| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of
*
Surgery,
Medicine, and
Pathology, Roswell Park Cancer Institute, Buffalo, NY 14263; and
§
Division of Hematology/Oncology and Comprehensive Cancer Center, Ohio State University, Columbus, OH 43210
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-herpesvirus that is carried as an asymptomatic and persistent
infection within pharyngeal epithelium and mature B lymphocytes by
>90% of adults worldwide. While EBV-infected B lymphocytes have the
potential for transformation and uncontrolled proliferation in vivo,
this is thought to be normally prevented by virus-specific CTL
surveillance (1, 2, 3). However, the identification of
cellular components that confer this protection in vivo remains
uncertain. The suppression or elimination of this protection is thought
to result in the development of an EBV-associated lymphoproliferative
disorder (4, 5). EBV-associated lymphoproliferative disorder occurs with high frequency in certain congenital, acquired, and iatrogenic immunodeficient states, including the profound immunosuppression following solid organ or bone marrow transplantation (BMT).4 The use of T cell-depleted donor bone marrow, or the administration of intensive anti-T cell therapy in the setting of severe graft-vs-host disease (GVHD), escalates the risk of developing posttransplant lymphoproliferative disorder (PTLD) following allogeneic BMT (4). The spectrum of PTLD ranges from polyclonal B cell hyperplasia to monoclonal immunoblastic lymphoma (5, 6). Although polyclonal PTLD has been shown to regress following withdrawal of immune suppressive therapy (7), monoclonal disease demonstrates intrinsic resistance to conventional therapy and usually runs a fatal course (8). Recent success with the delivery of in vitro-generated CTLs to patients with PTLD suggests that restoration of host immunity may be the most promising therapeutic strategy in the control of these "opportunistic" malignancies (43).
The in vitro inhibition of EBV-induced B cell transformation in the presence of autologous T lymphocytes (1) and the demonstration of CTL-mediated reversal of EBV+ lymphoblastoid cell line outgrowth in xenografted SCID mice (3) also strongly support the notion that T cells are critical in the control of EBV infection. Likewise, the clinical observation that PTLD regresses following relaxation of immunosuppressive therapy is consistent with this. However, as recently reviewed by Rickinson et al. (11), direct evidence that endogenous EBV-specific CTLs can mediate regression of PTLD in these patients is still lacking.
In the current report, we provide a detailed characterization of the emergence of a donor-derived, monoclonal PTLD following allogeneic BMT and document its regression following the endogenous expansion of donor-derived CD3+CD8+ CTL in blood.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A 31-yr-old Japanese man was diagnosed with Philadelphia chromosome positive, t(9;22), chronic myelogenous leukemia. He underwent allogeneic BMT with a conditioning regimen that consisted of etoposide (1.8 g/m2 over 2 days), cyclophosphamide (60 mg/kg/day for 3 days), and total body irradiation (10 G in five fractions). He received an infusion of bone marrow (2.62 x 108 cells/kg) from an HLA-matched unrelated donor (donor and recipient HLA typing: A2, B46, DR8, DQ1, DQB1-06, DQB1-08) that was not T cell depleted. This treatment was performed after obtaining informed consent, approved by the Institutional Review Board at Roswell Park Cancer Institute.
Prophylaxis against GVHD was initiated 48 h prior to bone marrow infusion (i.e., day -2) with cyclosporine A (CsA, 5 mg/kg/day). One dose of methotrexate (10 mg/m2) was administered on day +1. Solumedrol was initiated on day +3 at a dose of 1 mg/kg/day and was increased to 9 mg/kg/day on day +21 following the development of skin GVHD. Colonoscopy was performed on day +29 to evaluate high volume diarrhea and showed GVHD. OKT3 mAb was therefore administered at 1 mg/kg/day. Severe GVHD persisted, so two consecutive 8-day courses of 2-chlorodeoxyadenosine (2-CDA; 0.02 mg/kg/day) were administered. By day +56 the severity of the GVHD had decreased.
