1. anti-FIV on their surface would sort as B cells with anti-Ig.
(B) Blot from the same two cats as in panel A, but cells were To overcome these problems, we identified a MAb, desig-
analyzed at 11 weeks postinfection. Lane markers represent un-
sorted PBMCs (PBM), CD8+ cells (CD8), Ig+ cells (IG), and CD4+ nated B5, that recognizes a non-Ig epitope on feline B cells.
cells (CD4 or CD4+). All lanes represent PCR amplification of B5 was originally raised against an epitope on normal canine
lysates from 1 x 105 cells by using the gag primers FIV-7 and FIV-8, B cells and was found to also bind feline B cells.
except lanes + and -, in which lysates from 2 x 104 FIV-infected Two-color analyses using lymphocytes from either SPF or
and 2 x 104 uninfected FCD4-D cells, respectively, were included as FIV-infected cats indicate that B5 and our pan-T-cell marker
controls. Bands represent DNA segments of approximately 840 bp stain separate lymphocyte populations and that all B5-
as determined by molecular weight standards on the original gels.
positive cells are also positive for feline Ig (Fig. 5). In
addition, two-color analyses using B5 and CM277 (the anti-
monocyte MAb) indicate they also stain separate popula-
the increase in proviral burden in Ig+ cells, their numbers (as tions of cells (Fig. 6). Preincubation of B5 with feline Ig does
determined by complete blood count and differential and not alter the ability of B5 to recognize B cells, suggesting
flow analysis) were not significantly different than preinfec- that it binds to a membrane protein other than surface Ig.
tion numbers. The percentage of Ig+ cells that stain with B5 varies among
The Ig+, provirus-positive cell lacks T-cell markers. Al- cats, ranging from 70 to 95%.
though the presence of FIV provirus in Ig+ cells from FIV Using either magnetic beads or FACS, we sorted B5-
antibody-negative cats supports the suggestion that these positive cells from six different FIV-infected cats to 95 to
cells are not antibody-coated T cells, the shift in proviral 99% purity and analyzed the DNA for the presence of FIV
burden may actually be due to the coating of T cells with provirus. In all six cats, the B5 population was positive for
anti-FIV as the antibody response develops. Therefore, we provirus signal. Figure 4A shows a typical Southern blot
performed two-color flow cytometric analysis on lympho- from one of these cats. Furthermore, with three additional
cytes from the chronically infected cats both presorting and cats, we used two-color FACS to separate Ig+ B5+ and Ig+
postsorting. We found that no CD4+, CD8+, or pan-T- B5- populations and found provirus in the Ig+ B5+ cells
marker-positive cells were also positive for surface Ig and (Fig. 4B). These results support the suggestion that the Ig+
that no Ig+ cells were positive for either CD4 or CD8 (data cell is neither a T cell nor a monocyte and may be a B cell
not shown). These data suggest that T cells were not sorted infected with FIV.
into the Ig+ cell population. The broad lymphocyte tropism is not limited to experimen-
The Ig+ provirus-positive cell is not a monocyte/macro- tally infected cats. Because the 16 cats examined in this study
phage. Macrophages have been shown to be susceptible to were all experimentally infected with FIV-NCSU1, we were
FIV infection (3), and thus circulating monocytes could concerned that serial in vivo passage of this isolate may have
potentially contribute to the provirus signal measured in the altered its original tropism. Therefore, we collected lympho-
sorted lymphocyte subpopulations. Although it is unlikely cytes from the naturally infected cat from which this isolate
that the cultured subpopulations contained monocytes 2 and originated. PBMCs were separated into CD4+, CD8+, and
6 days after adherence to plastic, we used the antimonocyte Ig+ cell populations and analyzed for the presence of FIV5180 ENGLISH ET AL. J. VIROL.
A : J sorted lymphocyte culture supernatants, but the greatest
activity was consistently detected in the Ig+ cell cultures.
Amplification of the RT signal was achieved by adding the
4-day culture supernatants to FIV-susceptible FCD4-D cells.
RT activity developed in the cultures inoculated with super-
natants from all three sorted lymphocyte populations, with
peak activity developing earlier in cultures inoculated with
Ig+ cell supernatant (Fig. 8B). This pattern was similar for
IL IL to IL lymphocyte subpopulations from all six cats that were ex-
* ,w amined. These data demonstrate that infectious virus could
be produced by all three lymphocyte populations in culture.
