• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Biol Blood Marrow Transplant. Author manuscript; available in PMC Oct 5, 2007.
Published in final edited form as:
PMCID: PMC2000975
NIHMSID: NIHMS25244

High Content Flow Cytometry and Temporal Data Analysis for Defining a Cellular Signature of Graft versus Host Disease

Abstract

Acute graft versus host disease (GvHD) is diagnosed by clinical and histologic criteria that are often non-specific and typically apparent only after the disease is well established. Since GvHD is mediated by donor T-cells and other immune effector cells, we sought to determine whether changes within a wide array of peripheral blood lymphocyte populations could predict the development of GvHD. Peripheral blood samples from 31 patients undergoing allogeneic blood and marrow transplant were analyzed for the proportion of 121 different subpopulations defined by 4-color combinations of lymphocyte phenotypic and activation markers at progressive time points post-transplant. Samples were processed using a newly developed high content flow cytometry technique and subjected to a spline- and FLDA-based temporal analysis technique. This strategy identified a consistent post-transplant increase in the proportion and extent of fluctuation of CD3+CD4+CD8β+ cells in patients who developed GvHD compared to those that did not. While larger prospective clinical studies will be necessary to validate these results, this study demonstrates that high content flow cytometry coupled with temporal analysis is a powerful approach for developing new diagnostic tools and may be useful for developing a sensitive and specific predictive test for GvHD.

Keywords: graft versus host disease, high content flow cytometry, bioinformatics

INTRODUCTION

Acute graft-versus-host disease (GvHD) occurs in allogeneic hematopoietic stem cell transplant (SCT) recipients when donor immune cells in the graft initiate an attack on the skin, gut, liver and other tissues of the recipient (1-6). The pathophysiology of GvHD is currently felt to occur through several phases (2, 4, 6). In the first phase, damage by the chemotherapy or radiotherapy used in the transplant preparatory regimen causes host tissues to secrete inflammatory cytokines. This results in activation of alloreactive donor T-cells that recognize HLA and minor histocompatability antigen disparities on host cells. Subsequently, the donor T cells and other immune effectors elaborate a variety of inflammatory cytokines including TNF-α, IFN-γ, IL-13, IL-5 and others resulting in the widespread tissue damage observed clinically. A variety of other immune cells, including dendritic cells, B-cells, natural killer cells and macrophages may also play important roles in GvHD through additional mechanisms (7-11).

It is likely that the outcome of GvHD could be improved if it were treated as early as possible in a pre-emptive fashion, before the full-blown clinical syndrome develops. In addition, if the diagnosis of GvHD could be made more definitively, only those patients who absolutely required steroids and other immunosuppressive medications would be treated. Currently however, there are no definitive methods for detecting GvHD during its preclinical stage or to unequivocally distinguish GvHD from infections or drug toxicities unless there are advanced histologic features on biopsy samples. For these reasons, it is important to develop rapid, reliable and accurate tests for the prediction and diagnosis of GvHD.

The development of microarray technologies has ushered in a new era of diagnostic tests (12). Attempts have been made to use microarrays to identify gene expression patterns in peripheral blood leukocytes that would be diagnostic of GvHD (13, 14). However, gene microarray data provides averaged gene expression information from peripheral blood leukocyte populations that are actually quite heterogeneous in nature. Consequently, the averaging effects of microarray analysis may miss important variations in expression of individual genes within different subsets of cells. In addition, contributions from small populations of immune cells that may be important to the development of GvHD could be missed altogether.

Flow cytometry (FCM) offers a potential alternative to gene microarray analysis for rapidly defining complex changes in heterogeneous populations (15-17). Multiple dyes, lasers and detectors can be used to simultaneously collect multiple fluorescence emission signatures from cell populations representing as little as 0.1% of the total sample (18). Thousands of fluorescently conjugated antibodies and fluorescent dyes are now commercially available to provide a wide array of cellular measurements including cell phenotype, intracellular cytokine expression, cell cycle status and recently, signal transduction pathway activation. Recently, we and others have developed high throughput or high content FCM systems that can analyze thousands of individual samples per day to provide rich data sets on various types of cells (19, 20).

