| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,§
*
Division of Oncology, Departments of Medicine,
Comparative Medicine, and
Pathology, University of Washington, Seattle, WA 98195; and
§
Fred Hutchinson Cancer Research Center, Seattle, WA 98104
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The classical manifestations of GVHD involve damage to host skin, liver, and gut. One view is that donor T cells recognize Ags expressed by host skin, liver, and gut or possibly epithelial stem cells present in skin, liver, and gut (3). In that circumstance, the Ag-bearing cells would be directly lysed. An alternative view is that GVHD toxicity is mediated by cytokines released by donor T cells, i.e., "cytokine storm" (4). In that circumstance, Ags expressed by skin, liver, and gut need not necessarily be direct targets of donor T cells. The theories of GVHD are perplexing, in part because the target cells and target Ags have not been adequately defined.
One method to elucidate the role of cytokine storm in GVHD would be to induce GVHD with T cells specific for Ags expressed only by hemopoietic cells. If direct T cell cytolytic activity against skin, liver, and gut is a necessary component of GVHD damage to those organs, T cell responses to hemopoietic cells would not induce the classical manifestations of GVHD. It is known that donor T cells can recognize Ags expressed by host hemopoietic cells. Several studies have shown with functional T cell assays that some MiH have a distribution limited to hemopoietic cells (5, 6, 7, 8). However, MiH Ags expressed exclusively by host hemopoietic cells have not yet been identified at the molecular level, and the character of GVHD directed against such Ags has not been defined.
To examine the role of donor T cell reactivity against host hemopoietic cells in the causation of GVHD, we selected CD45 (common leukocyte Ag) for study as the target Ag. CD45 is one of the most abundant leukocyte cell surface glycoproteins and is expressed by all cells of the hemopoietic system. CD45 is known to be polymorphic (9, 10). T cell responses to the polymorphic regions of CD45 had not been reported previously, but the fact that Ab to polymorphic CD45 protein can be generated provided some assurance that Th cells could also be generated (9).
The polymorphic segments of CD45 protein are in the distal extracellular region. Different isoforms of CD45 exist and are used to define functional lymphocyte subpopulations. The isoforms are created by alternative splicing and are expressed in cell type-specific patterns on functional subpopulations of lymphocytes. The regions containing the allelic polymorphic segments are present on all splice variants and are therefore expressed on all cells of hemopoietic origin regardless of the CD45 isoform (11, 12).
The current study demonstrates that the polymorphic segment of the CD45
is immunogenic to T cells. CD4+ T cells generated by
immunizing to peptides representing the polymorphic region of CD45
proliferated vigorously in response to lymphocytes expressing the
appropriate CD45 allelic protein. When tested in vivo, CD45-specific T
cells mediated substantial lung injury. The results confirm that T
cells specific for hemopoietic differentiation Ags can mediate a
substantial graft-vs-host response (GVHR), but provide no evidence that
such T cells can induce classic toxicities of clinical GVHD. The T cell
clone had a Th1 phenotype in secreting IFN-
, but secreted no
detectable IL-2. Further studies with T cell clones specific for CD45,
but capable of secreting a different family of cytokines might further
elucidate the role of cytokine storm in GVHD.
One purpose of examining donor T cell immunity to host hemopoietic differentiation Ags is the possibility that T cell responses to host hemopoietic differentiation Ags might be used to eliminate residual host leukemia cells following BMT, i.e., to mediate a graft-vs-leukemia (GVL) response. The results validate that T cells specific for abundant, polymorphic hemopoietic differentiation Ags can mediate substantial effects in vivo. Although the substantial pulmonary toxicity observed could preclude targeting widely distributed hemopoietic Ags with similar CD4+ T cells for a GVL effect, the demonstration of substantial and severe GVHR mediated by donor T cells directed against a difference between donor and host of only three amino acids in one protein illustrates the potential power of MiH-specific T cells and indicates the need for additional experiments to learn how to harness that power.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Six- to eight-week-old female C57BL/6 mice that normally express the CD45b allele are denoted B6(CD45b). B6.SJL-Ly5aPep3b mice are congenic to C57BL/6 mice for expression of the CD45a allele and are denoted B6(CD45a). B6(CD45a) mice were obtained from The Jackson Laboratories (Bar Harbor, ME) and maintained in the specific pathogen-free animal care facilities of the University of Washington (Seattle, WA).
