|
|
||||||||


Departments of
*
Microbiology,
Comparative Medicine, and
Medicine, University of Alabama, Birmingham, AL 35294
| Abstract |
|---|
|
|
|---|
and
light chains were expressed by 62% and 38% of
the Ig-containing cells, respectively. J chain expression occurred in
most cells irrespective of the Ig isotype. In the SCID mice engrafted
with human lymphocytes, a few IgM-containing cells were found in the
spleen; plasma cells were not found in other tissues, including the
intestine. The aggregation of plasma cells did not appear to be a
result of infection with EBV. T cells were rarely found in the lymphoid
aggregates but were recovered from the spleen and peritoneal lavage.
Human Ig levels in the serum of engrafted mice reflected the isotype
distribution of the cells with IgG > IgM
IgA. | Introduction |
|---|
|
|
|---|
Engrafted SCID mice have been used to study responses of the human immune system to infections with HIV and EBV, bacterial infections, and vaccines, as well as to determine lymphocyte homing patterns (6-8). Previous work using the hu-PBL-SCID model has shown that the majority of human cells remain in the peritoneal cavity, repopulating only the spleen, thymus, and lymph nodes (5, 9, 10). Human Igs are present in the serum and peritoneal lavage fluids of the mice, indicating engraftment of, and differentiation into, Ig-producing plasma cells (7, 11, 12, 13, 14, 15, 16). Long-term B cell engraftment may be influenced by the EBV status of the donor. In mice repopulated with lymphocytes from EBV+ donors, B cell lymphomas appeared after 8 wk (11, 17, 18, 19). However, early after reconstitution of SCID mice with human B cells, the phenotype and activity of the B cells can be compared with those of the normal human immune system (8, 12, 14, 20).
While several groups have studied B cell populations in the hu-PBL-SCID model, mice that have been engrafted with nonhuman primate B cells have not been characterized. Because nonhuman primates are often infected with pathogens similar to those of humans, SCID mice repopulated with nonhuman primate lymphocytes would be an effective tool for direct comparisons between human and nonhuman primate immune responses to vaccines developed against these pathogens. To determine the early engraftment properties of B cells in SCID mice, we investigated the location, phenotype, and EBV status of human B cells 3 weeks after repopulation with human PBMC. In addition, we obtained tissues from nonhuman primate PBMC-engrafted SCID mice for comparisons of primate and human B cell repopulation in SCID mice.
| Materials and Methods |
|---|
|
|
|---|
C.B-17 (scid/scid) mice were obtained from the University of Alabama at Birmingham Animal Resources Program. The mice were maintained in cages fitted with microisolators. Cages, bedding, food, and water were autoclaved before use, and no antibiotics were administered.
Engraftment of SCID mice with PBMC
Venous blood was collected from seven healthy adult human donors. PBMC were isolated by centrifugation on a Ficoll-Hypaque density gradient (Sigma, St. Louis, Mo). The cells were washed three times in Dulbeccos PBS and resuspended in RPMI 1640 containing 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin (all from Mediatech, Herde, VA). PBMC were treated for 2 days in culture with either the addition of 0.5% PWM (Life Technologies, Gaithersburg, MD), 10 ng/ml PMA (Sigma), or 0.1% PHA (Difco, Detroit, MI). The cells were then washed extensively, and viability was determined by trypan blue dye exclusion. The cells were resuspended at 2 to 3 x 107 cells/250 to 1000 µl in RPMI 1640.
To enhance engraftment, six mice were pretreated 18 h before
receiving unstimulated or stimulated human PBMC by i.p. administration
of 20 µl of anti-asialo GM1 ganglioside (Wako Bioproducts,
Richmond, VA) to remove NK cells (21, 22). Human growth hormone (5 µg
HuGH,4 Protropin, Genentech,
San Francisco, CA) was administered to each animal 18 h after
injection of donor cells, and the HuGH treatment continued every other
day until the animals were sacrificed. A total of 22 mice were injected
i.p. with 2 to 3 x 107 freshly isolated or cultured
human PBMC (Table I
). Mice were
sacrificed 3 to 4 wk later. One mouse died during the study and only
two hu-PBL-SCID mice were found to have cells containing murine Igs
(Table I
).
|
One day before sacrifice, saliva was collected by pipette after stimulation by i.p. injection with 2 µg of carbamylcholine chloride (Carbocol, Sigma). After the animals were anesthetized, blood was collected by brachial bleed. The peritoneal cells were recovered by flushing the cavity with 3 ml of PBS. Bile was collected in disposable syringes into 100 µl of PBS. In all mice spleen, mesentery, and small intestine were harvested; in addition, liver, kidney, large intestine, thymus, pleural fat, lungs, salivary glands, peritoneal lining, and diaphragm from some animals were collected for histology.
