Graft-versus-host disease T cell infiltration and chemokine ... - Nature

Bone Marrow Transplantation (2002) 29, 979–986  2002 Nature Publishing Group All rights reserved 0268–3369/02 $25.00 www.nature.com/bmt

Graft-versus-host disease T cell infiltration and chemokine expression: relevance to the disease localization in murine graft-versus-host disease JY New1, B Li1, WP Koh1, HK Ng1, SY Tan1, EH Yap1, SH Chan1 and HZ Hu1,2 1

Departments of Microbiology, Faculty of Medicine, National University of Singapore, Republic of Singapore; and 2Department of Surgery, University of Wisconsin, Madison, WI, USA

Summary: Acute graft-versus-host disease (GVHD) involves mainly skin, liver and intestines. Other organs such as heart, muscle and central nervous system are seldom affected, although their parenchymal cells also express alloantigens, such as MHC class I antigens. The mechanism of this selective involvement of distinct organs in acute GVHD is not well understood. We postulated that it might be related to the selective migration of activated alloreactive T cells. Indeed, T cell infiltration, revealed by examination of serial samples using flow cytometry and immunohistology, occurred early and continuously in the target organs such as the liver, but not in a nontarget organ, the heart, in a murine acute GVHD model. Since T cell migration is largely controlled by the expression of chemokine and chemokine receptors, we investigated the chemokine spectrum in target/non-target organs of mice with acute GVHD. We found that in the spleen and liver MIP-1␣, MIP-2 and Mig were the predominant chemokines expressed. In another target organ, the skin, MIP-1␣, MIP-2, MCP-1 and MCP-3 were all highly expressed. In a non-target organ of acute GVHD, the heart, the predominant chemokines expressed were MCP-1 and MCP-3. This distinct pattern of chemokine expression in these organs may contribute to the preferential recruitment of inflammatory cells into the liver and skin, but not into the heart, in acute GVHD. Bone Marrow Transplantation (2002) 29, 979–986. DOI: 10.1038/sj/bmt/1703563 Keywords: T lymphocytes; chemokine; graft-versus-host disease; spleen; liver; skin

Graft-versus-host disease (GVHD) occurs in recipients of allogeneic bone marrow transplantation and in immune compromised patients after whole blood transfusion or organ transplantation.1–3 GVHD exhibits acute and chronic Correspondence: Dr HZ Hu, Department of Surgery, CSC Room H4/735, University of Wisconsin, 600 Highland Avenue, Madison, WI 53792, USA Received 2 January 2002; accepted 2 March 2002

patterns.4 Acute GVHD is initiated by allogeneic T cells. After entering the recipient blood circulation, these T cells home into the secondary lymphoid tissues where the alloreactive T cells become activated. These activated T cells proliferate, resulting in an expanded population.5,6 Activated alloreactive T cells migrate from the secondary lymphoid tissues into the peripheral target organs, and initiate the inflammatory process. They destroy the tissue structure and cause various clinical symptoms.7–9 Acute GVHD involves mainly skin, liver and intestines. Other organs such as heart, muscle and central nervous system are usually not affected, although their parenchymal cells also express alloantigens, such as MHC class I antigens.1,10–12 The mechanism of this seemingly selective involvement of distinct organs in acute GVHD is not well understood. We hypothesized that it might be related to the selective migration of activated alloreactive T cells. T cell migration is largely controlled by a group of protein molecules called chemokines. Chemokines are secreted by various cells, such as macrophages, dendritic cells, activated T cells, endothelial cells etc. They are classified as C, CC, CXC and CX3C chemokines according to the N-terminal amino acid sequences. Cells expressing chemokine receptors are attracted to the sites with the highest concentration of chemokines following the concentration gradient.13–15 Expression of MIP-1␣ was elevated and a large number of CD8⫹ cells expressing chemokine receptors CCR1, CCR4, CCR5, CXCR3 were detected in the liver of mice with acute GVHD.12,16 Blockade of T cell migration into liver using anti-CCR5 antibody down-modulated disease activity.12 It seems that the occurrence of acute GVHD is related to chemokine expression. We wondered whether chemokine expression pattern is a factor determining the preferential organ involvement in acute GVHD. As an initial step, we examined T cell infiltration and chemokine expression in target and non-target organs of mice developing acute GVHD. The results revealed a distinct chemokine expression pattern, suggesting that different T cell sub-populations were recruited into the target/nontarget organs.