In vivo tropism of other FIV isolates. To determine whether
the broad lymphocyte tropism was unique to FIV-NCSU1,
lymphocytes from cats infected with three other isolates of
FIV were examined for FIV provirus. Lymphocytes were
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
B +
sorted into subpopulations from two cats experimentally
infected with the Petaluma isolate of FIV, one cat infected
with the Mount Airy isolate of FIV, and one naturally
infected cat presented to the North Carolina State Univer-
sity veterinary teaching hospital. All cats were clinically
normal except for the naturally infected cat, which had
chronic gingivitis, vitritis, and focal retinal degeneration.
Similar to the FIV-NCSU1-infected cats, provirus was de-
tected in CD4+, CD8+, and Ig+ cells from all four cats
infected with the different FIV isolates (results for the Mount
FIG. 4. Southern blots of the PCR-amplified FIV gag gene from Airy and Petaluma isolates are depicted in Fig. 9; results for
two cats experimentally infected (>18 months) with FIV-NCSU1. the naturally infected cat are not shown). The strongest
Samples in panels A and B were amplified together, run on a double signal from each of the three isolates was again in the Ig+ cell
gel, blotted onto a single membrane, and then hybridized. (A) population.
Monocytes (MON) were removed from the PBMCs (PBM) of cat 7
by using MAb CM277-coated magnetic beads, and the remaining
cells (PBL1 and PBL2) were divided into two aliquots. By using DISCUSSION
two-color FACS, PBL1 cells were sorted into T cells (PanT) and B
cells (BC) with MAbs 1.572 and B5, respectively. Similarly, PBL2 Previous studies have demonstrated that FIV is capable of
cells were sorted into T cells (PanT) and Ig+ cells (IG) with MAb infecting feline T lymphocytes, macrophages, and astrocytes
1.572 and anti-cat Ig. (B) A blot from a chronically infected cat (cat in vitro (3, 6, 23). Highly productive and lytic infection of T
6), demonstrating that the FIV provirus in the Ig+ population resides lymphocytes in vitro (23) and observed alterations in T-cell
in the B5+ population. After removal of CD8+ cells (CD8) from numbers (1, 21, 33) and functions (15) in FIV-infected cats
PBMCs (PBM), the Ig+ cells (IG) were collected by FACS, stained have supported the conclusion that FIV is predominately a
with biotinylated MAb B5, and separated into B5+ (BC+) and B5- T-lymphotropic virus. The results presented herein indicate
(BC-) cells by FACS. All lanes represent PCR amplification of that mononuclear cells, in addition to T lymphocytes and
lysates from 1 x 105 cells by using the gag primers FIV-7 and FIV-8, monocytes, are targets for FIV infection.
except lane MON, for which only 5 x 104 cells were available.
Lysates from 2 x 104 FIV-infected (lane +) and uninfected (lane -) PCR and Southern blot analysis of highly purified lympho-
FCD4-D cells were included as controls. Bands represent DNA cyte subpopulations from 21 experimentally and naturally
segments of approximately 840 bp as determined by molecular infected cats revealed FIV provirus in CD4+ cell-, CD8+
weight standards on the original gels. cell-, and Ig+ cell-enriched populations. This pattern of FIV
provirus expression was found in 19 experimentally infected
cats: 10 with long-term (212-month), asymptomatic FIV-
NCSU1 infection, six with acute-stage (2- to 4-week) FIV-
provirus. As in the 16 experimentally infected cats, the NCSU1 infection, two with the Petaluma isolate, and 1 with
strongest signal was seen in the Ig+ cell population (Fig. 7). the Mount Airy isolate. In addition, two naturally infected
Production of infectious virus by lymphocyte subpopula- cats, the original source cat for FIV-NCSU1 and a second
tions. To determine whether the provirus-positive lympho- cat with symptomatic FIV infection, had FIVgag sequences
cyte populations were able to produce infectious virus, in CD4+, CD8+, and Ig+ cells.