Peptides
A 13-amino acid peptide (EPVSKPESASKPH) identical with the polymorphic region (p257–269) of CD45b protein and a 13 mer synthetic peptide (EPASKPDPASKPH) spanning the same polymorphic region (p257–269) of CD45a protein were constructed (differences in sequences are underlined). All peptides were synthesized by Dr. Patrick S. H. Chou (Biopolymer Facility, Department of Immunology, University of Washington) using F-moc chemistry in an automated peptide synthesizer (model 433A, Applied Biosystems, Foster City, CA).
Tumor cell lines
B61710.5 is a CD45b+, MHC class II (Iab)-positive B cell lymphoma of C57BL/6 (H-2b) mouse origin (13), provided by Dr. William R. Green (Department of Microbiology, Dartmouth Medical School, Hanover, NH). FBL-3 is a Friend murine leukemia virus-induced, CD45b+, MHC class II (Iab)-negative leukemia of C57BL/6 (H-2b) mouse origin. LSTRA is a Moloney virus-induced leukemia of BALB/c (H-2d) mouse origin (14).
Immunization protocols
Mice were inoculated by s.c. injection in the hind flanks with CFA (Sigma, St. Louis, MO; 50 µl) emulsified 1/1 with the CD45a or CD45b peptide (100 µg in 50 µl) or with CFA (50 µl) emulsified 1/1 with PBS as the control. Mice were boosted twice at 2-wk intervals by s.c. injection of IFA (Sigma; 50 µl) emulsified 1/1 with the CD45a or CD45b peptide (100 µg in 50 µl) or IFA (50 µl) emulsified 1/1 with PBS as control. Ten to 14 days after the final immunization, spleens of each group of mice were harvested.
Generation and characterization of T cell lines and clones
Peptide-specific CD4+ T cell lines and clones were generated using a protocol described previously (15). In brief, single cell suspensions of the immune spleens were prepared in culture medium consisting of a 1/1 mixture of RPMI 1640 medium (Life Technologies, Grand Island, NY) and EHAA medium (Biofluids, Rockville, MD) with 5 x 10-5 M 2-ME, 100 U/ml penicillin, 100 mg/ml streptomycin, 10 mM L-glutamine, and 10% FCS. The same medium was used in proliferation assays. Lymphocytes were cultured in 24-well culture plates at 4 x 106 cells/well with 2 x 106 irradiated syngeneic spleen cells and CD45a or CD45b peptide at 25 µg/ml (total volume, 2 ml). The plates were cultured in a humidified atmosphere under 5% CO2 tension at 37°C for 5 days, then split 1/2. On day 10, lymphocytes were restimulated at 1 x 106 lymphocytes/well in 24-well culture plates with 5 x 106 irradiated syngeneic spleen cells and peptide (10 µg/ml) and were tested for proliferative activity to the immunizing peptides. Starting with the third in vitro stimulation, lymphocytes were stimulated alternately with irradiated congenic spleen cells or irradiated syngeneic spleen cells plus peptides. Three days after the second in vitro stimulation, T cells were plated at 1 or 0.33 cell/well in 96-well flat-bottom plates with 1 x 106 irradiated syngeneic spleen cells and rIL-2 (100 U/ml). Derived T cell clones were maintained by periodic stimulation with peptide (5 µg/ml) plus irradiated syngeneic spleen cells followed by expansion with rIL-2 (10 U/ml) every 2 wk. T cell clones were stained with fluorescein-conjugated anti-Thy 1.2 (1%), fluorescein-conjugated anti-Lyt2 (2%), or phycoerythrin-conjugated anti-L3T4 (3%) mAbs (Becton Dickinson, Mountain View, CA) at 4°C for 45 min. Fluorescence intensity and positive cell percentages were measured on a flow cytometer (Coulter, Hialeah, FL).