T cell infiltration and chemokine expression in acute GVHD JY New et al

980

Materials and methods Animals Eight-week-old C57BL/6J (H-2b) mice were used as donors. CB-17 SCID mice (H-2d) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA), and maintained in a flexible film isolator. SCID mice do not have T and B lymphocytes, and acute GVHD can be easily induced.7–9 These mice were used at the age of 4–6 weeks as recipient for studying the chemokine expression profiles during the development of acute GVHD. Reagents and antibodies A green fluorochrome 5-carboxyfluorescein diacetate succinimidyl ester (CFSE) was purchased from Molecular Probes (Eugene, OR, USA). Monoclonal antibodies used in the FACS analysis for the homing study were of rat origin and directed at mouse CD4 (PE), CD8 (PE), and H-2Kb (biotin-conjugated, specific for donor origin cells; Pharmingen, San Diego, CA, USA). Strepavidin-RED670 (Gibco BRL, Gaithersburg, MD, USA) was used for visualizing the antibody binding to H-2Kb. Rat anti-mouse CD3, goat anti-rat Ig (biotin-conjugated) and strepavidin-conjugated horse radish peroxidase (Pharmingen) were used for the immunohistochemical staining of T lymphocytes. Induction of acute GVHD Induction of acute GVHD was previously described by us.7–9 Briefly, mononuclear cells (MNCs) were isolated from the spleens of C57BL/6J donor mice by density centrifugation on Ficoll–Paque (Pharmacia Biotech, Uppsala, Sweden). After washing twice with PBS, 10 ⫻ 106 cells were injected into each recipient CB-17 SCID mouse via the tail vein. The day of injection was recorded as day 0, and thereafter the recipient mice were observed for acute GVHD and weighed daily. FACS Analysis on the Cells Homing to Various Organs Donor splenocytes were labeled with CFSE prior to their injection into the recipient SCID mice. Donor splenocytes were resuspended at 50 ⫻ 106 cells/ml in RPMI 1640 (Gibco BRL), and CFSE was added to a final concentration of 10 ␮m. After incubation at 37°C for 10 min with constant shaking, the reaction was terminated by adding on equal volume of ice-cold RPMI 1640 containing 3% fetal calf serum (FCS), followed by two washes with the same medium. The cells were then resuspended in RPMI 1640. Ten million cells were injected into the recipient mice via the tail vein. Recipient mice were killed on days 1, 3, 4, 7 and 21 post donor cell injection and total body perfusion was performed by infusing 40 ml of PBS into the right and left ventricles through cardiac puncture to flush out the blood cells in each organ. Spleen, liver, heart, mesenteric lymph nodes (MLN) and peripheral lymph nodes (PLN, including inguinal and axillary lymph nodes) were procured from the recipient mice. MNCs of the spleen and liver were isolated using Ficoll–Paque density centrifugation after

Bone Marrow Transplantation

homogenization. Heart, MLN and PLN cells were obtained by mincing the tissues and the tissue debris filtered using a 41 ␮m mesh filter (Sefar, Ruschlikon, Switzerland). All cells were washed twice prior to staining with mAbs. The isolated cells were washed with PBS containing 5% FCS and 5 mm sodium azide. Aliquots of 106 cells were stained in 96-well round bottom PVC microtiter plates. Cells were firstly incubated on ice with biotin-conjugated anti-H-2kb mAb and either PE-labeled anti-CD4 or antiCD8 mAb, followed by strepavidin RED670 staining for 30 min. The stained cells were resuspended in PBS supplemented with 0.5% paraformaldehyde at 4°C until FACS analysis was performed. Cells were counted, 1–5 ⫻ 105 for each preparation, with a Coulter flow cytometer (Epics Esp; Coulter Corporation, Hialeah, FL, USA), and tri-color data analysis was performed with the aid of computer software (FlowJo; Treestar, San Carlos, CA, USA). Lymphocytes were gated based on their low forward scatter and side scatter property, and H-2kb⫹ CD4⫹ or H-2kb⫹ CD8⫹ cells were presented as a percentage of this gated population. CFSE fluorescence reduction was used as an indication for cell proliferation. Immunohistochemical staining The recipient SCID mice were killed on days 1, 3, 7 and 21 after injection of donor cells, and the spleen, liver, heart and skin were procured and snap-frozen in liquid nitrogen and stored until use. Organs derived from non-treated SCID mice were used as controls. A three-step immunoperoxidase staining procedure was performed at room temperature unless otherwise specified. Briefly, 8 ␮m thick sections were prepared from frozen tissues and fixed in a solution containing acetone and chloroform (1:1 ratio). Sections were then rehydrated in Tris-buffered saline (TBS, pH 7.6). Endogenous peroxidase was blocked by using a peroxidase blocking reagent (Dako, Glostrup, Denmark). Non-specific binding of the secondary antibody was prevented by incubating sections with goat serum (1:30 dilution) for 30 min. The primary antibody, rat anti-mouse CD3, was used at 1:300 dilution in PBS. The sections were incubated at 4°C overnight in a humidity chamber. Three rounds of washing with TBS were performed before incubating the sections with the biotin-labeled goat anti-rat mAb (1:250) for 30 min. The sections were washed three times with TBS before the strepavidin-HRP (1:200) was applied. After 30 min incubation, the sections were rinsed and color development was conducted using diamino-benzidine substrate solution (Dako). Finally, the sections were counterstained with Mayer’s hematoxylin and mounted in Fisher Permount (Fisher, Pittsburgh, PA, USA). Real time polymerase chain reaction The recipient SCID mice were killed at various time points after injection of donor cells, and the spleen, liver, heart and skin were procured for RNA isolation. After homogenizing in lysis buffer with a tissue grinder (Polytron PT1200; Kinematica, Littau-Lucerne, Switzerland), RNA was extracted and reverse-transcribed into cDNA using kits from Boehringer Mannheim (Mannheim, Germany). The