supernatants from short-term cultures of sorted lymphocytes The presence of FIV provirus in CD4+ and CD8+ cells
from FIV-infected cats were assayed for RT activity. Short- was not unexpected, as Brown et al. (2) previously demon-
term cultures were used for two reasons: (i) to allow deter- strated that the Petaluma and Peepers isolates of FIV were
mination of basal-level virus production with minimal in capable of infecting both CD4+ cells and CD8+ cells cultured
vitro autoinfection occurring and (ii) because long-term in vitro. Our studies show that FIV-NCSU1 also infects
culture of the Ig+ cells was not possible. Sorted lymphocytes IL-2-dependent CD4+ and CD8+ cultured cells, as well as
from infected cats were cultured for 2 days, washed, placed T-null cells that express neither the CD8 nor the CD4
in fresh medium, and cultured for an additional 4 days. The marker. It is possible that the T-null cells represent NK cells
supernatants were then collected and assayed for the pres- or IL-2-activated LAK cells, which have recently been
ence of RT activity. Figure 8A shows the RT activity in observed for the cat (32). We were unable to confirm Ig+ cell
4-day culture supernatants from lymphocyte subpopulations infection in vitro because of the difficulty in maintaining
from two of the experimentally infected cats and from the viable Ig+ cells in culture.
original FIV-NCSU1 source cat. RT activity was low in the Although macrophages have been shown to be susceptible.
VOL. 67, 1993 CELL TROPISM OF FIV 5181
-0 -
O .
o
IC
>U
0 -9
S_.
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
o
0. _
1l l in3
Control Mouse Hg (FITC) Pan T call (FITC)
Z
-a-
Z Pi
Wc Co
z
is _ t..
r
0 _
0 "
x 1-r
m0 O 1
La
M 4 A
~~~~~~ I a I w l
ol I9 8 .. I S 1- 4 9 _1.
190 o2 3 i4 100 101 102 a
CDS (FITC) CD4 (FITC)
Z
z 019
0-.
Is
0 X
~0-
am
um
Cs
Is .F
le 1 102 los
ANTI-CAT IgG (FITC)
FIG. 5. Two-color flow cytometric analysis of feline PBMCs (from a normal, SPF cat) recognized by MAb B5. The y axis represents
phycoerythrin fluorescence of biotinylated antibodies, and the x axis represents fluorescence of FITC-conjugated antibodies. Numbers
represent the percent positively staining cells in each quadrant. Note that cells recognized by MAb B5 contain surface Ig (bottom left panel,
37.73% dual positive cells) but none of the feline T-cell markers (less than background staining in all cases). Similar two-color analysis using
PBMCs from FIV-infected cats yielded the identical pattern.V
h.
-W
;F -0~~~~~~~~~~~A
0
0
0-
S1
0
Forward Scatter Forward Scatter
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
00-
z z
in
0 en
w
0 P in
0
M
a 0~
IL.4
>
*n *
0
a ea
(L0-
91 0
12
1131 102 tog lo04 10' 1'2 103
Goat anti-Mouse ig (FITC) Goat anti-Mouse 1g (FITC)
_ m. 45.97% 1.72%
0
-
.1*
Z ,,
CC
z
LI
aa
0
Oa
O S .:.. 25.22%
to0-
In a I.;
m -
.:.:~~~~~
:. ~~~~~
I>, T IIIIII
km
LSI
1 l01 102 103 104 10 1i1 tI 2 1 a3 la4
Antl-Cat Monocyte (FITC) Anti-Cat Monocyte (FITC)
FIG. 6. Two-color flow cytometric analysis of feline PBMCs (collected from a normal, SPF cat) recognized by MAb B5 and MAb CM277.
The left panels represent staining of the small mononuclear cell population (predominantly lymphocytes; the scatter pattern and gate are
shown in the top left panel), and the right panels represent staining of the large mononuclear cell population (large lymphocytes and
monocytes; the scatter pattern and gate are shown in top right panel). Middle panels represent the cells stained with biotinylated mouse Ig
(y axis) and FITC-conjugated goat anti-mouse Ig (x axis) (controls). The cells in the bottom two panels were stained with CM277 (anti-feline
monocyte MAb), washed, stained with FITC-conjugated goat anti-mouse Ig (x axis), washed, and then stained with biotinylated B5 (y axis).
Considering that background staining is 1.23% (middle right panel), less than 1% of the large mononuclear cells stain with both B5 and CM277.