Managing the large amounts of data generated by high-throughput/high content flow cytometry system is a significant challenge. When analyzing parameters such as changes in immune effector cell populations over time that could contribute to GvHD, the temporal aspect of the data introduces an additional level of complexity as well. GvHD can occur several days to several months after allogeneic transplant and data analysis needs to account for this variation within a study population. Recently, several algorithms treating time as a continuous variable have been introduced to deal with this issue, including spline-based methods such as linear, cubic and B-splines (21-29). Splines are mathematical representations of smooth curves that pass through two or more points. Using splines to model time series data can avoid problems such as data overfitting, and the methodology is well-suited for the analysis of small numbers of data points (22, 30).

Based on these developments, the potential now exists for combining the advantages of high content flow cytometry with the power of modern bioinformatics to determine if there are patterns of cells in the peripheral blood that correlate with a variety of physiologic or disease states including GvHD. Using the high content FCM and the spline-based analysis methods outlined above in a pilot study of 31 patients undergoing allogeneic transplantation, we analyzed whether increases or decreases in the proportion of one or more of 121 different peripheral blood leukocyte populations predicted the subsequent development of acute GvHD. We found that, of these populations, an increase in the proportion of CD3+CD4+CD8β+ cells 7-21 days post transplant best correlated with the subsequent development of acute GvHD. These findings suggest that this population should be studied further for its possible biologic role in GvHD and that larger prospective clinical studies should be conducted to validate its predictive accuracy. In addition, this work demonstrates that high content FCM and spline-based analysis is a promising approach to developing diagnostic tools for GvHD and other processes.

MATERIALS AND METHODS

Study Patients

Thirty-one patients undergoing HLA matched sibling and unrelated donor allogeneic blood and marrow transplantation were enrolled at the Moffitt Cancer Center. Of the 31 patients enrolled in this study, 21 patients were diagnosed with acute GvHD while 3 patients did not develop either acute or chronic graft versus host disease and thus were used as controls. The remaining patients were not included in the analysis either because they i) died prior to 100 days post transplant and it could not be determined if they would have subsequently developed GvHD or not (n=2), ii) developed de-novo chronic graft versus host disease which may have confounded the analysis for acute GvHD (n=3), iii) were lost to follow-up (n=1) or iv) had insufficient clinical samples collected (n=1). The details of the patient demographics, stem cell source, transplant related treatments and acute GvHD are summarized in Table 1. The diagnosis and grading of GvHD was made using previously published criteria (31).

Table 1
Summary of patient information

Isolation of peripheral blood mononuclear cells

Samples of peripheral blood were collected using IRB approved processes into EDTA containing tubes pre-transplant and then weekly for at least 100 days post transplant. Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll-Hypaque and then were cryopreserved for subsequent batch analysis. At the time of analysis, cells were thawed and aliquoted into 96 well plates at 1 ×104 to 1×105 cells per well in IMDM/10%FCS. The 96 well plates were then stained and analyzed as described below.

Flow cytometric high content screening

The 96 well plates were stained with 10 different 4-color antibody combinations as described in Table 2. All antibodies were obtained from BD Biosciences except anti-CD45-PE and anti-CD8ß-PE, which were from Immunotech. All staining and flow cytometric analysis was done using a flow cytometric high content screening (FC-HCS) technique described previously (19). Briefly, in the FC-HCS technique, all staining and analysis procedures were miniaturized so that small numbers of cells could be stained in 96 well plates with optimally diluted fluorescently conjugated antibodies. All staining procedures were performed by a Biome® 2000 robotic fluid handler (Beckman Instruments, Schaumburg, IL) using a series of mini-programs developed with BioWorks™ software (Beckman Instruments). Cell washings were performed either by centrifugation or by vacuum aspiration through a filter bottom plate (Silent Screen Plate, Nunc™, Rochester NY), using a vacuum manifold (96-Filtration System-Beckman Coulter) controlled by the Bioworks software. Flow cytometric analysis was performed on a FACSCalibur device equipped with a 488 nm argon laser and a ~635 nm red dye laser (Becton Dickinson (BD), San Jose, CA). Samples of cells were delivered directly from 96 well plates to the FACSCalibur using prototypes of a Microtiter Well Plate device (BD). The BD Multiwell Plate Manager/Multiwell Autosampler (MPM/MAS) software (BD) was used for the collection and annotation of data. A typical example of the primary flow cytometric data collected in this fashion is depicted in Figure 1.