mRNA expression of chemokine was detected using the ABI Prism 7700 Sequence Detection System (Perkin Elmer, Wellesley, MA, USA). The Taqman probes and primers specific for MIP-1␣, MIP-2, MCP-1, MCP-3, Mig and ␤-actin were synthesized by Applied Biosystem (Foster City, CA, USA). All probes were labeled at the 5⬘ end with the reporter dye molecule FAM (6-carboxy-fluorescein) and at the 3⬘ end with the quencher dye molecule TAMRA (6carboxytetra methyl-rhodamine). Probe and primer sequences are presented in Table 1. The MIP-1␣ gene was amplified by RT-PCR and cloned into pcDNA 3.1 vector (Invitrogen, Groningen, The Netherlands). This plasmid DNA containing MIP-1␣ was used as the standard. The concentration (␮g/ml) of the recombinant plasmid was determined by a spectrophotometer (␭ ⫽ 260 nm) and then converted into picomole (pmol) using the following formula: ␮g ⫻ 106 pg/1 ␮g ⫻ pmol/660 pg ⫻ 1/N ⫽ pmol, where N is the number of nucleotide pairs and 660 pg/pmol is the average molecular weight of a nucleotide pair. The quantity in pmol can be further converted into plasmid copy number by using the constant of 1 mole ⫽ 6.023 ⫻ 1023 molecules (or copy number). The recombinant plasmid was then serially diluted into copy numbers ranging from 108 to 102 and employed as a standard for other primer/probe and target combinations. Primers and probes applied to the standard were specific for MIP-1␣ gene as shown in Table 1. It was run simultaneously with each sample during every round of real time PCR. Thermal cycler parameters included 2 min at 50°C, 10 min at 95°C and then 45 cycles of denaturation at 95°C for 40 s, annealing at 55°C for 40 s and extension at 60°C for 40 s. Real time monitoring of the fluorescent emission from cleavage of sequence-specific probes by the nuclease activity of Taq polymerase allowed definition of the threshold cycle during the early exponential Table 1

phase of amplification. Quantitative real time PCR was performed to determine the MIP-1␣, MIP-2, MCP-1, Mig and MCP-3 mRNA copy number in each tissue and normalized to copies of the ␤-actin mRNA from the same sample.

981

Enzyme linked immunosorbent assay for chemokine levels in the organs Enzyme linked immunosorbent assay (ELISA) was performed on spleen, liver, heart and skin tissues collected from SCID mice before and after donor cell injection. Mice were killed on indicated days, and total body perfusion was carried out with 40 ml of PBS to avoid chemokine contamination from the blood prior to the isolation of the organs. Whole organs were homogenized using a tissue grinder in the presence of a proteinase inhibitor cocktail (Calbiochem, San Diego, CA, USA). ELISA for assaying the MIP-1␣, MIP-2 (R&D System, Minneapolis, MN, USA) and MCP1 (Pharmingen) was performed on supernatants from the tissue homogenate according to the manufacturer’s instructions. The expression levels of chemokines were normalized to the total protein level of the supernatant from the same tissue measured with the Bio-Rad protein assay dye (Hercules, CA, USA).

ELISA for chemokine levels in the blood Blood was collected weekly from the retro-orbital plexus of mice with acute GVHD until day 42 after donor cell injection. Sera were isolated and stored at ⫺20°C until use. Serum levels of MIP-1␣, MIP-2 and MCP-1 were determined using ELISA kits as previously mentioned. Noninjected mice served as controls.