Similar two-color analysis using PBMCs from FIV-infected cats yielded the identical pattern.
5182VOL. 67, 1993 CELL TROPISM OF FIV 5183
populations demonstrated that no cells expressed both a
T-cell antigen (pan-T, CD4, or CD8) and surface Ig. In
addition, flow analysis of the sorted, plastic-nonadherent
a0
U() populations with the antimonocyte antibody CM277 (34)
UQ I 0 demonstrated that all three purified lymphocyte subpopula-
tions were negative for monocytes.
Although the data support the Ig+ cell population being
>95% B cells, we can not exclude the possibility that part of
the provirus signal observed in this population is generated
-I- from a low number of Ig-coated null lymphocytes, NK cells,
or LAK cells. We have demonstrated that a CD4- CD8- cell
will support FIV infection in vitro, and NK cells have been
shown to support HIV infection in vitro (4). However, in our
in vitro studies, the CD4- CD8- cells were positive for our
pan-T-cell marker, while we could not demonstrate the
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
presence of any T-cell marker on the Ig+ cells. It is also
possible that the provirus-positive cells in the Ig+ cell
population are either Ig-coated T cells or monocytes that
have had their identifying surface markers down regulated
FIG. 7. Southern blot of the PCR-amplified FIV gag gene from because of FIV infection. However, the expression of Ig on
the naturally infected cat (cat TH) from which FIV-NCSU1 was
originally isolated. For all samples, PCR was performed with gag
these cells after magnetic bead sorting suggests that this is
primers FIV-1 and FIV-2 and 1 ILg of extracted DNA. Lanes not so. Flow cytometric analysis of Ig+ cells after 24 h of
contained CD8+ cells (CD8), Ig+ cells (IG), CD4+ cells (CD4), and culture revealed they not only had released their beads but
controls consisting of uninfected FCD4-D cells (-) and FIV- also had weak to no Ig expression. By 48 h of culture, over
NCSU,-infected FCD4-D (D+) cells. Bands represent DNA seg- 95% of the cells expressed Ig. This ability to strip and then
ments of approximately 580 bp as determined by molecular weight reexpress antibody suggests that the majority of these cells
standards on the original gels. were B cells.
The broad lymphocyte tropism of FIV represents a major
departure from what is observed in HIV-1 infections. Al-
to FIV infection (3), little to no provirus was demonstrated though a number of studies have demonstrated in vitro
by PCR and Southern blot analysis of purified monocytes. infection of several CD4-negative cell types, including B
Indeed, we were able to detect a weak signal in only 1 of 10 cells (14, 18, 19), this does not appear to be a characteristic
cats examined. Because fewer monocytes (5 x 104) than of the in vivo infection. Analysis of highly purified lympho-
lymphocyte subsets (1 x 105) were available for evaluation, cyte subpopulations from HIV-1-infected individuals by
it is difficult to compare their levels of viral burden. How- PCR demonstrated, similar to what is observed with FIV,
ever, our findings are similar to those of Massari et al. (16), provirus in CD4+ cells from all patients tested and in
who reported that provirus can be detected in circulating monocytes in about 10% of patients. Unlike with FIV
monocytes from only 10% of HIV patients. infection, no patient showed provirus in CD8+ T cells or
The presence of FIV provirus in CD8+ cells is interesting CD19+ B cells (25, 27).