Mesenteric tissues were obtained from SCID mice reconstituted with PBMC from chimpanzees (19 mice) and pig-tailed macaques (15 mice). PBMC from pig-tailed macaques were Con A stimulated and infected with SIVsmmPBj14 in vitro before engraftment, and each mouse received 2.5 x 107 to 5 x 107 cells i.p. (23). Mesenteric tissues from control age-matched BALB/c and C57BL/6 mice were also obtained at time of sacrifice.
Ig levels
Serum IgA, IgG, and IgM isotypes were determined by ELISA. Bile and saliva were analyzed for IgA. Briefly, 96-well polyvinyl microtiter plates (Dynatech, Chantilly, VA) were coated with the F(ab')2 fragments of goat Abs specific for human IgG (2.5 µg/ml), IgA (2.5 µg/ml), or IgM (1.0 µg/ml) (Jackson ImmunoResearch, West Grove, PA) to capture the respective Ig. Biotinylated goat anti-human IgA, IgG, and IgM (Tago Immunologics, Burlingame, CA) were used to detect bound Ig. A serum pool (Moni-Trol E, Baxter, McGaw Park, IL) with known concentrations of all three isotypes served as a standard for total IgA, IgG, and IgM levels. Standard curves were constructed using a computer program based on the four-parameter logistic model (Delta Soft, BioMetallics, Princeton, NJ). Sera from SCID mice that did not receive cells and normal BALB/c mice were used as controls.
Phenotype analysis
Peritoneal lymphocyte populations were analyzed by flow cytometry. A mAb against MHC class I (HLA-A, -B, and -C, Olympus Immunochemicals, Lake Success, NY) Ag was used to distinguish the human cells from the resident mouse cells. In addition, Abs against the surface pan-B (CD19) and -T cell markers (CD3, CD2, and CD5) were used to characterize the lymphocyte populations. All Abs were purchased from Becton Dickinson (San Jose, CA) unless stated otherwise.
Histology
Tissues were fixed in acid alcohol (95% ethanol, 5% glacial
acetic acid), embedded in low-melting point paraffin, and processed
according to the method of Sainte-Marie (24). For initial analysis,
serial sections were stained with hematoxylin and eosin (H&E) or
primary polyclonal Abs against human and mouse Ig (Southern
Biotechnology Associates, Birmingham, AL). The T cell population was
identified using mAb to CD2. To determine the isotypes of the
Ig-containing cells, polyclonal Abs against human IgG, IgM, and IgA
labeled with tetraethylrhodamine isothiocyanate (TRITC), FITC, or
biotin (Southern Biotechnology), respectively, were incubated together
on the same section. After an extensive wash in PBS, the secondary
reagent, streptavidin-conjugated fluorochrome
7-amino-4-methyl-coumarin-3-acetate (AMCA) (Jackson), was added to
visualize the biotinylated IgA Ab, yielding three-color
immunofluorescence (25). Similar techniques were used to determine the
isotype of cells containing cytoplasmic J chain and the presence of
and
light chains (Southern Biotechnology) (26). The polyclonal Abs
were absorbed with mouse liver powder (Rockland, Gilbertsville, PA) to
remove any nonspecific reactivity with murine B cells. The polyclonal
Abs were shown to cross-react with other primate but not murine
lymphocytes.
EBV status
Sections positive for primate Ig-containing cells were stained for the Epstein-Barr nuclear Ag (EBNA-1, Accurate Chemicals, Westbury, NY). The phenotype of EBV+ B cells was determined as described above.
Statistical analysis
All statistical analyses were performed with the InStat 2.0 (GraphPad, San Diego, CA) software package. Differences in the percentage of Ig-containing cells between human- and primate-engrafted SCID mice were determined by the Mann-Whitney test for nonparametric data.
| Results |
|---|
|
|
|---|
An average of 2.4 x 106 cells (range:
0.56.5 x 106) was recovered from the peritoneal
lavage. To determine the percentage of human lymphocytes, Abs against
MHC class I Ags were used. Human lymphocytes represented an average of
44% (range: 1986%) of the recovered peritoneal cells (Fig. 1
A). T cells represented 41%
(1080%) of the peritoneal population, and less than 2% (06%)
stained for the B cell surface Ag, CD19 (Fig. 1
B).