Sequences of primers and probes employed in real time PCR

Target

Sequences

MIP-1␣

Forward primer: Reverse primer: Probe:

5⬘ GCGCCATATGGAGCTGACAC 3⬘ 5⬘ TCAGGCATTCAGTTCCAGGT 3⬘ 5⬘ CCTGCTGCTTCTCCTACAGCCGGA 3⬘

MIP-2


5⬘ TGACTTCAAGAACATCCAGAGCTT 3⬘ 5⬘ CTTGAGAGTGGCTATGACTTCTGTCT 3⬘ 5⬘ AGGACCCCACTGCGCCCAGA 3⬘

MCP-1


5⬘ GCTGGAGAGCTACAAGAGGATCA 3⬘ 5⬘ TCTCTCTTGAGCTTGGTGACAAAA 3⬘ 5⬘ CTACAGCTTCTTTGGGACACCTGCTGCT 3⬘

Mig


5⬘ AGTGATAAGGAATGCACGATGCT 3⬘ 5⬘ TGAGGTCTTTGAGGGATTTGTAGTG 3⬘ 5⬘ CAGCACCAGCCGAGGCACGA 3⬘

MCP-3


5⬘ GGGAAGCTGTTATCTTCAAGACAAA 3⬘ 5⬘ CTCCTCGACCCACTTCTGATG 3⬘ 5⬘ CTTCAGCGCAGACTTCCATGCCCTT 3⬘

␤-actin


5⬘ TTCAACACCCCAGCCATGTA 3⬘ 5⬘ TGTGGTACGACCAGAGGCATAC 3⬘ 5⬘ TAGCCATCCAGGCTGTGCTGTCCC 3⬘

Primers were confirmed for the specificity by regular PCR (data not shown). Probes are labeled at the 5⬘ end with the reporter dye molecule FAM (6-carboxy-fluorescein) and at the 3⬘ end with the quencher dye molecule TAMRA (6-carboxytetra methyl-rhodamine). Bone Marrow Transplantation


982

MLN

spleen

liver

heart

MLN

spleen

liver

heart

Day 1

CD8PE

CD4PE

Day 4

Day 7

Day 21 CFSE

CFSE

Figure 1 Representative distribution and proliferation of donor T cells in various recipient organs. Donor spleen cells were isolated and labeled with CFSE prior to the injection into the recipient mice. Mononuclear cells were then isolated from recipient spleen, liver, heart, MLN at different time points post transplantation and analyzed by flow cytometry. Representative flow cytometric profiles show the expression of CD4 or CD8 vs CFSE by the H-2kb⫹ donor cells isolated from the recipient mice on days 1, 4, 7 and 21 after the allogeneic cell transfer. Five mice were examined.

a

An acute GVHD murine model was employed as reported previously.7–9 Spleen, liver, skin and heart were harvested from euthanized C.B-17 SCID mice (H-2d) that received spleen cells obtained from C57BL/6J (H-2b) mice. These organs were chosen to represent the target (spleen, liver and skin) and non-target (heart) organs of acute GVHD. The infiltration of T cells was examined initially with FACS analysis in these organs. Freshly harvested organs were minced and mononuclear cells were isolated. Cells were stained with antibodies directed at CD4, CD8 and H2kb molecules. As presented in Figures 1 and 2, both donor CD4⫹ and CD8⫹ cells were detected in the secondary lymphoid tissues and the liver in the first 4 days after donor cell infusion. In the samples obtained 7 and 21 days after donor cell infusion, CD4⫹ and CD8⫹ cells were observed in the heart as well, and the percentages of these two cell subsets increased substantially in all organs with the lowest being in the heart. From day 7 onwards, the majority of infiltrating lymphocytes were CD8⫹ T cells, consisting of up to 60% of total T cell population in the target organs. Donor cell proliferation was examined by labeling the donor splenocytes with CFSE before they were injected into the recipient mice. Cells were then isolated from various organs of the recipient mice at different time points. As shown in Figure 1, the donor cells homed to the lymphoid organs (LNs, spleen) and to a much lesser extent to the liver as well 1 day after the injection. Cell division started on day 3, peaked on day 4, and subsided by day 21. No proliferation, however, was observed in the target organ (liver). The liver infiltrating T cells were effector cells that had divided. Since FACS analysis can only show the percentage of a specific cell population and is not able to indicate the absolute infiltrating cell numbers, we examined the infiltrating T cells in these organs by immunohistology on tissue sections. T cells were visualized with an anti-CD3 antibody staining, and they were easily observed in the liver and skin Bone Marrow Transplantation