in light of what is known of the pathogenesis and lymphocyte In addition to finding FIV provirus in all three lymphocyte
subset changes in FIV-infected cats. We (30) and others (1, subpopulations, we were able to show that these subpopu-
33) have demonstrated that experimentally FIV-infected, lations were able to produce infectious virus. We were able
asymptomatic cats have normal or even increased numbers to demonstrate low levels of Mg2e-dependent RT activity in
of CD8+ cells, yet we have shown here that a fraction of culture supernatants from all three subpopulations. Adding
these cells appear to carry FIV provirus. Whether the CD8+ these culture supernatants to FCD4-D led to much greater
infection is lytic in vivo is not known, but we have demon- RT activity, suggesting that these supernatants contained
strated in vitro that FIV infection is lytic for CD8+ cells. The infectious virus. In both culture situations, the RT activity in
role of CD8+ cells in FIV pathogenesis remains to be the Ig+ population developed much more rapidly, and in the
determined. original cultures it achieved much higher levels, than the RT
The presence of FIV provirus in Ig+ cells is somewhat activity produced by either of the T-cell-enriched popula-
surprising and appears to represent a major deviation from tions. This rapid induction of RT activity in the Ig+ cell
the situation in other lentivirus infections. The major host population is interesting in the light of our sorting technique
cells for replication of other animal lentiviruses (simian and supports the suggestion that it is indeed the B cell that is
immunodeficiency virus, equine infectious anemia virus, infected. We purified this population by using anti-cat Ig
visna-maedi virus), as well as those for HIV, are T cells bound to magnetic beads, which leads to cross-linking and
and/or macrophages (20). Not only do the Ig+ cells contain stripping of surface Ig. Cross-linking of B-cell surface Ig can
FIV gag sequences, but in all but the acutely infected cats, lead to B-cell stimulation. If indeed B cells are infected, such
the Ig+ cell-enriched populations yielded a stronger provirus stimulation could lead to more rapid virus production, as has
signal on the gel and Southern blot than either CD4+ or been described for mitogen-stimulated, HIV-infected CD4+
CD8+ cells. In addition, limiting-dilution PCR analysis indi- cells (36).
cated that a greater number of Ig+ cells than either CD4+ or Of particular importance to understanding the pathogene-
CD8+ lymphocytes are infected. sis of FIV, and perhaps HIV-1, is our observation that the
Several findings support the hypothesis that the Ig+ pop- CD4+ cell has the greatest provirus burden during the acute
ulation was predominantly an enriched B-cell population stage (2 to 6 weeks) of infection but not during the later
with minimal contamination of either T cells or monocytes. asymptomatic stage. This high CD4+ provirus burden pre-
Two-color analysis of both unfractionated and sorted PBMC cedes the early reduction of CD4+ cells following FIV5184 ENGLISH ET AL.
infection, and the loss of a strong provirus signal in CD4+
cells correlates with their decrease in numbers in the circu-
lation. However, by 11 weeks postinfection, when CD8+
cells are increasing but CD4+ cells are not, the level of
provirus burden in the remaining CD4+ cells has greatly
decreased. It is possible that the shift in provirus burden, as
well as the decrease in CD4' cell numbers, reflects the rapid
lysis of a subpopulation of CD4+ cells that are highly
susceptible to FIV infection. Continued viral replication may
depend on the slow renewal of these highly susceptible cells,
S*4
1
8,9
2
J. VIROL.
FIG. 9. Southern blot analysis of PCR amplification of FIV from
lymphocyte subpopulations from a cat experimentally infected with
FIV-Mount Airy (1) and a cat experimentally infected with FIV-
Petaluma (2). Both cats were infected for 18 months. Double
amplification was performed by amplifying 1 pg of DNA first
through 30 cycles with primers FIV-7 and FIV-8 and then through 20
A additional cycles with FIV-1 and FIV-2 oligonucleotides as nested
_ 19+ CELL primers. Blotting, hybridization, and autoradiography were per-
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
1500
formed as described in Materials and Methods. Lane markers
M CD8+ represent CD8' cells (8), Ig+ cells (B), CD4' cells (4), and CRFK
E}CD4+ cells infected with the Black strain of FIV (+). Bands represent
1200 DNA segments of approximately 580 bp as determined by molecular
0 El CONT weight standards on the original gels.
a.
900
'I
leading to a lower proviral burden in the population. Such
IL differential susceptibility of subsets of cells has been dem-
0 onstrated for both HIV and simian immunodeficiency virus.
600 The memory subset of T cells as defined by CD45RA+
I--
expression is preferentially infected by HIV-1 in vitro, and
PCR amplification and Southern blotting indicate that this
300 same subset of cells contains the majority of provirus in vivo
(26). Likewise, a memory subset of cells is preferentially
susceptible to simian immunodeficiency virus in vitro and
carries the greatest proviral burden in vivo (9, 35).