Phenotypes of the recovered PBMC were not altered by preincubation of
the cells with any of the specified mitogens (Table I
).
|
H&E staining of the mesentery from reconstituted SCID mice
revealed that lymphocytes were beneath the peritoneal lining, organized
into microscopic, vascularized, and sometimes encapsulated aggregates
that did not have a typical lymph node structure with germinal centers
(Fig. 2
A). Similar lymphocytic
aggregates containing only IgM+ cells were found in the
mesentery of normal BALB/c and C57BL/6 mice (Fig. 2
B), but
not in nonengrafted SCID mice. A second type of microscopic structure
that appeared to be a small lymph node was also found in the mesentery
in both engrafted (Fig. 2
C) and nonengrafted SCID mice, as
well as in normal mice (Fig. 2
D). In normal animals, these
structures contained mainly IgA+ (cytoplasmic and surface)
and surface IgM+ cells, with few cytoplasmic
IgG+ cells.
|
|
Further characterization of the lymphocytic aggregates determined
that the Ig-containing cells displayed morphologic features of
lymphoblasts or plasma cells. Evaluation of light chain expression of
the Ig-containing cells within the lymphoid aggregates revealed that
62% expressed
and 38% expressed
light chains (Fig. 3D
).
Analysis of the isotypes of the cytoplasmic Igs showed that IgG
predominated and that IgA- and IgM-containing cells were generally
found in approximately equal numbers (Table II
; Fig. 3
, E, F,
and G). There were no significant differences in isotype
distribution of Ig-containing cells from mice engrafted with fresh or
mitogen-stimulated PBMC from the same donor, or between mice engrafted
with PBMC from different donors. Therefore, we combined the phenotypic
data from the hu-PBL-SCID mice for comparisons with nonhuman
primate-engrafted mice. Lymphocytic aggregates from SCID mice
reconstituted with nonhuman primate PBMC were similar to lymphoid
aggregate tissues from hu-PBL-SCID mice in that a greater
percentage of the lymphoblasts contained cytoplasmic IgG. Significant
differences in the percentages of IgG- and IgM-containing cells were
observed in the hu-PBL-SCID mice when compared with the
chimpanzee-PBL-SCID mice, but no significant differences were found
between the mice engrafted with human and macaque lymphocytes, or
between mice engrafted with macaque and chimpanzee lymphocytes (Table II
). In the hu-PBL-SCID mice, spleen sections showed repopulation only
by IgM-containing cells, and aggregated lymphoblastic cells were found
occasionally in the pleural cavity near the thymus. Interestingly, 69%
of IgA-, 92% of IgM-, and 78% of IgG-containing cells expressed
cytoplasmic J chain (Fig. 3
H). We did not observe any
significant repopulation of the gut, liver, thymus, lungs, or salivary
glands by human or nonhuman primate B cells.
|
To determine whether the human Ig-containing cells were infected with EBV, we stained the sections with Abs against the latent EBV nuclear Ag 1 (EBNA-1). At three weeks postinjection, less than 1% of the human lymphocytes were positive for EBNA-1 in the majority of tissues analyzed. In only two mice from the same donor did we find high numbers of EBNA-1+ cells.
Human Ig levels in serum, saliva, and bile
Human Igs were detected in sera and secretions of the reconstituted mice. Again, IgG was the predominant isotype found in the serum (range: 260-5000 µg/ml). Lower levels of IgM (range: 0.2192 µg/ml) or IgA (3.7331 µg/ml) were detected. Due to low sample volumes, only IgA was analyzed in bile and saliva samples. Human IgA was detected in the bile (54477 ng/ml) of 5 of the 11 engrafted animals tested (45%) but was not detected in the saliva.
| Discussion |
|---|
|
|
|---|
Since most of the primate T cells were recovered from the peritoneal lavage or spleen, the state of activation of the T cell population may determine their localization in SCID mouse tissues. On the other hand, cells of B lineage were differentiated into lymphoblastoid or plasma cells, which characteristically lose CD19, thus explaining the decreased numbers of recovered CD19+ cells. In addition, the localization of B lymphoblasts into organized lymphoid structures may prevent their removal by peritoneal flushing. This localization may also explain why B lineage cells were not detected in the intestinal tissues. In other experiments in which mice were reconstituted with mouse (33, 34) or rat lymphocytes (H.H.D., unpublished data), B cells were detected in the intestinal lamina propria; in the SCID mice engrafted with rat lymphocytes, no mesenteric lymphocytic aggregates were found. Perhaps species differences in B cell homing receptors influence their tissue distribution.
T cells were not observed in the intestinal tissues. A previous report (8) suggested that T cells found in peripheral tissues, such as the spleen, had characteristics of activated cells, which down-regulate their homing receptors. This may explain the lack of T cell migration into the intestine.