70

MLN

PLN

Spleen

Liver

Heart

60 50 40 30 20 10 0

1 3 4 7 21

1 3 4 7 21 1 3 4 7 21 1 3 4 7 21 Days post transplantation

1 3 4 7 21

1 3 4 7 21 1 3 4 7 21 1 3 4 7 21 Days post transplantation

1 3 4 7 21

b 70 Percentage of CD8+ cells

T cell infiltration in the target and non-target organs

Percentage of CD4+ cells

Results

60 50 40 30 20 10 0

1 3 4 7 21

Figure 2 Kinetics of infiltration of (a) CD4 and (b) CD8 donor cells into various recipient organs at different time points after the allogeneic T cell transfer. Donor spleen cells were isolated and labeled with CFSE prior to the injection into the recipient mice. Mononuclear cells were then isolated from recipient spleen, liver, heart, MLN and PLN at different time points post transplantation and analyzed by flow cytometry. Lymphocyte population was gated based on the low FS and SS property. The mean percentage of H-2kb⫹ CD4⫹ or H-2kb⫹ CD8⫹ cells within the gated lymphocyte population was presented. The data represent the mean ⫾ s.d. of five mice in each group.

7 days after donor cell infusion, but only rarely in the heart (Figure 3). mRNA expression of chemokines in the target and nontarget organs Spleen, liver, skin and heart were harvested from SCID mice at various time points after injection of the donor


levels on day 21. Expression of MIP-1␣ and MIP-2 was higher in the spleen. In the liver, MIP-1␣, MIP-2 and Mig predominated. MIP-2 and MCP-3 were expressed at a higher level in the skin. Expression of MCP-1 and MCP3 was more prominent in the heart. Figure 3 Immunohistochemical staining for detection of infiltrating CD3⫹ cells in various organs of the recipient mice. Target and non-target organs were procured from the recipient mice at different time points post transplantation. Cryosections were prepared and stained with an anti-CD3 mAb to visualize infiltrating T cells. Representative cryosections of tissues collected at day 21 post transplantation show a large number of T cells infiltrating in the liver (a) and the skin (b), but not in the heart (c). These results correlate with those observed in Figures 1 and 2.

spleen cells. RNA was isolated for cDNA synthesis, and real time PCR was performed to quantify chemokine mRNA expression in these organs. As depicted in Figure 4, expression of MIP-1␣, MIP-2, Mig, MCP-1 and MCP3 increased 3 days after donor cell injection, peaked on day 7, and, except for MIP-2 in the skin, returned to baseline a

983

Expression of chemokines at the protein level in the target and non-target organs Spleen, liver, skin and heart were harvested from SCID mice at various time points after injection of the donor spleen cells. ELISA was conducted on the supernatant collected from the tissue homogenate, and the quantity of chemokines was normalized to that of the total protein in each organ. The quantity of three chemokines, MIP-1␣, MIP-2 and MCP-1, was determined (Figure 5). Expression of the chemokines at the protein level was correlated with that at the mRNA level. MIP-1␣ and MIP-2 were predominant in the spleen and the liver. In the heart, a high level of expression of MCP-1 was observed on day 7 after donor

b

c

d MIP-1a

MIP-2

MCP-1

Mig

MCP-3

liver

heart

Chemokine/ -actin level

spleen

skin

Days post transplantation

Figure 4 Kinetics of chemokine expression in various organs of the recipient mice after the injection of donor cells. (a) Amplification plot of the standard used in real time PCR. Ten-fold serial dilutions of MIP-1␣-pcDNA 3.1 recombinant plasmids (from 108 to 102 copy number) were amplified for 45 cycles. The amplification plot shows that copy number of target gene as low as 102 can be reliably detected. (b). Standard curve derived from the amplification plot shown in (a). The regression line was obtained using the Sequence Detection System software provided by PE Biosystem. Correlation coefficient ⫽ 0.997. (c). Representative amplification plot of RNA samples extracted from different tissues. RNA was isolated from day 7 spleen, liver, skin and heart of each recipient mouse and reverse-transcribed into cDNA. Real time PCR was performed with primers and probes specific for MIP1␣ sequence (Table 1). From left to right, the curves indicate spleen, liver, skin and heart cDNA samples as substrate respectively. (d) Expression profiles of MIP-1␣, MIP-2, MCP-1, Mig and MCP-3 mRNA in various recipient organs. The quantity of mRNA for each chemokine was normalized to the expression of ␤-actin in the same sample. (䉬) The ratio in each sample; (----) the mean of the three samples. Bone Marrow Transplantation