Proliferative assays
To determine the proliferative responses to peptides and syngeneic or congenic lymphocytes, CD45-reactive T cells were cultured in 96-well plates at 2 x 104 cells/well with 5 x 105 irradiated syngeneic spleen cells (3000 rad) and the designated peptide (total volume of 200 µl/well). To determine the proliferative responses to congenic spleen cells or leukemia cells, lymphocytes or cloned T cells were cultured in 96-well plates at 2 x 104 cells/well with 5 x 105 irradiated congenic spleen cells (3,000 rad) or with irradiated congenic leukemia cells (10,000 rad, 5 x 104) in each well of a 24-well culture plate supplemented with complete medium (total volume of 2 ml/well). Plates were incubated in a humidified atmosphere under 5% CO2 at 37°C for 96 h and pulsed for 8 h with 1 mCi of [3H]thymidine/well before harvesting. All determinations were conducted in triplicate wells, and [3H]thymidine incorporation was determined.
Adoptive transfer studies
To study in vivo effects of CD45-specific T cell clones, CD45a anti-CD45b T cell clone 1A4 (1 x 107 cells/recipient) was adoptively transferred via lateral tail vein injection into B6(CD45a) mice or B6(CD45b) mice. Mice from each group were sacrificed on days 1, 2, 3, 8, and 17, and tissues (lung, liver, gut, kidney, spleen, and lymph nodes) were taken for histopathologic analysis.
To study in vivo effects of CD45-specific T cell clones into bone marrow (BM) chimeric mice, B6(CD45a) mice were transplanted with B6(CD45b) BM, and B6(CD45b) mice were transplanted with B6(CD45a) BM. The BM chimeras were prepared as described previously (16). In brief, recipient mice received 14 Gy of total body irradiation (TBI) in a single fraction from dual opposed 60Co sources at an exposure rate of 20 Gy/min. on the day before transplantation. Donor marrow was obtained by femur flush. Chimerism was validated by FACS analysis of peripheral blood leukocytes using Abs specific for CD45a or CD45b on day 63. On day 82 after transplantation, chimeric mice were injected i.v. with CD45-specific T cell clones. Mice from each group were sacrificed on day 1, 3, and 7. Lungs were removed, inflated, fixed in formalin, and sectioned and stained with hemoxylin and eosin. The histology was scored (0 to 4+) according to the extent and severity of the vasculitis by a pathologist (R.C.H.) who was blinded to the experimental treatment groups.
Cytokine production assays
Each CD45a anti-CD45b T cell clone (5 x
105/ml) was stimulated with irradiated congenic spleen
cells (5 x 106/ml) from B6(CD45b) mice in a 24-well
culture plate supplemented with complete medium. After 24-h incubation
at 37°C, culture supernatant of each T cell clone was collected and
stored at -70°C until used for analysis of cytokine production. IL-2
activity of the culture supernatant was quantified by measuring the
proliferative response of an IL-2-dependent cell line, CTLL-2. IL-4 and
IFN- were measured using ELISA kits from Genzyme
(Cambridge, MA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine whether T cell immunity to CD45 alleles can be
elicited, B6(CD45a) mice were immunized with a 13 mer peptide
(EPVSKPESASKPH) identical with the
p257–269 polymorphic region of CD45b protein. CD45
expressed by B6(CD45a) mice and B6(CD45b) mice differ by the three
underlined amino acids. After two in vivo immunizations, splenocytes
were stimulated in vitro with the immunizing peptide. Lymphocytes
derived from immunized B6(CD45a) mice proliferated specifically in
vitro in response to the CD45b peptide in a dose-dependent manner, but
not in response to the CD45a peptide (Fig. 1
A).
|
B).
T cell lines from B6(CD45a) mice specific for CD45b peptides were
cloned by limiting dilution and maintained by episodic restimulation
with irradiated syngeneic spleen cells plus the immunizing CD45b
peptide. Restimulation with syngeneic spleen cells plus peptide was
alternated by restimulation with irradiated congenic spleen cells. All
clones examined were Thy-1.2+ CD4+
CD8- by FACS analysis using fluoresceinated Abs (data not
shown). Proliferative data from eight representative CD4+ T
cell clones confirmed that T cells derived from B6(CD45a) mice
immunized with CD45b peptide responded specifically to the CD45b
peptide, but not to the CD45a peptide (Table I). The stimulation indexes of CD45
peptide-specific T cell clones were generally >300.
|
).
|
Peptide-specific T cells need not necessarily respond to whole
parental protein or to cells that synthesize the whole protein. The
response of CD45 peptide specific T cells to hemopoietic cells
synthesizing CD45 protein requires that CD45 protein be processed so
that the nominated peptide segment survives processing. The response
also requires that the nominated peptide segment be presented by class
II MHC molecules in high enough concentration to stimulate the immune T
cells. In experiments designed to determine whether CD45
peptide-specific T cells can recognize hemopoietic cell populations
synthesizing CD45 protein, B6(CD45a) T cell clones specific for CD45b
peptide were shown to respond vigorously to irradiated B6(CD45b) spleen
cells, but not to B6(CD45a) spleen cells. A representative clone is
presented (Fig. 2A).