The lymphoblasts found in the aggregates were polyclonal, as evidenced by expression of all three major Ig heavy chain and both light chain isotypes. The predominant class of Ig in the aggregates, as well as in serum, was IgG. IgA- and IgM-secreting cells comprised a higher percentage of the cells in the aggregates than the corresponding levels of these Igs in the serum. One explanation for this difference comes from studies that have shown the half-life of human serum IgA and IgM (36.9 and 23.1 h, respectively) to be much shorter than that of human IgG (11.9 days) in engrafted SCID mice (14). Our finding of IgA in the bile would indicate that there is transport of IgA from the circulation (35). An interesting observation was the expression of cytoplasmic J chain by the majority of Ig-containing cells, regardless of the isotype. This finding suggests that the B lineage cells are in a state of activation, typical of Ig- and J chain-producing cells (36, 37, 38). B lymphoblasts were found on occasion in the spleen and in the pleural fat, indicating that they can disseminate to other regions. These findings were consistent whether the mice were reconstituted with human, chimpanzee, or macaque PBMC, indicating that primate B cell engraftment is comparable among species.
Lymphomas that arise in hu-PBL-SCID mice have been shown to contain latent EBV Ags characteristic of EBV-infected cell lines (11, 39). Therefore, we examined the lymphocytic aggregates for the expression of EBV Ags. The majority of human lymphocytes were negative for EBV Ags at 3 weeks postinfection. This finding suggests that the phenotype of the Ig-containing cells in early engraftment of SCID mice is similar to that of mitogen-activated human peripheral blood lymphocytes (36) and not due to lymphoproliferative outgrowth of an EBV-transformed B cell. The lymphocytic tumors that are often described in engrafted SCID mice (10, 11, 18, 19, 39) may occur after long-term engraftment if the few EBV+ cells found in early engraftment become long-lived and invade areas such as the small lymph node-like structures, which are generally devoid of primate cells.
A possibility that the observed lymphoid structures, generated in SCID
mice after the injection of human PBMC, may be analogous to the omental
"milky spots" (also known as Milcheflecken or taches laiteuses)
(40, 41, 42, 43, 44) has been considered. These structures are found with a high
density in the omentum of many species, including humans and mice (41, 43). In mice and humans, the omental lymphoid structures contain
self-replenishing lymphocytes of the B1 (CD5+) lineage and
may serve as an additional site of B cell generation (45, 46, 47). However,
the structural features of milky spots and the observed lymphoid
aggregates described in this paper are different. The milky spots are
found mainly in the omentum, are relatively small (difficult to see
without magnification), contain mainly macrophages, T cells, B cells,
and a few plasma cells, and are covered by porous epithelium, which
allows the influx and efflux of cells (42, 44, 48, 49). In contrast,
the lymphoid aggregates we observed in hu-PBL-SCID mice were found
frequently in retroperitoneum (in the vicinity of the pancreas),
hepatic hilus, and attached to organs such as the spleen. Typically,
many of the aggregates were larger than putative milky spots, were
encapsulated and contained fully differentiated plasma cells as the
dominant population (Fig. 2
, C and D, and Fig. 3
D). Nevertheless, we cannot exclude the possibility that
some of the structures observed (e.g., Fig. 2
, A and
B, and Fig. 3
, EG) are indeed milky
spots induced in SCID mice by human PBMC. However, in contrast to
mesenteric lymphoid structures of mice, which are known to populate
other tissues with IgA plasma cell precursors, especially the
intestines (45), we did not detect human plasma cells in significant
numbers in the gut and other mucosal tissues of SCID mice. Thus, it
appears that the cells remain in the aggregates without a marked
dissemination into the periphery. Whether this is due to
down-regulation of expression of mucosal or lymph node homing receptors
on human B cells, as a consequence of their differentiation stage, is
at present unknown. This does not seem to be the case with human T
cells, which, especially after infection with HIV, migrate into the gut
of SCID mice (this paper, and 32 .
Transfer of murine Peyers patch (PP) and peripheral and mesenteric lymph node cells (PBMC were not examined) into SCID mice resulted in repopulation of the immune system (50). However, there were some qualitative differences in repopulation of various host tissues depending on the source of the donor lymphoid cells. For example, transfer of PP cells yielded reconstitution of both lamina propria and intraepithelial cell compartments by IgA-producing plasma cells and T cells, while peripheral lymph node cells gave rise to T cells in lamina propria but not to IgA plasma cells (50). Because primate lymphoid cells from PP, spleen, and lymph nodes were unavailable for our experiments, we could not examine their repopulation potential and restricted our studies to PBMC.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Present address: National Institutes of Dental Research, National Institutes of Health, 30 Convent Drive, Bethesda, MD 20892. ![]()
3 Address correspondence and reprint requests to Jiri Mestecky, Department of Microbiology, University of Alabama at Birmingham, 756 BBRB, 845 19th Street South, Birmingham, AL 35294. E-mail address: ![]()
4 Abbreviations used in this paper: HuGH, human growth hormone; H&E, hematoxylin and eosin; AMCA, 7-amino-4-methyl-coumarin-3-acetate; EBNA, Epstein-Barr nuclear Ag. ![]()
Received for publication January 16, 1998. Accepted for publication April 6, 1998.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |