combined immunodeficient mice transplanted with ...

Immunology 1998 93 524-532

Peritoneal macrophages play an important role in eliminating human cells from severe combined immunodeficient mice transplanted with human peripheral blood lymphocytes S. SHIBATA,*§ T. ASANO,* A. NOGUCHIt M. NAITO,t A. OGURAt & K. DOI§ *Division for Experimental Animal Research, tDepartment of Veterinary Science, National Institute of Infectious Diseases, Tokyo, tthe Second Department of Pathology, Niigata University School of Medicine, Niigata, and §Laboratory of Veterinary Pathology, Faculty of Agriculture, University of Tokyo, Tokyo, Japan

SUMMARY To elucidate the mechanism of human cell elimination from severe combined immunodeficient (SCID) mice transplanted with human peripheral blood lymphocytes (hu-PBL-SCID mice), we explored the immunocytes in the peritoneal cavity in SCID mice where human PBL were transferred. When the phenotype of peritoneal exudate cells (PEC) was compared by flow cytometry among three congenic strains of SCID mice that differ in their acceptability for human PBL, the PEC in NOD-scid mice, which exhibit the highest acceptability, contained the smallest number of F4/80l1/-Mac-1 +-activated macrophages. Moreover, the proportions of natural killer cells in PEC of the three strains of SCID mice were not always correlated with the acceptability. These findings suggest the possibility that peritoneal macrophages eliminate human cells in hu-PBL-SCID mice. To verify this hypothesis, we evaluated the engraftment of human PBL into SCID mice that were treated with liposome-encapsulated dichloromethylene diphosphonate, which selectively depletes macrophages by inducing apoptosis, or 8-aminoguanidine hemisulphate salt, an inhibitor of inducible nitric oxide synthase of macrophages. As a result, both of these regimens improved engraftment of human PBL, indicating that peritoneal macrophages take part in human cell elimination in the peritoneal cavity of hu-PBL-SCID mice and that it is mediated, at least in part, by direct macrophage cytotoxicity utilizing nitric oxide.

scid (CB 17-scid) mice,6,7 the most commonly used SCID mouse strain. Interestingly, the engraftment of human PBL is dependent on the SCID mouse strain used as recipient. For example, NOD-scid mice show higher acceptability for human cells than CB17-scid mice.89 This observation also supports the notion that innate immunity, which is not affected by scid mutation but is controlled by other genetic background genes, plays an important role in the host's immune response against transferred human cells in hu-PBL-SCID mice. NOD-scid mice are reported to show multiple deficiency in innate immunity such as impaired natural killer (NK) cell activity, deficient macrophage function and lack of complement activity.'0 But the critical factor for human PBL engraftment has not yet been addressed. NK cells have long been considered to be the major effector cells that eliminate human cells in hu-PBL-SCID mice."'3 However, recent studies throw doubt upon that notion. The scid and beige double-mutant mice, which show impaired NK activity as well as deficiency in acquired immunity, do not show higher engraftment of xenogenic tissues than SCID mice."' In addition, NK cell-depleted C57BL/6-scid (B6-scid) mice by anti-NKl.1 monoclonal antibody treatment and B6scid beige mice did not exhibit superior engraftment of human

INTRODUCTION Hu-PBL-SCID mice, which are severe combined immunodeficient (SCID) mice transplanted with human peripheral blood lymphocytes (PBL) into their peritoneal cavity,' are utilized for studies on human haematopoiesis and human infectious diseases, most notably on human immunodeficiency virus infection.2'3 SCID mice accept human PBL since they lack mature functional T and B lymphocytes,4 but the engraftment is not always sufficient.5 That suggests a role of innate immunity in recipient mice, whose integrity is preserved in C.B-17Received 29 July 1997; revised 3 November 1997; accepted 10 December 1997. Abbreviations: 8-AG, 8-aminoguanidine hemisulphate salt; FSC, forward-scatter; hu-PBL-SCID mice, SCID mice transplanted with human peripheral blood lymphocytes; iNOS, inducible nitric oxide

synthase; liposome-MDPCl2, liposome-encapsulated dichloromethylene diphosphonate; liposome-PBS, liposome containing phosphatebuffered saline; NO, nitric oxide; PEC, peritoneal exudate cells; SCID, severe combined immunodeficient; SSC, side-scatter. Correspondence: Dr S. Shibata, Division for Experimental Animal Research, National Institute of Infectious Diseases, 1-23-1 Toyama,

Shinjuku-ku, Tokyo 162, Japan.