Cat I Cat 5 TH
B
Cat I Cat 5 TH
12r
C
0
2
0
I I
I--
0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24
DAYS POST INOCULATION OF FCD4-D CELLS
FIG. 8. Mg2-dependent RT activity in culture supernatants of purified lymphocyte subpopulations from FIV-infected cats. (A)
Percoll-purified lymphocytes were sorted into CD4+, CD8+, and Ig+ cell populations, placed in culture for 48 h, and then washed, assayed
for population purity by flow cytometry (>95% pure), and placed in fresh culture medium. After 4 days of culture, 10 Fl of supernatant was
collected and assayed for RT activity. CONT represents RT activity of the supernatant from uninfected FCD4-D cells. Sorted lymphocyte
cultures from two SPF cats (cat 1 and cat 5) infected with FIV-NCSU1 for 18 months and from the naturally infected cat from which
FIV-NCSU1 was originally isolated (TH) are shown. (B) Mg2e-dependent RT activity in FCD4-D cells inoculated with supernatant from
cultured sorted lymphocytes from FIV-infected cats. One hundred microliters of supernatant from the 4-day cultures described above was
added to 2 x le3 DEAE-dextran treated FCD4-D cells. After 24 h, the cells were washed and resuspended in fresh culture medium. Culture
supernatant was collected at various times postinoculation and assayed for RT activity. The RT activity from uninfected FCD4-D cells wasVOL. 67, 1993 CELL TROPISM OF FIV 5185
Our data show that the Ig+ cell fraction bears the major (ed.), Current veterinary therapy XI. W. B. Saunders Co.,
FIV provirus burden during the asymptomatic infection and Philadelphia.
suggest that B cells are a target cell in this population. 13. Klatzmann, D., F. Barre-Sinoussi, M. T. Nugeyre, C. Danquet,
However, we cannot rule out the possibility that some E. Vilmer, C. Griscelli, F. Brun-Veziret, C. Rouzioux, J. C.
Gluckman, J. C. Chermann, and L. Montagnier. 1984. Selective
yet-to-be-identified non-T, nonmonocyte cell exists at low tropism of lymphadenopathy associated virus (LAV) for helper-
levels in this population and is also a target of FIV. Because inducer T lymphocytes. Science 225:59-63.
the immunological and clinical consequences of FIV infec- 14. Levy, J. A., J. Shimabukuro, T. McHugh, C. Casavant, D. Stites,
tion, characterized by a reduction in CD4+ cells, an inver- and L. Oshiro. 1985. AIDS-associated retroviruses (ARV) can
sion of the CD4+/CD8+ ratio, and the development of AIDS productively infect other cells besides human T helper cells.
(1, 21, 33), are so similar to those of HIV infection (17), it Virology 147:441-448.
will be important to resolve this issue of tropism. Despite 15. Un, D. S., D. D. Bowman, R. H. Jacobson, M. C. Barr, M.
what appears to be a major difference in cell tropism, these Fevereiro, J. R. Williams, F. M. de Noronha, F. W. Scott, and
two viruses have a similar immunopathogenesis, and in this R. J. Avery. 1990. Suppression of lymphocyte blastogenesis to
regard, FIV infection should be considered an excellent mitogens in cats experimentally infected with feline immunode-
ficiency virus. Vet. Immunol. Immunopathol. 26:183-189.
model for HIV-1 pathogenesis. 16. Massari, F. E., G. Poli, S. M. Schnittman, M. C. Psallidopoulos,
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guest
V. Davey, and A. S. Fauci. 1990. In vivo T lymphocyte origin of
ACKNOWLEDGMENTS macrophage-tropic strains of HIV: role of monocytes during in
vitro and in vivo infection. J. Immunol. 144:4628-4632.
We thank Robert Olmsted (Georgetown University and the Na- 17. McChesney, M. B., and M. B. A. Oldstone. 1989. Virus-induced
tional Institute of Allergy and Infectious Diseases) for providing immunosuppression: infections with measles virus and human
blood samples from two FIV-Petaluma-infected cats and one FIV- immunodeficiency virus. Adv. Immunol. 45:335-380.
Mount Airy-infected cat for analysis, and we thank N. Gengozian 18. Monroe, J. E., A. Calender, and C. Mulder. 1988. Epstein-Barr
(University of South Florida, St. Petersburg) for providing the CM277 virus-positive and -negative B-cell lines can be infected with
MAb. We also thank Chris Parrish for excellent technical assistance human immunodeficiency virus types 1 and 2. J. Virol. 62:3497-
and Wayne Tompkins for critical review of the manuscript. 3500.