The response of the CD45b peptide-specific T cell clones to irradiated
B6(CD45b) spleen cells was dependent upon the responder to stimulator
ratio, with a stimulation index of >300 at responder to stimulator
ratios of 1:6.25 or greater (Fig. 3).
Similarly, specific and vigorous proliferative responses were observed
when T cell clones from B6(CD45b) mice specific for CD45a peptide were
tested in response to spleen cells from B6(CD45a) mice (Fig. 2B).
|
|
To examine the character of GVHD induced by T cells specific for CD45, 1 x 107 T cells (clone 1A4) specific for CD45b were injected i.v. into B6(CD45b) mice. On days 1, 2, 3, 8, and 17, mice were sacrificed and evaluated for histopathologic changes.
The most striking histologic lesions were observed in the lung. On day
1, a mononuclear cell vasculitis involving primarily pulmonary veins
was observed, including diffuse to focally intense margination of
mononuclear cells to endothelium, dense accumulation within and
surrounding the vascular wall, and focal vascular wall necrosis (Fig. 4B). Pulmonary arteries
displayed less mononuclear cell margination and perivasculitis.
Infiltrates were present in areas of alveolar interstitium closely
adjacent to pulmonary veins. Increased numbers of alveolar macrophages
were present within the vicinity of damaged interstitium. Comparative
normal lung is shown in Figure 4, A and E.
|
C) and day 3 (Fig. 4, D,F, and G), with more widespread involvement of
the interstitial and alveolar space. Perivascular inflammation was also
increased. Mononuclear cells remained the predominant inflammatory cell
type in vascular and interstitial locations, with alveolar macrophages
accumulating in alveolar spaces. Neutrophils were occasionally observed
in more severely involved vascular and interstitial sites. There was no
substantial pleural or airway involvement, although peribronchiolar
regions in close proximity to affected pulmonary veins contained
increased numbers of mononuclear cells. By day 8 the intensity of the
vasculitis waned slightly, but interstitial and alveolar mononuclear
inflammatory cells were more prominent. Mice injected with syngeneic
cells were histologically normal (Fig. 4, H andI). A mild increase in perivascular, interstitial, and
alveolar cellularity remained present on day 17 (not shown). The
classical inflammatory and apoptotic manifestations of GVHD were not
observed (data not shown). No mice died in the experiments presented.
However, higher doses of CD45-specific T cells injected alone or with
exogenous IL-2 often induced fatal pulmonary vasculitis (data not
presented). Administration of CD45-specific T cell clones i.v. results in a mild focal hepatic triaditis
In the liver, small foci of parenchymal inflammatory cells were
only occasionally noted on day 1 when clones were injected i.v. On day
8 there was a slight, but definite, increase in the number of
mononuclear cells accumulated in portal triads (Fig. 5). When clones were injected i.p., there
was a more substantial increase in the number of lymphocytes in hepatic
portal triads and around central veins. In addition, there was
widespread, severe mononuclear vascular and perivascular inflammation
in the mesentery (not shown).
|
The above experiments show that T cells specific for CD45 induce a substantial GVHR with a predominant pulmonary vasculitis. CD45 is expressed by hemopoietic cells. If CD45 were expressed exclusively by hemopoietic cells, the GVHR observed could be assumed to be initiated by donor T cells that recognize and are activated by interaction with host hemopoietic cells. However, there is one report of CD45 being expressed in vitro by cultured endothelial cells provided that the cultured endothelial cells are stimulated by IL-1 for several days (17). To confirm that the predominant targets of donor CD45-specific T cells are host hemopoietically derived cells and to confirm the necessity and sufficiency of host-derived hemopoietic cells to initiate the observed GVHR, experiments were performed using BM chimeric mice.