Figure 1
Representative FC-HCS analysis of PBMCs from patients with GvHD
Table 2
Markers used in high content flow cytometric analysis

Statistical analysis

The strategy for analysis of the FC-HCS data is summarized in Figure 2. Batch analysis of the FCS files was performed using FlowJo Software (TreeStar, Palo Alto, Ca) to determine the proportion of each gated population within peripheral blood mononuclear cells. A total of 121 distinct cell populations were defined in the 10 four-color staining panels using gates on two-dimensional contour plots as depicted in the representative example in Figure 1. This data was then exported as text files to generate tabular data. Data from samples taken between 7 to 21 days post-transplant were selected to develop the predictive test since we sought to develop a test that would be useful 1-2 weeks prior to the onset of GvHD. Patient data was separated into two groups representing either affected (developed acute GvHD) or unaffected (developed neither acute nor chronic GvHD) patients. Splines were fit to the time series proportion data for each cell type for each patient. Functional Linear Discriminant Analysis (FLDA) was then employed to generate an average curve for each group, based on a signal plus noise model as previously described (32). FLDA was performed in MATLAB (Natick, MA) based on linear basis B-splines with three vertices (i.e., one for each week). All the primary flow cytometric data from measurements with high estimated sensitivity and specificity were inspected visually to confirm the spline based data analysis strategy. Lastly, the leave-one out validation (LOOV) technique was used to estimate the error of the FLDA classifier.

Figure 2
Summary of the data analysis pipeline

RESULTS

To determine if a cellular signature of acute GvHD could be defined in the weeks prior to the onset of the full-blown clinical syndrome, PBMCs from weekly blood samples were collected starting immediately after allogeneic transplant through ~day +100. These samples were stained using 10 panels of 4-color antibody combinations (Table 2) and flow cytometry data was collected using the FC-HCS technique. Gating was performed on different populations within each stained sample to generate a total of 121 populations for subsequent statistical analysis. An example of a representative 4-color, 6-parameter staining combination and gated subpopulations is depicted in Figure 1. The data was then analyzed as outlined in Figure 2 for any changes in cell populations occurring between 7 and 21 days post-transplant that correlated with the development of acute GvHD, which in this series of patients occurred on average 35 days (± 17 days) post-transplant.

Of the 121 populations analyzed in this fashion, a population of lymphocytes that co-stained with antibodies directed against CD3, CD4 and CD8ß (CD3+CD4+CD8β+ cells) was found to have the highest correlation with the development of acute GvHD (Figure 3A and 3B and data not shown). Patients who did not develop acute GvHD had little variation in the proportion of CD3+CD4+CD8β+ cells in the 7-21 days following transplant, while acute GvHD patients had a higher and more variable proportion of CD3+CD4+CD8β+ cells within this time period (Figure 3A and 3B). To confirm whether these findings from the spline-based analysis and FLDA reflected actual changes in the flow cytometric analysis of the patient's peripheral blood lymphocytes rather than an anomaly arising from the analysis strategy, the original flow data for the CD3+CD4+CD8ß+ cells was inspected manually for each patient at each time point. The gating strategy for defining the CD3+CD4+CD8ß+ population detected in the spline analysis is presented in Figure 4A. The entire CD3+CD4+CD8ß+ population in many patients consisted of 2-3 distinct subpopulations defined by varying levels of staining with antibodies directed against CD4, CD8β and CD8 (Figure 4B). While these populations could not be clearly distinguished in all patients at all time points, a subpopulation which co-expressed high levels of both CD8ß and CD8 (CD3+CD4+CD8ßbrCD8+ cells) could typically be detected. As depicted in Figure 5 (and data not shown), a visually distinct difference was noted for the proportion of CD3+CD4+CD8ß+ and CD3+CD4+CD8ßbrCD8+ cells between the patients who developed GvHD and those who did not, supporting the temporal data analysis strategy. A leave-one out classifier built on these observations correctly predicted the absence of GvHD with 100% specificity and 86% sensitivity when based on the entire CD3+CD4+CD8ß+ population (Table 3). Most of the predictive power of this population resided in the CD3+CD4+CD8ßbrCD8+ cells as opposed to other populations seen in this four color staining panel that did not co-stain with one or more of the antibodies to CD4, CD8 or CD8ß (Figure 4B and Table 3).