On day +62 flow cytometric analysis of blood showed a phenotype
consistent with B cell excess (see Fig. 1
). The patient received
IFN-
(2 x 106 U/m2/day for 35 days),
i.v. Ig (IVIg, 25 g/day for 3 days), and acyclovir (ACV, 645 mg/day for
17 days). In addition, the immunosuppressive regimen was reduced (CsA,
0.6 mg/kg/day; solumedrol, 1.5 mg/kg/day). Over the ensuing weeks,
sequential phenotypic and molecular analyses of blood lymphocytes were
performed as described below. Severe GVHD affecting skin,
gastrointestinal tract, and liver subsequently reoccurred, and the
patient expired from fulminant liver failure on day +113. Permission
for full autopsy was obtained and confirmed histological evidence of
severe GVHD in these organs. There was no evidence of malignancy in any
tissues examined histologically.
|
Two-color phenotypic analysis of blood using directly conjugated mAbs was performed as described (12) using a lymphocyte gate. Routine methods for DNA isolation, digestion, electrophoresis, Southern blotting, and 32P radiolabeling of probes were followed (9). DNA probes for the JH region of the IgH gene (10), for EBV DNA (13), and for the constant region of the TCR ß-chain gene (14) have been described.
Identification of B cell and T cell gene rearrangement by PCR
To identify the specific IgH V-D-J gene rearrangement associated with the B cell lymphocytosis, DNA was isolated from PBMC on day +75. The sequences of the seven IgH-V family-specific 5' primers and the four IgH-J generic 3' primers, along with PCR amplification conditions, have been described (15). Following amplification, one distinct band appeared which was excised and sequenced. DNA from day +65 and day +68 PBMC were also amplified and sequenced; they were found to be identical. Subsequently, IgH-V5 (5'-TGCGCCAGATGCCCGGGAAAG-3') and IgH-J4 (5'-GAGGAGACGGTGACCAGGGTTCCCTGG-3') primers were used to detect this IgH gene rearrangement in DNA extracted from serial blood and tissue samples.
Analysis of the TCR Vß gene repertoire was undertaken at indicated
time points and from indicated tissues. RNA was isolated using RNAzol
(Tel-Test, Freindswood, TX) and reversed transcribed into first-strand
cDNA following the manufacturer’s recommendations (Clontech, Palo
Alto, CA). Five microliters of the cDNA was equally distributed to 26
tubes, and PCR amplification was performed with 25 different TCR Vß
5' primers, each paired with a single Cß-specific 3' primer as
directed (Clontech). An identical aliquot of the cDNA was used for
amplification of the TCR gene constant region using C 5' and C
3' primers as an internal control to verify the integrity of the
cDNA (data not shown). Water was used as negative control.
Identification of EBV latent membrane protein-1 (LMP-1)
RNA isolation and reverse transcription (RT) were performed as above. Thirty-five cycles of PCR amplification (94°C for 1 min, 58°C for 1 min, and 72°C for 1 min) was next performed with primers specific for LMP-1 mRNA (5'-CCTCCGCACCCTCAACAAG-3'; 5'-GAGATGATGACGACCCCCA-3'). ß-Actin was amplified from cDNA samples with commercially available primers (Perkin-Elmer, Norwalk, CT).
Microsatellite PCR analysis
The details of performing PCR amplification of the microsatellite locus 220yh4 for purposes of distinguishing donor versus recipient cells have been described (16).
DNA sequencing
PCR products were first purified using the Qiaquick PCR purification kit (Qiagen, Chatsworth, CA) and sequenced on an Applied Biosystems model 373 Stretch DNA sequencing system (Perkin-Elmer). Sequences were analyzed using GenBank databases (IntelliGenetics, Campbell, CA).
Cytotoxicity assays
Target PBMC (day +62) and effector PBMC (day +113) were collected from the patient and immediately frozen viably without further sensitization in vitro. On the evening prior to the cytotoxicity assay, cells were thawed, washed, resuspended in RPMI 1640 medium supplemented with 10% FCS, and cultured separately overnight for equilibration. A total of 50 pM of IL-2 (10 U/ml) was added to day +113 PBMC. Viability was >80%. Day +113 effector cells were then washed, plated with W6/32 mAb (anti-HLA-A, -B, -C shared determinant; Accurate Chemicals, Westbury, NY) and anti-CD8 mAb (Becton Dickinson, Mountain View, CA), or nonreactive isotype control mAb (mouse IgG2a, negative control; Dako, Carpinteria, CA) all at 1:200 dilution. 51Cr-labeled day +62 PBMC target cells were then added (E:T = 20:1), incubated for 4 h, and assessed for specific cell lysis as described previously (17).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The temporal course of the B cell lymphoproliferation and the
endogenous, autologous cellular immune response that followed were
assessed with serial flow cytometric analyses over a 3-mo period (Fig. 1
, A and B).