B6(CD45b) mice were transplanted with B6(CD45a) BM following 14 Gy of
total body irradiation. The peripheral blood of the BM chimeric mice
was confirmed to be entirely of donor phenotype by FACS analysis on day
63 following BMT (data not presented). On day 82 following BMT,
B6(CD45b) mice chimeric for B6(CD45a) BM were injected i.v. with 5
x 106 T cells specific for CD45b. No pulmonary
vasculitis or other manifestations of GVHD were observed (Table III). As a positive control, T cells
specific for CD45b induced GVHD when injected into normal B6(CD45b)
mice that had not received BMT and were not chimeric for CD45a BM.
Thus, removal of host CD45b hemopoietic cells by BMT eliminated the
induction of the GVHR by CD45b-specific T cells.
|
). As a negative control, T cells
specific for CD45b did not induce a GVHR when injected into normal
B6(CD45a) mice. Direct cell to cell contact is necessary for response of CD45 specific T cells to spleen cells expressing CD45 protein
The lack of GVHR induced by CD45-specific T cells injected into BM chimeras implies that T cell recognition of hemopoietic cells is sufficient to elicit the observed donor antihost response. However, the possibility exists that CD45 is shed from hemopoietic cells and presented by APC-expressing class II MHC, including pulmonary vascular endothelium. CD45 glycoprotein is abundantly present at the lymphocyte cell surface, but there is no published evidence as to whether it is shed. Our experiments used functional proliferative assays to examine whether CD45 protein is shed and presented as an exogenous Ag by APC.
Initial experiments performed proliferative assays in a dual-chamber
transwell culture system to determine whether CD45b-positive
splenocytes and leukemia cells release CD45b protein in adequate
concentration to stimulate CD45b-reactive T cells. Transwell chambers
were used containing medium, CD45b peptide, irradiated B6(CD45a) spleen
cells, irradiated B6(CD45b) spleen cells, or irradiated FBL-3 leukemia
(a CD45b-positive leukemia cell line). The transwell chambers were
placed into culture wells containing B6(CD45a) T cells specific for
CD45b plus irradiated B6(CD45a) spleen cells as APC. Transwell chambers
containing CD45b peptide induced strong proliferative responses, but no
responses were observed to transwell chambers containing B6(CD45b)
spleen cells or FBL-3 (Fig. 6A). Thus, CD45 peptide
could traverse the membrane between chambers, but there was no evidence
for the presence of soluble CD45b protein traversing the membrane. To
rule out the possibility that CD45 protein is shed, but cannot traverse
the transwell membrane, CD45b-specific T cells were cultured directly
with FBL-3 with B6(CD45a) spleen cells as APC (Fig. 6B). No responses were observed.
|
). CD45b-specific T cells proliferated
in response to L1710.5 leukemia cells (CD45b positive, Iab
positive), even without the addition of splenic APC. By contrast,
no proliferative responses were observed when CD45b-specific T cells
were incubated with FBL-3 leukemia cells (CD45b positive,
Iab negative) alone or with splenic APC (Fig. 6B).
|
To begin to decipher the mechanism(s) of the observed predominant
pulmonary GVHD response, experiments were designed to identify effector
capabilities of the particular CD45 specific T cell clones used. Some
subsets of CD4+ T cells are able to mediate direct
lysis of Ag-positive targets. The necessity of cell to cell contact for
proliferative response to CD45 implies that direct lysis of target
cells expressing CD45 and class II MHC molecules might occur. To
evaluate cytolytic activity, CD4+ T cell clones specific
for CD45b were activated by short term culture with Ag and tested in a
standard 4-h chromium release assay for ability to lyse LPS-activated
B6(CD45b) spleen cells. No direct cytolytic activity of the
CD45+ class II MHC+ targets was detected (Table IV). Similarly, T cells specific for
CD45a did not lyse B6(CD45a) LPS-activated spleen cells. Control
targets included FBL-3 leukemia cells and LSTRA, a Moloney leukemia
virus-induced lymphoma of BALB/c origin.