Figure 3
The pattern of CD3+CD4+CD8β+ cells following allogeneic transplant differentiates between patients with and without GvHD
Figure 4
Representative gating strategy for CD3+CD4+CD8β+ cells and co-expression of CD4, CD8β and CD8 on these cells in a patient with GvHD
Figure 5
Representative flow cytometric analysis of CD3+CD4+CD8β+ cells in patients who developed GvHD compared to patients who did not develop GvHD
Table 3
Leave one out validation (LOOV) of the predictive power of CD3+CD4+CD8β+ and other populations with the development acute GvHD.

DISCUSSION

In the current study, we hypothesized that a sensitive and specific flow cytometric cellular signature for acute GvHD could be developed based on analyzing a wide array of peripheral blood cell populations for phenotypic and activation markers using high content flow cytometry and temporal based statistical tools. We found that the FC-HCS technique could be readily applied to analyzing a large clinical sample set and that it efficiently yielded a robust data set for subsequent analysis. As we have reported previously, the FC-HCS technique was able to process 500-1000 samples per day with excellent flow cytometric staining and analysis profiles (e.g., Figure 1) (19). While other high throughput flow cytometry approaches are more efficient, FC-HCS is well suited for detecting small populations of cells in different samples using multiparameter analysis since there is almost no carryover between individual samples that could cause false positive results (data not shown) (20). Since the development of acute GvHD occurred at different time points, it was important to use temporal analysis tools (e.g. spline- and FLDA based methods) to look for statistically meaningful differences in the various leukocyte populations between patients diagnosed with GvHD and the controls and to build a predictive model. This approach may also be suitable for other projects where clinical outcomes occur at variable time points.

In this preliminary study, the FC-HCS technique coupled with the temporal analysis determined that elevations in the proportion of CD3+CD4+CD8β+ cells within the first 7-21 days post-transplant predicted the development of GvHD with the highest sensitivity and specificity of any of the tested populations. A subset of these cells that also co-stained with an anti-CD8 antibody (i.e. CD3+CD4+CD8ßbrCD8+ cells) also predicted the development of GvHD with a high sensitivity and specificity whereas other populations which lacked co-expression of either CD4,CD8 or CD8β had relatively low predictive value (Table 3). To conclusively determine whether the CD3+CD4+CD8β+ or CD3+CD4+CD8ßbr CD8+ cell value is a reliable predictor of GvHD, a larger prospective clinical study will need to be conducted which is sufficiently powered to yield statistically stronger results than obtained in this initial pilot study. It will be important in this future study to control for reactivation of CMV and development of other infectious diseases as well as additional clinical events to ensure that the flow cytometric signature is specific to GvHD. In addition, other antibody panels should be incorporated to determine if cell populations that were not looked for in this initial pilot study, including T-regulatory cells and others, have greater sensitivity and specificity than the CD3+CD4+CD8β+ population identified in this study (33, 34). Ultimately, if the results presented in this initial report or more predictive populations identified in future studies are validated, a simple weekly blood test could be constructed where multi-color staining and flow cytometric analysis are performed to predict the development of GvHD. This could potentially allow for pre-emptive treatment of GvHD, similar to the pre-emptive strategies used for CMV reactivation following allogeneic transplantation (35). Subsequent clinical trials could then test whether pre-emptive treatment with steroids or other immunosuppressive agents could improve the outcome compared to traditional treatments that are initiated only after the clinical syndrome is established. Other uses of this test could include determining how much immunosuppressive treatment should be initiated, when treatment failure has occurred and as a guide for tapering immunosuppressive therapy. A similar approach may also be applicable to developing a test for chronic GvHD or other transplant related outcomes.