984

MIP-1 spleen

skin

Chemokine/ total protein level

heart

MCP-1

0.0045

0.09

0.04

0.003

0.06

0.02

0.0015

0.03

0

liver

MIP-2

0.06

0

1

3

7

0

21

0

1

3

7

21

0

0.06

0.0045

0.09

0.04

0.003

0.06

0.02

0.0015

0.03

0

0

1

3

7

0

21

0

1

3

7

21

0

0.06

0.0045

0.09

0.04

0.003

0.06

0.02

0.0015

0.03

0

0

0

1

3

7

21

0

1

3

7

21

0

0.36

0.036

0.42

0.24

0.024

0.28

0.12

0.012

0.14

0

0

0

1

3

7

21

0

1

3

7

21

0

0

1

3

7

21

0

1

3

7

21

0

1

3

7

21

0

1

3

7

21

Days post transplantation

Figure 5 Expression profiles of MIP-1␣, MIP-2 and MCP-1 at the protein level in various organs of the recipient mice. Spleen, liver, heart and skin were procured from the recipient mice at various time points after allogeneic cell transfer. ELISA was performed on the supernatant collected from the tissue homogenate to determine the chemokine levels. The chemokine levels were then normalized to that of the total protein of the same sample tissue. The data represent the mean ⫾ s.d. of three samples in each condition.

cell injection. Increased levels of expression of MIP-1␣, MIP-2 and MCP-1 were observed in the skin.

Blood level of chemokines in the SCID recipient mice Ten SCID mice were each injected with 10 ⫻ 106 donor spleen cells, and five others were injected with PBS as controls. Their body weights were monitored daily. The control mice all gained weight during the observation period. However, the weight of the treated mice started to decline by day 7 after a brief increase in the first several days and six mice died by day 35 because of acute GVHD (data not shown) which was similar to our previous observations.7–9 Blood samples were taken weekly from the SCID mice receiving donor cell infusion and from the control SCID mice. Sera were separated and used for ELISA to determine the concentrations of chemokines MIP-1␣, MIP-2 and MCP-1. Levels of the three chemokines increased in mice that received donor cell infusion, but not in the control mice (Figure 6). The kinetics was similar to that of the tissues examined. MIP-1␣ was elevated 3 days after the cell infusion, peaked on day 14 and remained higher than before infusion until day 28. MIP-2 started to increase 7 days after cell infusion, peaked on day 21 and remained higher than before infusion until day 35. Serum MCP-1 began to increase 3 days after donor cell injection, reached a prominent peak on day 7 and subsided to a slightly higher level than before infusion until day 35 in most of the mice. Bone Marrow Transplantation

Discussion The inflammatory reaction in acute GVHD begins with T cell infiltration. Although all nucleated cells express major histocompatibility complex class I molecules on the cell surface which are major antigens for activating alloreactive T cells, T cell infiltration in acute GVHD does not occur evenly in the body. T cells preferentially infiltrate certain organs, such as liver, skin and intestines, and cause serious damage to these organs, but other organs such as heart, muscle and neural system are apparently unaffected.1,10,11 This discrepancy may be due to the factors that govern T cell migration. Of these factors, chemokines play a central role by determining the direction and T cell subpopulation in the trafficking and infiltration process. Chemokines are secreted by various cells, and form a local high concentration to recruit inflammatory cells. Inflammatory cells usually express several chemokine receptors on the cell surface, and each cell type with distinct effector functions expresses different chemokine receptors. CD4⫹ effector T cells generated in vitro express a broad range of chemokine receptors, some of which are preferentially expressed on Th1 cells (CXCR3, CCR5) or Th2 cells (CCR3, CCR4, CCR8).17–20 This suggests that distinct chemotactic signals direct the localization of Th1 and Th2 cells to the sites of inflammation. In the spectrum of chemokines we investigated, chemokine receptors CCR1 and CCR5 are specific for MIP-1␣, CXCR2 for MIP-2, CXCR3 for Mig, CCR2 for MCP-1, and CCR1, CCR2 and


MIP-1 (pg/ml)

100 80 60 40 20 0

0

3

7 14 21 28 Days post transplantation

35

42

0

3


35

42

0

3


35

42

MCP-1 (pg/ml)

1200 1000 800 600 400 200 0

MIP-2 (pg/ml)

700 600 500 400 300 200 100 0

Figure 6 Kinetics of MIP-1␣, MIP-2 and MCP-1 in the blood during acute GVHD. Blood samples were obtained from the recipient mice once a week up to week 6 after the transfer of allogeneic T cells. ELISA was performed to quantitate the serum chemokine levels. Data represent 10 recipient mice with acute GVHD (closed box) and five controls (open circle).