|
The 1A4 clone used in vivo in the above GVHD experiments was
analyzed for cytokines released upon activation and was shown to
secrete substantial quantities of IFN-, but no detectable IL-2 or
IL-4 (Table V). Most of the clones
derived in a similar fashion have been shown to similarly secrete
IFN-, but also secrete IL-2 (Table VI). The clones
secreting IL-2 have not yet been compared in vivo to the 1A4 clone.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CD45a and CD45b alleles differ by a total of five amino acids. One segment has a three-amino acid difference, and two segments have single amino acid differences (11, 12). CD4+ T cells specific for CD45b were elicited by immunization of B6(CD45a) mice with a 13 mer peptide from the segment of CD45 differing by three amino acids between strains. T cells specific for the 13 mer CD45b peptide proliferated vigorously and secreted cytokines in response to splenic lymphocytes as well as B cell leukemia cells from normal B6 mice. Similar T cells specific for CD45a protein were obtained by immunization of B6(CD45b) mice with synthetic peptides corresponding to the polymorphic region segment of CD45a protein. Thus, that particular segment of CD45 protein is processed and presented by host class II MHC molecules in high enough concentration to stimulate a vigorous response. Of note, the level of proliferation was as great as that normally seen for allogeneic responses, providing strong inferential evidence that CD45 is presented in abundant quantity.
The demonstrated ability to generate T cells specific for CD45 provided the opportunity to address whether T cells directed against Ags not expressed by skin, liver, and gut induce the classic manifestations of GVHD of skin, liver, and gut toxicity. CD45-specific T cells induced a GVHR, but the predominance of pulmonary vasculitis observed was unexpected. Pulmonary manifestations of GVHD occur, but usually in association with more typical manifestations of GVHD involving damage to skin, liver, and gut.
Histologic changes in humans with acute GVHD and pulmonary pathology include diffuse alveolar damage, interstitial pneumonitis, lymphocytic bronchitis/bronchiolitis, and bronchiolitis obliterans organizing pneumonia (22). However, all of these are nonspecific patterns of injury and inflammation (23). The histologic changes sometimes include vascular alterations, but isolated vasculitis, like that in the current experiments, has not been reported. Changes in pulmonary function and histology are commonly observed following human BMT. However, patients are usually prepared with high dose chemotherapy and/or TBI, which can result in toxic damage that confounds the analysis of possible GVHD. In one rodent study using the transfer of parental T cells into allogeneic F1 rats without TBI, interstitial pneumonitis and lymphocytic bronchiolitis/bronchitis were interpreted to be direct results of acute lethal GVHD. Of note, the histopathology observed duplicated that of lung allograft rejection, including the presence of early vasculitis (24).
The analysis of pulmonary GVHD in humans is complicated not only by toxicity alterations from chemotherapy and radiation, but also by other factors, including the absence of a "gold standard" for GVHD, the frequent presence of infection, and the difficulty of obtaining lung tissue contemporaneous with the onset of pulmonary symptoms. Thus, it is not clear whether vasculitis is a common manifestation of early acute GVHD following human BMT. If acute pulmonary GVHD develops in humans, vasculitis may be the initial histologic alteration that precedes other suggested changes, such as lymphocytic bronchitis and interstitial pneumonitis. In this regard, the demonstrated association between acute GVHD and terminal pulmonary hemorrhage (25) indicates that injury to the vascular endothelium may be a critical early event.
There are also chronic pulmonary syndromes associated with GVHD after allogeneic BMT (26). Chronic syndromes are associated with clinical bronchiolitis and often occur with obliterative bronchiolitis and patchy interstitial pneumonitis (26, 27). In patients without obvious bronchiolitis, chronic airflow obstruction is commonly observed in long term survivors of allogeneic BMT (28) and is a significant complication, even in children (29). The development of chronic pulmonary syndromes is usually associated with signs and symptoms of acute and chronic GVHD involving other organs and is thus most likely a manifestation of GVHD. We have not yet determined whether mice surviving the acute insult of vasculitis induce by CD45-specific T cells ultimately develop long term manifestations of pulmonary GVHD such as obliterative bronchiolitis and airway obstruction.
Drawing the conclusion that pulmonary vasculitis is induced by donor T cell recognition of CD45 expressed by host hemopoietic cells requires that CD45 be expressed only by hemopoietic cells. The conclusion also requires that CD45 not be shed and presented by nonhemopoietic APC, such as vascular endothelium. CD45 has long been considered to be expressed only by hemopoietic cells. However, there is one report that CD45 is expressed by cultured endothelial cells (17). The RNA message for the CD45 was demonstrated by PCR to be constitutively present in cultured endothelial cells and to increase substantially with IL-1 stimulation. Endothelial cells are not known to constitutively express CD45 in vivo. Whether and the extent to which CD45 is expressed by activated endothelial cells in vivo are not known.