If the CD3+CD4+CD8β+ or CD3+CD4+CD8ßbrCD8+ cell populations identified in this report are validated as accurate and reliable predictors of GvHD in future studies, it will also be important to further clarify their biology and relevance to the clinical manifestations of GvHD. This cell type appears to be a T-cell subset that co-expresses CD4, CD8αβ heterodimers (detected by the 2ST8.5H7 antibody, Table 2) and CD8αα homodimers (detected by the SKI antibody, Table 1). One type of CD4+CD8+ (double positive) T-cell is found in the thymus as an intermediate stage of T-cell development. These can be found at low levels in the peripheral blood of healthy individuals and at higher levels during viral infection and other conditions where they are thought to represent premature egress of immature thymic T-cells (36-39). However, the CD3+CD4+CD8β+ CD8+ population detected in patients after allogeneic transplant in our study appears to express lower levels of CD4 than typical intra-thymic double positive T-cells, and may represent a different cell type altogether (40). This possibility is supported by reports of double positive T-cells in the lamina propria of rhesus macaques and in mice following epicutaneous immunization that have a variety of phenotypic and functional properties that suggest they are mature T-cells rather than premature release of immature thymic T-cells (41, 42). Similarly, in humans, CD4dimCD8+ T-cells have been observed in healthy blood donors, HIV infected persons and kidney transplant recipients that also appear distinct from intrathymic immature double positive T-cells (43). CD4+CD8+ T-cells have been found in the blood of both HCV infected persons and normal controls that have properties consistent with differentiated effector memory cells as well (36). Together, these observations suggest that the double positive T-cells we have observed in the peripheral blood of persons following allogeneic transplant may be mature antigen specific cells. It will be particularly interesting in future studies to characterize the CD3+CD4+CD8β+ CD8+ population for a variety of properties including activation and differentiation status, expression of tissue and lymph node homing markers, the pattern of intracellular cytokine secretion, TCR Vß repertoire diversity and antigen responsiveness as previously described in virally infected persons (36). In addition, if a similar population were observed in murine models of allogeneic transplant, more direct studies of its relevance to GvHD could be performed (44).

ACKNOWLEDGEMENTS

We thank Simon Dablemont for his assistance in performing FLDA. JL was supported by the Canadian Institutes of Health Research/MSFHR Bioinformatics Training Program. CS is the recipient of a Canadian Research Chair, Michael Smith Scholar grants and BC Transplant Society grants.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