This study was supported by grants CA-43676 and CA-54723 from 19. Montagnier, L., J. Gruest, S. Chamaret, C. Dauguet, C. Axler,
the National Cancer Institute and AI-27665 from the National D. Guetard, M. T. Nugeyre, F. Barre-Sinoussi, J. C. Chermann,
Institute of Allergy and Infectious Disease. J. B. Brunet, D. Klatzmann, and J. C. Gluckman. 1984. Adap-
tation of lymphadenopathy associated virus (LAV) to replica-
REFERENCES tion in EBV-transformed B lymphoblastoid cell lines. Science
1. Ackley, C. D., J. K. Yamamoto, N. Levy, N. C. Pedersen, and 225:63-66.
M. D. Cooper. 1990. Immunologic abnormalities in pathogen- 20. Narayan, O., and J. E. Clements. 1989. Biology and pathogen-
free cats experimentally infected with feline immunodeficiency esis of lentiviruses. J. Gen. Virol. 70:1617-1639.
virus. J. Virol. 64:5652-5655. 21. Novotney, C., R. English, J. Housman, M. Davidson, M. Nasisse,
2. Brown, W. C., L. Bissey, D. S. Logan, N. C. Pedersen, J. H. C. R. Jeng, W. Davis, and M. Tompkins. 1990. Lymphocyte
Elder, and E. W. Collisson. 1991. Feline immunodeficiency virus population changes in cats naturally infected with feline immu-
infects both CD4+ and CD8+ T lymphocytes. J. Virol. 65:3359- nodeficiency virus. AIDS 4:1213-1218.
3364. 22. O'Connor, T. J., S. Tanguay, R. Steinman, R. Smith, M. C.
3. Brunner, D., and N. C. Pedersen. 1989. Infection of peritoneal Barr, J. K. Yamamoto, N. C. Pedersen, P. R. Andersen, and
macrophages in vitro and in vivo with feline immunodeficiency Q. J. Tonelli. 1989. Development and evaluation of immunoas-
virus. J. Virol. 63:5483-5488. say for detection of antibodies to the feline T-lymphotropic
4. Chehimi, J., S. Bandyopadhyay, K. Prakash, B. Perussia, N. F. lentivirus (feline immunodeficiency virus). J. Clin. Microbiol.
Hassan, H. Kawashima, D. Campbell, J. Kornbluth, and S. E. 27:474-479.
Starr. 1991. In vitro infection of natural killer cells with different 23. Pedersen, N. C., E. Ho, M. L. Brown, and J. K. Yamamoto.
human immunodeficiency virus type 1 isolates. J. Virol. 65: 1987. Isolation of a T-lymphotropic virus from domestic cats
1812-1822. with an immunodeficiency-like syndrome. Science 235:790-793.
5. Dalgleish, A. C., P. C. L. Beverley, P. R. Clapham, D. H. 24. Pedersen, N. C., J. K. Yamamoto, T. Ishida, and H. Hansen.
Crawford, M. F. Graves, and R. A. Weiss. 1984. The CD4 (T4) 1989. Feline immunodeficiency virus infection. Vet. Immunol.
antigen is an essential component of the receptor for the AIDS Immunopathol. 21:111-129.
retrovirus. Nature (London) 312:763-766. 25. Psallidopoulos, M. C., S. M. Schnittman, L. M. Thompson HI,
6. Dow, S., M. Dnetz, and E. Hoover. 1992. Feline immunodefi- M. Baseler, A. S. Fauci, H. C. Lane, and N. P. Salzman. 1989.
ciency virus neurotropism: evidence that astrocytes and mac- Integrated proviral human immunodeficiency virus type 1 is
rophages are the primary targets. Vet. Immunol. Immuno- present in CD4+ peripheral blood lymphocytes in healthy sero-
pathol. 35:23-35. positive individuals. J. Virol. 63:4626-4631.