The current experiments used CD45 chimeric mice to validate that host hemopoietic cells need to be present for induction of the observed pulmonary GVHR and mitigate against the possibility that endothelial cell CD45 plays a substantial role. Transfer of CD45b-specific T cells into B6(CD45b) mice chimeric for CD45a bone marrow failed to induce vasculitis, implying that following BMT, no CD45b-positive tissue was present and/or any CD45b expressed by endothelial cells was inadequate to stimulate donor T cell responses. In the experiments presented, the CD45b-specific T cells were injected into bone marrow chimeras on day 83 post-BMT. By FACS examination of peripheral blood, there were no host hemopoietic cells remaining.
It is known that cytokines secreted by stimulated T cells in the vicinity of endothelial cells can up-regulate class II MHC and adhesion molecules on endothelium and thus potentially render endothelium capable of functioning as APC for soluble Ags (30, 31). However, there is no evidence that CD45 is present in vivo as a soluble molecule. Our in vitro studies showed that CD45 is not present as a soluble molecule in any appreciable amount despite its abundance as a cell surface molecule. In those experiments CD45 peptide specific T cells responded vigorously to CD45-positive spleen cells, but not detectably to supernatants from CD45-positive spleen cells or to class II MHC-negative leukemia cells expressing CD45. One implication is that cell to cell contact is necessary for the response of CD45 peptide-specific T cells to spleen cells expressing CD45 protein. We are not aware of other studies that have otherwise assayed for shed CD45. The other implication is that CD45 is processed endogenously in the class II MHC-processing pathway, as has been shown for other Ags (32).
One of the original goals of the current experiments was to determine whether T cells specific for polymorphic hemopoietic differentiation Ags can mediate a GVL effect without GVHD toxicity. Following allogeneic BMT, leukemia recurs less commonly in patients who develop GVHD. Donor T cell populations that mediate GVHD can also mediate a GVL effect that is commonly curative (33, 34). The observation that donor T cells can eradicate host leukemia has been substantiated in studies treating human leukemia relapse by infusions of donor T cells (35). In that circumstance the recipient hemopoietic system, including BM, peripheral blood, and lymph nodes, is derived from donor hemopoietic cells, whereas the leukemia cells as well as nonhemopoietic tissue are host derived. Infusions of donor T cells can result in leukemia regression and cure, but usually also result in substantial GVHD toxicity. If the preponderance of GVHD damage is caused by T cells specific for polymorphic host proteins with a wide tissue distribution, T cells specific for polymorphic proteins expressed exclusively by host hemopoietic cells might result in a GVL effect without substantial GVHD. However, the current demonstration that CD45-specific T cells induce substantial pulmonary damage mitigates against the use of such cells in leukemia therapy.
Following BMT, host hemopoietic cells are not eliminated immediately, but, rather, decrease in number over time. In experiments not presented, CD45-specific T cells were injected into bone marrow chimeras on day 21 at a time when some host CD45-positive hemopoietic cells were still present. When CD45-specific T cells were injected into mice with only small numbers of residual CD45+ hemopoietic cells, some GVHD vasculitis was observed. Thus, responses of CD4+ T cells to even small numbers of residual leukemia cells might cause substantial toxicity.
The observed induction of a pulmonary GVHR even in mice containing only small numbers of residual host cells indicates that T cell clones similar to the particular CD45-specific CD4+ T cell clone used in the current in vivo experiments might not be appropriate for mediating GVL effects. However, other CD4+ T cell populations may mediate a more specific GVL effect. The currently used T cell clone secreted Th1 cytokines and was not cytolytic. In experiments not presented we showed that the lung injury induced by the CD45-specific Th1 cells used is characterized by host-derived mononuclear cell inflammation and activation of alveolar macrophages. Labeled cloned T cells were localized in inflammation foci in the lung, but the majority of cells in the foci were not labeled. Moreover, mature host T cells and B cells were not required, since lung injury was comparable in transgenic RAG-1 knockout host mice that lack these cells.6 It is possible that clones with the same Ag specificity but with different functions might yield markedly different results. T cell responses can be polarized toward synthesis of Th1 or Th2 cytokines. Others have speculated that T cells secreting Th2 cytokines, either by Th or Tc, might have a dampening effect on GVHD (36, 37, 38). It is conceivable that clones secreting Th2 cytokines and capable of directly lysing residual host hemopoietic cells would be able to mediate a GVL effect without substantial GVHD. An alternative ploy for using T cells specific for hemopoietic differentiation Ags to treat residual host leukemia might best focus on the use of CD8+ CTL, especially CD8+ CTL that secrete Th2-type cytokines.