524

5

1998 Blackwell Science Ltd

Peritoneal macrophages in hu-PBL-SCID mice PBL as compared to naive B6-scid mice, either.'5 These observations strongly suggest that NK cells are not the major effector cells that eliminate human cells in hu-PBL-SCID mice, though they may have relatively minor roles. Recently, we found that engraftment of human PBL is improved when recipient SCID mice are depleted of murine interferon-y (IFN-y) by monoclonal antibody treatment (submitted for publication). Since the major target cells for IFN-y are macrophages,'6 and macrophages activated by IFN-'y exhibit cytotoxicity against tumour cells;'7 we hypothesized that peritoneal macrophages play an important role in elimination of human cells in hu-PBL-SCID mice. Macrophage depletion in the spleen and liver has already been shown to facilitate migration and engraftment of human cells into the spleen;18 however, the role of macrophages in the peritoneal cavity where human PBL were transferred remains to be elucidated. The aim of the present study is to examine the role of peritoneal macrophages in the elimination of human cells in hu-PBL-SCID mice. We first described the immunocyte population in the peritoneal cavity where human PBL are transferred. And we compared the phenotype of peritoneal exudate cells (PEC) among three strains of congenic SCID mice, which differ in their acceptability for human PBL. This result suggests an involvement of activated macrophages in the elimination of human cells in hu-PBL-SCID mice. Subsequently, we evaluated engraftment of human PBL in peritoneal macrophage-depleted SCID mice and inducible nitric oxide synthase (iNOS) -inhibited SCID mice. Our findings indicate an important role of peritoneal macrophages in the elimination of human cells in hu-PBL-SCID mice. MATERIALS AND METHODS Mice CB17-scid (C.B-17/Icr-scid) mice were obtained from the breeding colony at National Institute of Infectious Diseases (Tokyo, Japan) or purchased from Clea Japan Inc. (Tokyo, Japan). NOD-scid (NOD/LtSz-scid) mice and B6-scid (C57BL/6J-scid/SzJ) mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME) and have been maintained at the National Institute of Infectious Diseases. They were used at 5-10 weeks of age. All mice were maintained under specific pathogen-free conditions. Human cell transfer Mononuclear cells were separated from human peripheral blood by Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) density gradient centrifugation. Each mouse was intraperitoneally injected with 20 x 106 human peripheral blood mononuclear cells. Some B6-scid mice were y-irradiated from a "'Cs source (Gammacell 40, Atomic Energy of Canada Ltd., Kanata, Canada) 1 day before human cell transfer. Two weeks after human cell transfer, the mice were killed and analysed in all the experiments.

Depletion ofmacrophages To deplete peritoneal macrophages in recipient mice, CB17scid mice were intraperitoneally given 0-1 ml of liposomeencapsulated dichloromethylene diphosohonate (liposomeMDPC12), which selectively depletes macrophages by inducing apoptosis.'9 Liposome-MDPCI2 was prepared as previously C 1998 Blackwell Science Ltd, Immunology, 93, 524-532

525

described.'9'20 Liposome-MDPCl2 was given 1 day before and 7 days after human cell transfer. Control mice were dosed with liposome containing only phosphate-buffered saline (liposome-PBS) or with PBS alone in the same way. Inhibition of iNOS To inhibit iNOS in recipient mice, CB17-scid mice were intraperitoneally given 8-aminoguanidine hemisulphate salt (8-AG, Sigma, St. Louis, MO), an inhibitor of iNOS as previously described21 with little modification. Briefly, 8-AG was dissolved in PBS and 1 mg of 8-AG was given intraperitoneally to a mouse twice per day. Control mice were given PBS alone.