1. Chao NJ. Graft-versus-host disease: the viewpoint from the donor T cell. Biology of Blood and Marrow Transplantation. 1997;3:1–10. [PubMed]
2. Couriel D, Caldera H, Champlin R, Komanduri K. Acute graft-versus-host disease: pathophysiology, clinical manifestations, and management. Cancer. 2004;101:1936–1946. [PubMed]
3. Gilliam AC. Update on Graft versus Host Disease. Progress in Dermatology. 2004;123:251–257. [PubMed]
4. Goker H, Haznedaroglu IC, Chao NJ. Acute graft-vs-host disease: pathobiology and management. Experimental Hematology. 2001;29:259–277. [PubMed]
5. Martin PJ, Schoch G, Fisher L, et al. A retrospective analysis of therapy for acute graft-versus-host disease: initial treatment. Blood. 1990;76:1464–1472. [PubMed]
6. Reddy P. Pathophysiology of acute graft-versus-host disease. Hematological Oncology. 2003;21:149–161. [PubMed]
7. Vakkila J, Thomson AW, Hovi L, Vettenranta K, Saarinen-Pihkala UM. Circulating dendritic cell subset levels after allogeneic stem cell transplantation in children correlate with time post transplant and severity of acute graft-versus-host disease. Bone Marrow Transplant. 2005;35:501–507. [PubMed]
8. Chan GW, Gorgun G, Miller KB, Foss FM. Persistence of host dendritic cells after transplantation is associated with graft-versus-host disease. Biol Blood Marrow Transplant. 2003;9:170–176. [PubMed]
9. Haraguchi K, Takahashi T, Matsumoto A, et al. Host-residual invariant NK T cells attenuate graft-versus-host immunity. J Immunol. 2005;175:1320–1328. [PubMed]
10. Nakamura A, Kobayashi E, Takai T. Exacerbated graft-versus-host disease in Pirb−/−mice. Nat Immunol. 2004;5:623–629. [PubMed]
11. Lo JW, Leung AY, Huang XR, et al. Macrophage migratory inhibitory factor (MIF) expression in acute graft-versus-host disease (GVHD) in allogeneic hemopoietic stem cell transplant recipients. Bone Marrow Transplant. 2002;30:375–380. [PubMed]
12. Margalit O, Somech R, Amariglio N, Rechavi G. Microarray-based gene expression profiling of hematologic malignancies: basic concepts and clinical applications. Blood Rev. 2005;19:223–234. [PubMed]
13. Akilesh S, Shaffer DJ, Roopenian D. Customized molecular phenotyping by quantitative gene expression and pattern recognition analysis. Genome Res. 2003;13:1719–1727. [PMC free article] [PubMed]
14. Ichiba T, Teshima T, Kuick R, et al. Early changes in gene expression profiles of hepatic GVHD uncovered by oligonucleotide microarrays. Blood. 2003;102:763–771. [PubMed]
15. Srour EF, Yoder MC. Flow cytometric analysis of hematopoietic development. Methods Mol Med. 2005;105:65–80. [PubMed]
16. Dunphy CH. Applications of flow cytometry and immunohistochemistry to diagnostic hematopathology. Arch Pathol Lab Med. 2004;128:1004–1022. [PubMed]
17. Goolsby C. A look at where we came from and where we are going. Cytometry. 2004;58A:53–54. [PubMed]
18. Owen RG, Rawstron AC. Minimal residual disease monitoring in multiple myeloma: flow cytometry is the method of choice. British Journal of Haematology. 2005;128:732–733. [PubMed]
19. Gasparetto M, Gentry T, Sebti S, et al. Identification of compounds that enhance the anti-lymphoma activity of rituximab using flow cytometric high-content screening. J ImmunolMethods. 2004;292:59–71. [PubMed]
20. Edwards BS, Oprea T, Prossnitz ER, Sklar LA. Flow cytometry for high-throughput, high-content screening. Curr Opin Chem Biol. 2004;8:392–398. [PubMed]
21. Bar-Joseph Z, Gerber G, Simon I, Gifford DK, Jaakkola TS. Comparing the continuous representation of time-series expression profiles to identify differentially expressed genes. Proceedings of the National Academy of Sciences of the United States of America. 2003;100:10146–10151. [PMC free article] [PubMed]
22. Bar-Joseph Z, Gerber GK, Gifford DK, Jaakkola TS, Simon I. Continuous representations of time-series gene expression data. Journal of computational biology : a journal of computational molecular cell biology. 2003;10:341–356. [PubMed]
23. De Hoon MJ, Imoto S, Miyano S. Statistical analysis of a small set of time-ordered gene expression data using linear splines. Bioinformatics (Oxford, England) 2002;18:1477–1485. [PubMed]