The currently described GVHR system is unique in that a substantial response is mediated by T cells against a short segment of a known and defined Ag. The system will allow testing of T cell clones and lines with the same Ag specificity, but with different cytokine profiles. Such experiments may allow separate characterization of effector mechanisms that predispose to GVL as opposed to GVHD effects.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Center for Surgery Research, The Cleveland Clinic Foundation, Cleveland, OH 44195.
3 Current address: Department of Medicine, University of Arkansas Medical Center, Little Rock, AR 72205.
4 Address correspondence and reprint requests to Dr. Martin A. Cheever, Corixa Corp., 1124 Columbia St., Suite 200, Seattle, WA 98104.
5 Abbreviations used in this paper: GVHD, graft-vs-host disease; BMT, bone marrow transplantation; MiH, minor histocompatibility Ags; GVHR, graft-vs-host response; GVL, graft-vs-leukemia; BM, bone marrow; TBI, total body irradiation.
6 J. G. Clark, D. K. Madtes, R. C. Hackman, W. Chen, M. A. Cheever, and P. J. Martin. Alloreactive Th1 cell transfer induces pulmonary inflammation and alveolar macrophage activation in mice. Submitted for publication.
Received for publication December 22, 1997. Accepted for publication March 24, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K.-H. Kim, K. Burkhart, P. Chen, C. W. Frevert, J. Randolph-Habecker, R. C. Hackman, P. D. Soloway, and D. K. Madtes Tissue Inhibitor of Metalloproteinase-1 Deficiency Amplifies Acute Lung Injury in Bleomycin-Exposed Mice Am. J. Respir. Cell Mol. Biol., September 1, 2005; 33(3): 271 - 279. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Clark, J. B. Mandac-Dy, A. E. Dixon, D. K. Madtes, K. M. Burkhart, J. M. Harlan, and D. C. Bullard Trafficking of Th1 Cells to Lung: A Role for Selectins and a P-Selectin Glycoprotein-1-Independent Ligand Am. J. Respir. Cell Mol. Biol., February 1, 2004; 30(2): 220 - 227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Thatte, V. Dabak, M. B. Williams, T. J. Braciale, and K. Ley LFA-1 is required for retention of effector CD8 T cells in mouse lungs Blood, June 15, 2003; 101(12): 4916 - 4922. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Dixon, J. B. Mandac, P. J. Martin, R. C. Hackman, D. K. Madtes, and J. G. Clark Adherence of adoptively transferred alloreactive Th1 cells in lung: partial dependence on LFA-1 and ICAM-1 Am J Physiol Lung Cell Mol Physiol, September 1, 2000; 279(3): L583 - L591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Dixon, J. B. Mandac, D. K. Madtes, P. J. Martin, and J. G. Clark Chemokine expression in Th1 cell-induced lung injury: prominence of IFN-gamma -inducible chemokines Am J Physiol Lung Cell Mol Physiol, September 1, 2000; 279(3): L592 - L599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Dixon, J. B. Mandac, P. J. Martin, D. K. Madtes, R. C. Hackman, and J. G. Clark Alloreactive Th1 Cells Localize in Lung and Induce Acute Lung Injury Chest, July 1, 1999; 116 (2009): 36S - 37S. [Full Text]
|
|
|
|
 |
|
 P. A. Eden, G. J. Christianson, P. Fontaine, P. J. Wettstein, C. Perreault, and D. C. Roopenian Biochemical and Immunogenetic Analysis of an Immunodominant Peptide (B6dom1) Encoded by the Classical H7 Minor Histocompatibility Locus J. Immunol., April 15, 1999; 162(8): 4502 - 4510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Clark, D. K. Madtes, R. C. Hackman, W. Chen, M. A. Cheever, and P. J. Martin Lung Injury Induced by Alloreactive Th1 Cells Is Characterized by Host-Derived Mononuclear Cell Inflammation and Activation of Alveolar Macrophages J. Immunol., August 15, 1998; 161(4): 1913 - 1920. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