Sixty-two days following BMT, the first changes indicating a clonal B
cell expansion were noted, with 86.7% of lymphocytes expressing CD19
(Fig. 1C). Ninety-six percent of CD19+ B
lymphocytes expressed surface Ig with 
light chain (LC),
indicating a clonal population of B lymphocytes. This same clonal
population expressed CD20 and CD23, consistent with a B cell
lymphoblastic phenotype commonly associated with EBV transformation
(Fig. 1C) (18). The monoclonal population of B
lymphocytes was confirmed by Southern blot analysis and also was shown
to contain the EBV genome (Fig. 2,
A and B). PCR amplification of microsatellite DNA
isolated from FACS-sorted subsets of lymphocytes from day +62 showed
the monoclonal B cell population was of donor origin (Fig. 2C). RT-PCR analysis of RNA extracted from blood drawn the
same day revealed an EBV-LMP-1 gene product that was present but
somewhat smaller than the B95.8 EBV strain LMP-1. Sequence analysis
demonstrated numerous point mutations and deletions between the
coordinates 168,611 and 168,225, virtually identical to the LMP-1
transcript variant described in the C15 nasopharyngeal carcinoma
isolate (data not shown) (19).
|
, ACV, and IVIg were administered without effect. 2-CDA,
administered earlier as an immunosuppressive agent, has no known
activity against EBV-associated lymphoma (20). On day +81,
18 days after the reduction of immunosuppressive therapy, an expansion
of CD3+CD8+ T cells was noted, with a
concomitant decrease in CD19+ B cells (Figs. 1A
and 3A). The CD4:CD8 ratio was
1:13, and nearly all T cells showed an activated phenotype with HLA-DR
expression (Fig. 3A). Without further pharmacologic or
immunologic manipulations, by day +97 the
CD3+CD8+ T cell surged to 84.4% of peripheral
blood lymphocytes and the clonal CD19+ population was
reduced to 14% (Fig. 1A). Absolute numbers of
CD19+ and CD3+ lymphocytes began to decline in
parallel at this time (Fig. 1B). Blood from day +97 was
analyzed for TCR gene rearrangement by Southern blot analysis and
showed no evidence of clonal rearrangement (data not shown).
Microsatellite analysis of DNA isolated from sorted T lymphocytes
showed complete donor origin of T lymphocytes (Fig. 2D). By
day +103 the clonal CD19+ B cell population had decreased
to 4%, whereas the CD3+CD8+ T cells persisted
at 78%, still with activation Ag (HLA-DR) expression (Figs. 1A and 3B). Ten days later (day +113), the
patient succumbed to fulminant liver failure secondary to GVHD.
Phenotypic analysis of blood at that time revealed 66% of the
lymphocytes to be CD3+CD8+ donor T cells, with
only 3% CD19+ B cells (Figs. 1, A and
B, and 3C), with the majority of CD8+
T cells still expressing HLA-DR.
|
PCR evidence for the IgH gene rearrangement associated with the
donor-derived malignant B cell clone was found only after BMT and
showed a decline in blood over time that was coincident with the
CD3+CD8+ autologous T cell surge (Fig. 4A). Persistent evidence of
the B cell clone could be found at autopsy by PCR in the spleen, bone
marrow, and lymph node, but not in the liver where GVHD had been most
severe (Fig. 4A). Despite molecular evidence documenting the
persistence of this clonal B cell population in the blood and certain
tissues at autopsy, the LMP-1 transcript, associated with EBV-driven B
cell proliferation (21, 22) was no longer detected in
these cells near or at the time of death (Fig. 4B). This was
despite a RT-PCR sensitivity for LMP-1 that was 10-fold greater than
that for the B cell clone.
|
We next characterized the TCR Vß gene repertoire in blood over
time. As all cells following BMT were of donor origin (Fig. 2, C and D), the TCR Vß profile of the donor T
cells before BMT were also analyzed. This profile showed a typical
polyclonal pattern (Fig. 5A).