24. Irizarry RA, Tankersley C, Frank R, Flanders S. Assessing homeostasis through circadian patterns. Biometrics. 2001;57:1228–1237. [PubMed]
25. James G, Hastie T. Functional Linear Discriminant Analysis for Irregularly Sampled Curves. Journal of the Royal Statistical Society: Series B. 2001;63:533–550.
26. James G, Sugar C. Clustering for Sparsely Sampled Functional Data. 2003;98:397–408.
27. Luan Y, Li H. Clustering of time-course gene expression data using a mixed-effects model with B-splines. Bioinformatics (Oxford, England) 2003;19:474–482. [PubMed]
28. Rice JA, Wu CO. Nonparametric mixed effects models for unequally sampled noisy curves. Biometrics. 2001;57:253–259. [PubMed]
29. Storey JD, Xiao W, Leek JT, Tompkins RG, Davis RW. Significance analysis of time course microarray experiments. Proceedings of the National Academy of Sciences of the United States of America. 2005;102:12837–12842. [PMC free article] [PubMed]
30. Friedman JH, Silverman BW. Flexible parsimonious smoothing and additive modeling (with discussion) Technometrics. 1989;31:3–39.
31. Gratwohl A, Hermans J, Apperley J, et al. Acute graft-versus-host disease: grade and outcome in patients with chronic myelogenous leukemia. Working Party Chronic Leukemia of the European Group for Blood and Marrow Transplantation. Blood. 1995;86:813–818. [PubMed]
32. James GM, Hastie TJ. Functional linear discriminant analysis for irregularly sampled curves. Journal of the Royal Statistical Society Series B. 2001;63:533–550.
33. Cohen JL, Boyer O. The role of CD4+CD25hi regulatory T cells in the physiopathogeny of graft-versus-host disease. Curr Opin Immunol. 2006;18:580–585. [PubMed]
34. Nguyen VH, Zeiser R, Negrin RS. Role of naturally arising regulatory T cells in hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2006;12:995–1009. [PubMed]
35. Boeckh M, Fries B, Nichols WG. Recent advances in the prevention of CMV infection and disease after hematopoietic stem cell transplantation. Pediatr Transplant. 2004;8(Suppl 5):19–27. [PubMed]
36. Nascimbeni M, Shin EC, Chiriboga L, Kleiner DE, Rehermann B. Peripheral CD4(+)CD8(+) T cells are differentiated effector memory cells with antiviral functions. Blood. 2004;104:478–486. [PubMed]
37. Ortolani C, Forti E, Radin E, Cibin R, Cossarizza A. Cytofluorimetric identification of two populations of double positive (CD4+,CD8+) T lymphocytes in human peripheral blood. Biochem Biophys Res Commun. 1993;191:601–609. [PubMed]
38. Tonutti E, Sala P, Feruglio C, Yin Z, Colombatti A. Phenotypic heterogeneity of persistent expansions of CD4+ CD8+ T cells. 1994;73:312–320. [PubMed]
39. Weiss L, Roux A, Garcia S, et al. Persistent expansion, in a human immunodeficiency virus-infected person, of V beta-restricted CD4+CD8+ T lymphocytes that express cytotoxicity-associated molecules and are committed to produce interferon-gamma and tumor necrosis factor-alpha. J Infect Dis. 1998;178:1158–1162. [PubMed]
40. Jimenez E, Sacedon R, Vicente A, Hernandez-Lopez C, Zapata AG, Varas A. Rat peripheral CD4+CD8+ T lymphocytes are partially immunocompetent thymus-derived cells that undergo post-thymic maturation to become functionally mature CD4+ T lymphocytes. J Immunol. 2002;168:5005–5013. [PubMed]
41. Pahar B, Lackner AA, Veazey RS. Intestinal double-positive CD4+CD8+ T cells are highly activated memory cells with an increased capacity to produce cytokines. Eur J Immunol. 2006;36:583–592. [PubMed]
42. Szczepanik M, Bryniarski K, Tutaj M, et al. Epicutaneous immunization induces alphabeta T-cell receptor CD4 CD8 double-positive non-specific suppressor T cells that inhibit contact sensitivity via transforming growth factor-beta. Immunology. 2005;115:42–54. [PMC free article] [PubMed]
43. Lambert C, Iobagiu C, Genin C. Persistent oligoclonal CD4dimCD8+T cells in peripheral blood. Cytometry B Clin Cytom. 2005;66:10–17. [PubMed]
44. Chen BJ, Deoliveira D, Cui X, et al. Inability of memory T cells to induce graft-versus-host disease is a result of an abortive alloresponse. Blood. 2006 [PMC free article] [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...