CCR3 for MCP-3.17,18 From the results generated by real time PCR and ELISA on various organs harvested from mice after injection of donor cells, we found that in the spleen and liver, target organs of acute GVHD, MIP-1␣, MIP-2 and Mig were the predominant chemokines expressed. In another target organ, the skin, MIP-1␣, MIP2, MCP-1 and MCP-3 were all highly expressed. In a nontarget organ of acute GVHD, the heart, the predominant chemokines expressed were MCP-1 and MCP-3. This distinct pattern of chemokine expression in these organs may contribute to the preferential recruitment of inflammatory cells into the liver and skin, but not into the heart, in acute GVHD. The predominantly expressed chemokines MIP-1␣ and Mig in the spleen, liver and skin recruit Th1 cells, which have been shown to play the central role in mediating alloimmune response in acute GVHD.1,21 A recent study by Murai et al12 characterized the liver infiltrating CD8⫹ T cells. They found a large population of donor CD8⫹ T cells in the liver, causing both portal hepatitis and non-suppurative destructive cholangitis. These cells expressed CCR5. In contrast, the chemokines expressed in the heart of SCID mice, after injection of

donor cells, were mainly MCP-1 and MCP-3. These two chemokines do not attract the acute GVHD mediating CCR5⫹ and CXCR3⫹ helper 1 or CD8⫹ effector T cells. MIP-2 was another chemokine that was up-regulated in target organs but peaked later than MIP-1␣ and Mig as shown in the serum and tissue ELISA results. Its receptor, CXCR2, is predominantly up-regulated on neutrophils, monocytes and eosinophils.19,22,23 This chemokine might be responsible for recruiting these inflammatory cells into the target organs at a later stage of acute GVHD. Serody et al16 demonstrated that the recruitment of CD8⫹ T cells to acute GVHD liver relies heavily on MIP-1␣. However, using MIP-1␣ knockout splenocytes as donor cells failed to block CD4⫹ cell infiltration into acute GVHD liver. Murai et al12 demonstrated that CD4⫹ T cells expressed much lower CCR5 than CD8⫹ T cells. These observations indicate that chemokines other than MIP-1␣ play a critical role in the recruitment of CD4⫹ T cells into the target organs. It was recently reported that CXCR3 was up-regulated on CD4⫹ T cells during allogeneic activation,24 and the use of anti-Mig Ab blocked the infiltration of CD4⫹ cells into skin allograft and significantly delayed the rejection.25 These results correlate with our distinct chemokine expression patterns in the target and non-target organs of murine acute GVHD. Blocking antibodies directed at CCR1, CCR5 and CXCR3 or chemokine receptor antagonists have been used in preventing kidney allograft rejection26,27 and acute GVHD12 in animal studies. These treatments could dramatically reduce the infiltration of CCR5 or CXCR3 expressing T cells, and consequently reduce the damaging inflammatory activity. In summary, we have documented the chemokine spectrum in target/non-target organs of mice with acute GVHD. A distinctive pattern of chemokines expressed in target/non-target organs has been observed. Although each chemokine may bind more than one receptor, and the complexity of the selective infiltration of certain organs in acute GVHD remains to be elucidated, at least the chemokine spectrum preferentially expressed in different organs may play an important role. Based on the difference of chemokine expression observed, it is possible to modulate acute GVHD by changing the micro-environmental chemokine repertoire in reducing the recruitment of alloreactive cells. The simultaneous upregulation of multiple chemokines during acute GVHD as demonstrated in our study suggests that the blocking of two or more chemokines or chemokine receptors might be required in order to effectively ameliorate the disease.

985

Acknowledgements This work was supported by grants 30030, NMRC/0348/1999, NMRC/0327/1999 (HZH) from the National Medical Research Council, Singapore, and a grant from Leukemia Research Foundation (HZH).

References 1 Ferrara JL, Deeg HJ. Graft-versus-host disease. New Engl J Med 1991; 324: 667–674. Bone Marrow Transplantation