At day 64, 2 days following the detection of a clonal population of B
lymphoblasts, the TCR Vß repertoire was unchanged with the exception
that TCR Vß1 became undetectable and Vß3 and -23 became apparent
for the first time (Fig. 5B). Twenty-three days after the
detection of circulating clonal EBV+ B cells (i.e., day
+83), the TCR Vß repertoire was unchanged with the exception of Vß1
again being detected (Fig. 5C). By day +97, PTLD was
diminishing (Fig. 1B), LMP-1 transcript became undetectable
in blood (Fig. 4B), and the percent
CD3+CD8+ CTL were at their peak (Fig. 1A). At this time there was a switch from a polyclonal to a
restricted pattern of TCR Vß usage, with a predominance of Vß1, -2,
-3, -4, -6, -8, -9, and -23 in blood (Fig. 5D).
|
E). Cytotoxic activity of ex vivo CTL against EBV+ B lymphoblasts
To determine if the
CD3+CD8+HLA-DR+ CTL present in
blood during the resolution of PTLD were indeed MHC class I-restricted
cytolytic effectors, a standard 51Cr-release cytotoxicity
assay was performed using PBMC from day +62 (86% CD19, CD20, CD23,
surface Ig LC, and EBV+) as targets (T) and PBMC from
day +113 (66% CD3+CD8+HLA-DR+) as
effectors (E). All cells were donor derived (Fig. 2, C and
D). No additional in vitro sensitization was performed prior
to the 4-h assay. Cytotoxicity at an E:T of 20:1 in the presence of
isotype control Abs was 76.3 ± 4.2%. In the presence of Abs
reactive against MHC class I and CD8, cytotoxicity was 6.3 ±
1.3% (Fig. 6). Thus, donor-derived CTL
that expanded in peripheral blood during the simultaneous reduction of
donor-derived CD19+, monoclonal, EBV+ B
lymphoblasts displayed MHC class I-restricted cytotoxicity against this
population.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B) despite the
persistence of severe clinical GVHD until death would, in addition to
the other data, also argue against this. The nearly complete abrogation
of cytotoxic activity in the presence of anti-MHC class I mAb is
consistent with earlier reports showing the same inhibition of
CTL-mediated lysis of EBV+ LCLs in vitro (32),
and argues against NK cells as effectors. This is because NK cell
cytotoxic activity against autologous EBV+ LCLs increases
in the presence of MHC class I blockade (33). The Vß-restricted population of CTL persisted while the circulating B cell population was reduced to <7% of lymphocytes, and the LMP-1 mRNA transcript became impossible to detect in blood, spleen, liver, or lymph node despite a highly sensitive RT-PCR assay. LMP-1 transcripts are believed to be expressed only in proliferating EBV+ B cell immunoblasts that are highly immunogenic and accordingly sensitive to EBV-specific CTL recognition in vivo (34). Furthermore, the LMP-1 gene product has been shown to possess immunodominant epitopes utilized by EBV-specific CD3+CD8+ T lymphocytes in HLA A2 individuals, similar to the patient in this study (35). In an immunocompetent host or an immunologically recovering bone marrow transplant recipient, LMP-1 expressing B cell immunoblasts should be effectively eradicated (36). Thus, the absence of LMP-1 transcript in tissues that showed evidence of the B cell clone by DNA PCR suggests that the patient in this study was likely to have maintained a nonimmunogenic, clonal population of EBV+ B cells expressing a latent gene profile restricted to EBV-encoded nuclear Ag 1 (EBNA1) and LMP-2A (latency type 1) following the T cell response. EBV has been shown to establish long-term infection and latency type 1 gene expression in mature, resting memory B lymphocytes of normal, asymptomatic carriers (37). Indeed, cellular immune-mediated mechanisms resulting in the transition of activated, proliferating EBV+ B lymphoblasts to a quiescent, resting state have been described in vivo (34, 37, 38, 39, 40). Thus, DNA PCR evidence documenting the presence of the B lymphocyte clone in tissues obtained at autopsy support recent findings that describe the plasticity of latent gene usage by EBV in maintenance of long-term persistence within the memory B cell compartment.