Flow cytometry The peritoneal cavity of each mouse was flushed by Hanks' balanced salt solution containing 2% fetal calf serum, and the number of recovered cells was regarded as the total number of PEC. Single cell suspensions of PEC were stained with the following antibodies: fluorescein isothiocyanate (FITC)-conjugated anti-mouse pan-NK cells (DX5) and anti-mouse Gr-l (RB6-8C5), and biotin-conjugated anti-mouse I-A (25-9-17) were purchased from Pharmingen (San Diego, CA). Phycoerythrin (PE) -conjugated anti-mouse macrophage (F4/80) was from Caltag (San Francisco, CA). Anti-human CD45 [4B2, American Type Culture Collection (ATCC) HB-196] was purified from ascites and labelled with FITC in our laboratory. Anti-H-2 (M1/42.3.9.8.HLK, ATCC TIB-126), anti-mouse macrophages (F4/80, ATCC HB-198) and anti-mouse Mac-I (M1/70.15.l 1.5.HL) were purified from serum-free tissue culture supernatant and labelled with FITC or biotin in our laboratory. Biotinylated antibodies were developed with UltraAvidin-PE (Linco technologies Inc., Ballwin, MO) or streptavidin-TRI-COLOR (Caltag). Cells were analysed by fluorescence-activated cell sorter (FACScan; Becton Dickinson, Mountain View, CA). Dead cells were excluded by incorporation of 7-aminoactinomycin D (Sigma) or based on morphology in forward-scatter (FSC) versus sidescatter (SSC) plots. Determination of human immunoglobulin concentration Sandwich enzyme-linked immunosorbent assay (ELISA) was performed to quantify serum human IgG and IgM concentration in each hu-PBL-SCID mouse. Microculture plates (#3690, Costar, Cambridge, MA) were coated with either affinity-purified goat anti-human IgG (Rockland, Gilbertsville, PA) or purified monoclonal anti-human IgM (Pharmingen). Affinity-purified alkaline phosphatase-conjugated goat antihuman IgG and anti-human IgM. (American Qualex, La Mirada, CA) were used to detect human IgG and IgM respectively. The absorbance at 405 nm was quantified on an ELISA reader (Multiskan Bichromatic, Labsystems, Helsinki, Finland). Statistics Statistical analyses were performed using Student's t-test or its variant, as appropriate.

526

S. Shibata et al.

RESULTS Comparison of human PBL engraftment among the three strains of congenic SCID mice We compared human PBL engraftment in the peritoneal cavity among three strains of congenic SCID mice, namely CB17scid, NOD-scid, and B6-scid mice (Table 1). As reported previously,8"9 NOD-scid mice showed superior acceptance for human PBL as compared to other strains of SCID mice, while human PBL engraftment in naive B6-scid mice is hardly observed. Since y-irradiated B6-scid mice show higher engraftment of human PBL than naive B6-scid mice, the poor engraftment in naive B6-scid mice is assumed to be due to higher activity of innate immunity existing in B6-scid mice

i-SSC ssc

C.)

Comparison of PEC phenotype among the three strains of congenic SCID mice We examined and compared PEC phenotype in the three strains of SCID mice by flow cytometry (Fig. la). SSC of PEC in NOD-scid mice is relatively lower than those in the other two strains of SCID mice. PEC in all of these SCID mice strains consist of a large proportion of macrophages and a relatively small ratio of NK cells, while the total number of PEC was not significantly different (Fig. 2a-c). Gr-1 + granulocytes are virtually absent (Fig. 2d). Interestingly, we can find

CB1 7-scid

(a)

o

than those in other strains of mice. Thus, acceptability for human PBL tends to be NOD-scid>CBI7-scid>B6-scid mice.

B6-scid

NOD-scid

i I

FSC

L-6

1^ .0

4.'I

I 87.8

M7 79-2

NK

NK 3.6

Is