7. Dow, S. W., M. L Poss, and E. A. Hoover. 1990. Feline 26. Schnittman, S., H. Lane, J. Greenhouse, J. Justement, M.
immunodeficiency virus: a neurotropic lentivirus. J. Acquired Baseler, and A. Fauci. 1990. Preferential infection of CD4+
Immune Defic. Syndr. 3:658-668. memory T cells by human immunodeficiency virus type 1:
8. Erlich, H. 1989. PCR technology: principles and application for evidence for a role in the selective T-cell functional defects
DNA amplification. Stock Press, New York. observed in infected individuals. Proc. Natl. Acad. Sci. USA
9. Gallatin, W., M. J. Gale, P. Hoffman, D. Willerford, K. Draves, 87:6058-6062.
R. Benveniste, W. Morton, and E. ClarL 1989. Selective repli- 27. Schnittman, S. M., M. C. Psallidopoulos, H. C. Lane, L.
cation of simian immunodeficiency virus in a subset of CD4+ Thompson, M. Baseler, F. Massari, C. H. Fox, N. P. Salzman,
lymphocytes. Proc. Natl. Acad. Sci. USA 86:3301-3305. and A. S. Fauci. 1989. The reservoir for HIV-1 in human
10. Gardner, N. B., and P. W. Luciw. 1989. Animal models of peripheral blood is a T cell that maintains expression of CD4.
AIDS. FASEB J. 3:2593-2606. Science 245:305-308.
11. Goff, S., P. Traktman, and D. Baltimore. 1981. Isolation and 28. Talbott, R. L., E. E. Sparger, K. M. Lovelace, W. M. Fitch,
properties of Moloney murine leukemia virus mutants: use of a N. C. Pedersen, P. A. Luciw, and J. H. Elder. 1989. Nucleotide
rapid assay for release of virion reverse transcriptase. J. Virol. sequence and genomic organization of feline immunodeficiency
38:239-248. virus. Proc. Natl. Acad. Sci. USA 86:5743-5747.
12. Jacobs, R., J. Lumsden, and W. Vernau. 1992. Canine and feline 29. Tompkins, M. B., D. H. Gebhard, H. R. Bingham, M. J.
reference values, p. 1250-1275. In R. Kirk and J. Bonagura Hamilton, W. C. Davis, and W. A. F. Tompkins. 1990. Charac-5186 ENGLISH ET AL. J. VIROL.
terization of monoclonal antibodies to feline T lymphocytes and dysfunction in cats experimentally infected with feline immuno-
their use in the analysis of lymphocyte tissue distribution in the deficiency virus. J. Virol. 65:2225-2230.
cat. Vet. Immunol. Immunopathol. 26:305-317. 34. Whitehurst, C. E., R. J. Hill, N. K. Day, and N. Gengozian.
30. Tompkins, M. B., P. D. Nelson, R. V. English, and C. Novotney. 1991. Phenotypic markers for the feline monocyte: rosette
1991. Early events in the immunopathogenesis of feline retrovi- formation with human erythrocytes and a monoclonal antibody
rus. J. Am. Vet. Med. Assoc. 199:1311-1315. which binds myeloid cells. Proc. Soc. Exp. Biol. Med. 197:317-
31. Tompkins, M. B., G. K. Ogilvie, R. A. Franklin, K. W. Kelley, 325.
and W. A. Tomplins. 1987. Induction of IL-2 and lymphokine
activated killer cells in the cat. Vet. Immunol. Immunopathol. 35. Willerford, D., M. J. Gale, R. Benveniste, E. Clark, and W.
16:1-10. Gallatin. 1990. Simian immunodeficiency virus is restricted to a
32. Tompkins, M. B., V. F. Pang, P. A. Michaely, R. I. Feinmehl, subset of blood CD4+ lymphocytes that includes memory cells.
E. J. Basgall, T. V. Baszler, J. F. Zachary, and W. A. Tompkins. J. Immunol. 144:3779-3783.
1989. Feline cytotoxic large granular lymphocytes induced by 36. Zagury, D., J. Bernard, R. Leonard, R. Cheynier, M. Feldman,
recombinant human IL-2. J. Immunol. 143:749-754. P. S. Sarin, and R. C. Gallo. 1986. Long-term cultures of
33. Torten, M., M. Franchini, J. E. Barlough, J. W. George, E. HTLV-III-infected T cells: a model of cytopathology of T-cell
Mozes, H. Lutz, and N. C. Pedersen. 1991. Progressive immune depletion in AIDS. Science 231:850-853.
Downloaded from http://jvi.asm.org/ on January 15, 2021 by guestYou can also read