986

2 Brok HP, Heidt PJ, van der Meide PH et al. Interferon-␥ prevents graft-versus-host disease after allogeneic bone marrow transplantation in mice. J Immunol 1993; 151: 6451–6459. 3 Dickinson AM, Sviland L, Hamilton PJ et al. Cytokine involvement in predicting clinical graft-versus-host disease in allogeneic bone marrow transplant recipients. Bone Marrow Transplant 1994; 13: 65–70. 4 Allen RD, Staley TA, Sidman CL. Differential cytokine expression in acute and chronic murine graft-versus-host disease. Eur J Immunol 1993; 23: 333–337. 5 Bouma GJ, van der Meer-prins E, van der Woude FJ et al. Relevance of pretransplant donor-specific T cell allorepertoire for human kidney graft survival. Transplantation 1995; 59: 969–976. 6 Hu HZ, de Jonge N, Doornewaard H et al. Cytotoxic T lymphocytes infiltrating the human cardiac allograft show a restriction in T-cell receptor V beta gene usage: a study on serial biopsy and blood specimens. J Heart Lung Transplant 1994; 13: 1058–1071. 7 Hu HZ, Li GL, Lim YK et al. Kinetics of interferon-␥ secretion and its regulatory factors in the early phase of acute graft-versus-host disease. Immunology 1999; 98: 379–385. 8 Hu HZ, New JY, Li B et al. Depletion of activated alloreactive T cells in prevention of acute graft-versus-host disease. Transplant Proc 2000; 32: 2456–2457. 9 Li B, New JY, Yap EH et al. Blocking L-selectin and ␣4integrin changes donor cell homing pattern and ameliorates murine acute graft versus host disease. Eur J Immunol 2001; 31: 617–624. 10 Nikolic B, Lee S, Bronson RT et al. Th1 and Th2 mediate acute graft-versus-host disease, each with distinct end-organ targets. J Clin Invest 2000; 105: 1289–1298. 11 Hakim F, Fowler DH, Shearer GM, Gress RE. Animal models of acute and chronic graft-versus-host disease. Curr Protocols Immunol 1998; S27: 4.3.1–4.3.21. 12 Murai M, Yoneyama H, Harada A et al. Active participation of CCR5⫹ CC8⫹ T lymphocytes in the pathogenesis of liver injury in graft-versus-host disease. J Clin Invest 1999; 104: 49–57. 13 Zlotnik A, Yoshie O. Chemokines: a new classification system and their role in immunity. Immunity 2000; 12: 121–127. 14 Baggiolini M. Chemokines and leukocyte traffic. Nature 1998; 392: 565–568.

Bone Marrow Transplantation

15 Rollins BJ. Chemokines. Blood 1997; 90: 909–928. 16 Serody JS, Burkett SE, Panoskaltsis-Mortari A et al. T-lymphocyte production of macrophage inflammatory protein-1␣ is critical to the recruitment of CD8⫹ T cells to the liver, lung and spleen during graft-versus-host disease. Blood 2000; 96: 2973–2980. 17 Bonecchi R, Bianchi G, Bordignon PP et al. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s. J Exp Med 1998; 187: 129–134. 18 Sallusto F, Lanzavecchia A, Mackay CR. Chemokines and chemokine receptors in T-cell priming and Th1/Th2-mediated responses. Immunol Today 1998; 19: 568–574. 19 Nelson PJ, Krensky AM. Chemokines, chemokine receptors and allograft rejection. Immunity 2001; 14: 377–386. 20 Zingoni A, Soto H, Hedrick JA et al. The chemokine receptor CCR8 is preferentially expressed in Th2 but not Th1 cells. J Immunol 1998; 161: 547–551. 21 Lu Y, Sakamaki S, Kuroda H et al. Prevention of lethal acute graft-versus-host disease in mice by oral administration of T helper 1 inhibitor, TAK-603. Blood 2001; 97: 1123–1130. 22 Wolpe SD, Sherry B, Juers D et al. Identification and characterization of macrophage inflammatory protein 2. Proc Natl Acad Sci USA 1989; 86: 612–616. 23 Mercer-Jones MA, Shrotri MS, Heinzelmann M et al. Regulation of early peritoneal neutrophil migration by macrophage inflammatory protein-2 and mast cells in experimental peritonitis. J Leukoc Biol 1999; 65: 249–255. 24 Ebert LM, McColl SR. Coregulation of CXC chemokine receptor and CD4 expression on T lymphocytes during allogeneic activation. J Immunol 2001; 166: 4870–4878. 25 Koga S, Auerbach MB, Engeman TM et al. T cell infiltration into class II MHC-disparate allografts and acute rejection is dependent on the IFN-␥-induced chemokine Mig. J Immunol 1999; 163: 4878–4885. 26 Grone HJ, Weber C, Weber KS et al. Met-RANTES reduces vascular and tubular damage during acute renal transplant rejection: Blocking monocyte arrest and recruitment. FASEB J 1999; 13: 1371–1383. 27 Hancock WW, Lu B, Gao W et al. Requirement of the chemokine receptor CXCR3 for acute allograft rejection. J Exp Med 2000; 192: 1515–1520.