The liver was the only organ examined that demonstrated elimination of both the EBV-LMP-1 transcript and the B cell clone transformed by the virus. The liver was also the site of fatal GVHD, known to be caused by alloreactive T cells (41). The TCR Vß repertoire detected within the liver parenchyma at the time of fatal GVHD showed some overlap with that which predominated during the CTL expansion in vivo and which demonstrated MHC class I-restricted cytotoxic activity against EBV+ B cell lymphoblasts in vitro. Thus, it is possible that the donor CTL responsible for the elimination of autologous PTLD could have also contributed to the fatal hepatic GVHD that occurred in the allogeneic tissues. However, we were unable to provide any proof of this in this analysis. If true, it would support earlier in vitro studies demonstrating dual specificity of CTL for autologous EBV-transformed cell lines or a single EBV epitope and HLA alloantigens (42). Further, it would underscore the importance of eliminating alloreactive CTL from autologous EBV-specific CTL generated ex vivo prior to their infusions in vivo for the treatment of PTLD (43).
Several aspects of this report provide new information with regard to the in vivo control of PTLD. We provide direct evidence that withdrawal of immune suppression in a patient with GVHD following allogeneic BMT resulted in the emergence of endogenous donor-derived CD3+CD8+HLA-DR+ T cells that demonstrated MHC class I-restricted cytolytic activity against the monoclonal donor-derived B cell PTLD ex vivo. The endogenous emergence of this oligoclonal population of CTL was associated with the simultaneous regression of the monoclonal PTLD in vivo. This report therefore provides additional evidence in support of the generally accepted notion that autologous EBV-specific CTL are responsible for the normal immune surveillance against PTLD.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 V.P.K. and R.A.B. contributed equally to this
work.
3 Address correspondence and reprint requests to Dr. Michael A. Caligiuri, Ohio State University, 458A Starling Loving Hal, 320 West 10th Avenue, Columbus, OH 43210. E-mail address:
4 Abbreviations used in this paper: BMT, bone marrow transplantation; ACV, acyclovir; CsA, cyclosporine; LC, light chain; PTLD, posttransplant lymphoproliferative disease; GVHD, graft-vs-host disease; 2-CDA, 2-chlorodeoxyadenosine; LCL, lymphoblastoid cell line; IVIg, i.v. Ig; LMP, latent membrane protein.
Received for publication September 24, 1997. Accepted for publication April 13, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. F. May Jr, S. Roychowdhury, D. Bhatt, E. Kocak, X.-F. Bai, J.-Q. Liu, A. K. Ferketich, E. W. Martin Jr, M. A. Caligiuri, P. Zheng, et al. Anti-human CTLA-4 monoclonal antibody promotes T-cell expansion and immunity in a hu-PBL-SCID model: a new method for preclinical screening of costimulatory monoclonal antibodies Blood, February 1, 2005; 105(3): 1114 - 1120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kuzushima, N. Hayashi, A. Kudoh, Y. Akatsuka, K. Tsujimura, Y. Morishima, and T. Tsurumi Tetramer-assisted identification and characterization of epitopes recognized by HLA A*2402-restricted Epstein-Barr virus-specific CD8+ T cells Blood, February 15, 2003; 101(4): 1460 - 1468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Porcu, C. F. Eisenbeis, R. P. Pelletier, E. A. Davies, R. A. Baiocchi, S. Roychowdhury, S. Vourganti, G. J. Nuovo, W. L. Marsh, A. K. Ferketich, et al. Successful treatment of posttransplantation lymphoproliferative disorder (PTLD) following renal allografting is associated with sustained CD8+ T-cell restoration Blood, September 18, 2002; 100(7): 2341 - 2348. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Quaia, P. Zancai, R. Cariati, S. Rizzo, M. Boiocchi, and R. Dolcetti Glucocorticoids promote the proliferation and antagonize the retinoic acid-mediated growth suppression of Epstein-Barr virus-immortalized B lymphocytes Blood, July 15, 2000; 96(2): 711 - 718. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
,
CD3-CD56-; 
,
CD19+CD20+CD3-) and T cells (
,
CD3+CD8+) in the blood of the patient
throughout the clinical course. Relevant therapeutic events occurring
during the clinical course are indicated below the horizontal axis.
B, Temporal course of absolute number of lymphocyte subsets
(
, total CD3+) in
peripheral blood (cells/mm3). C, Histograms of
lymphocyte subsets on day +62 show that 86% of gated lymphocytes are
CD19+, whereas 7% express CD3 (left
panel). B lymphocytes are predominantly
CD20+CD23+, consistent with a B lymphoblast
phenotype (middle panel). Ninety-six percent of
CD19+ B lymphocytes express surface Ig with 
1% (16).
