PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Cytometry A. Author manuscript; available in PMC 2010 November 1.
Published in final edited form as:
PMCID: PMC2821788
NIHMSID: NIHMS162406

Differentiation of Alloreactive versus CD3/CD28 Stimulated T-lymphocytes Using Raman Spectroscopy: a greater specificity for noninvasive acute renal allograft rejection detection2

Abstract

Acute rejection (AR) remains problematic in renal transplantation. As a marker, serum creatinine is limited, warranting a more effective screening tool. Raman spectroscopy (RS) can detect T-cell activation with high sensitivity. In this study we explore its specificity. Seventy-five inactivated, 40 alloantigen-activated, and 75 CD3/CD28-activated T-cells were analyzed using RS. CD3/CD28-activated peak magnitudes (PM) were 4.3% to 23.9% lower than Inactivated PM at positions: 903cm−1, 1031cm−1, 1069cm−1, 1093cm−1, 1155cm−1, 1326cm−1, and 1449cm−1, with a difference in peak ratio (PR) observed at the 1182:1195cm−1 position (0.91 ± 0.06 vs. 1.2 ± 0.01, respectively: p = 0.006). Differences in CD3/CD28- and alloantigen-activated PM were observed at: 903cm−1, 1031cm−1, 1093cm−1, 1155cm−1, 1326cm−1, and 1449cm−1, with no PR differences at the 1182:1195cm−1 position (0.91 ± 0.06 vs. 0.86 ± 0.09: p = 0.8). Spectral signature separation of CD3/CD28 - and alloantigen-activated groups was 100% specific and sensitive. We conclude that RS can differentiate T-cells activated by different stimuli with high sensitivity and specificity.

Keywords: Raman Spectroscopy, Renal Transplantation, Acute Allograft Rejection, Human T-Cell Activation and Detection, Cell Surface Receptors, CD3/CD28 Stimulation

Introduction

Renal transplantation has become the preferred treatment for the vast majority of patients with end-stage renal disease (1,2). However, its applicability to all in need has been limited by the continued shortage of available organs (3). Given the scarcity of donor kidneys, it is imperative that the functionality and survival of each of these grafts be maximized in the recipient. Acute rejection (AR), which is predominantly T-cell mediated, has continued to negatively impact the outcome of renal allografts (46). Diagnosis of AR following transplantation based on serum creatinine elevation (SCE) following transplantation has proven problematic. Despite the widespread use of SCE as a marker by most centers, it represents a late finding that becomes apparent only after significant histologic damage to the transplanted organ has already occurred. Moreover, SCE has a low specificity, reflecting the fact that other conditions such as urinary tract infection (UTI), dehydration, obstruction, and even immunosuppressive medications can cause false positives leading to unwarranted referrals for costly and potentially dangerous transplant biopsies (7), which may also be subject to sampling error. In addition to being a late non-specific marker, SCE has a low sensitivity, not taking into account subclinical AR (SCAR) which occurs with no change in serum creatinine (8). SCAR has prompted a greater implementation of protocol renal transplant biopsies in response to concerns over the accelerated development of chronic allograft nephropathy if the condition remains undetected and untreated (910).

A recent surge in proposed methodologies for the noninvasive diagnosis of AR has included Raman spectroscopy (RS). RS is a laser-based technology that utilizes a monochromatic incident photon to induce detectable molecular vibrations in a given material. These vibrations yield a unique “signature”, which we and others have shown in prior studies to be highly accurate in the differentiation of activated from non-activated T lymphocytes based upon molecular differences in cell surface receptors (11,12). Despite the establishment of RS as a methodology to identify T cell activation states, the capability of the system to distinguish signatures of T cells that are activated by different stimuli still requires examination. Therefore, the purpose of this study is to compare spectral signatures of alloantigen-activated and CD3/CD28 activated T cells. We hypothesize that the RS signatures of T cells activated via these two methodologies will differ significantly.

Materials and Methods

Alloantigen-activated T Cell Preparation

Prior approval for the study was obtained from the Wayne State University Human Investigation Committee. Mononuclear cells were obtained using sodium-heparinized venous blood collected from healthy participants and separated via density gradient as described by Boyum (13). The resultant cells were washed with Hank’s balanced salt solution (HBSS, Invitrogen, Carlsbad, CA) and suspended in complete medium. Cellular viability was >98% as checked by vital dye exclusion.

Three distinct T cell sample groups were created for the alloantigen-activated T cells: 1. activated; 2. inactivated; and 3. resting T lymphocytes. The activated T lymphocyte samples were created via a two-way mixed lymphocyte culture (MLC) as described by Dupont and colleagues (14), which brings into proximity T lymphocytes from non-related individuals, antigen presenting cells (APC; monocytes and macrophages), and the necessary components to model allograft rejection. T cell activation was confirmed by CD69 directed flow cytometry as described elsewhere (15,16) The inactivated group was created by utilizing T lymphocyte/stimulator samples from two non-related individuals which were pretreated for 20 minutes with Mitomycin C (0.025 mg Mitomycin C [0.5 mg/ml] added to each 1 ml of cell suspension) at 37 degrees Celsius. These inactivated T lymphocyte samples were washed twice with HBSS prior to the start of MLCs. Efficacy of Mitomycin C was verified using double inactivation control cultures which were run in parallel. Finally, the resting T lymphocyte group consisted of T cells that were neither stimulated via MLC nor inhibited by Mitomycin C treatment. These cells were the product of CD69 negative sorting via flow cytometry (15,16). Following MLC but prior to RS analysis, B cells were removed by negative selection using FluoroBeads-B immunomagnetic beads and magnetic sorter (One Lambda, Canoga Park, CA) while monocytes/macrophages were removed via steel wool columns to ensure that these signatures were not obtained erroneously. All cultures were incubated at 37 degrees Celsius in 5% CO2 in parallel over a 7-day period, with RS, antibody, and viability analysis carried out daily.

Non-specific Antigen Model of Activation

T lymphocytes were isolated directly from peripheral blood and urine using a Dynal T cell negative isolation kit (Invitrogen, Canoga Park, CA). In order to model a nonspecific process of activation, a CD3/CD28-coated Dynal bead system (Invitrogen, Canoga Park, CA) was utilized to cross-link cell membrane molecules. This consisted of T lymphocytes incubated with beads (bead volume titrated to correspond to 3 beads for every T cell at a concentration of 1 × 107 cells per volume) in complete media containing recombinant interleukin 2 (50U/ml). The CD3/CD28 beads were removed from T cells via magnetic sorter prior to Raman analysis.

Independent Verification of Activation/Inactivation

Independent of RS analysis, the activation status of all samples was verified by antibody staining and flow cytometry. For the alloreactive and CD3/CD28 stimulated samples, activation status was assessed using a monoclonal antibody, CD69 (FastImmune, Becton Dickinson, San Jose, CA), as described elsewhere (17). Stained cells were viewed using a Nikon Eclipse TE 2000-U inverted microscope, with images captured and processed with Metavue 6.2r5 software (Downingtown, PA).

Raman Spectroscopy

Prior to collection of Raman spectral data, all samples were washed with PBS (Invitrogen, Carlsbad, CA). T cells were allowed to settle within an analysis reservoir placed on a leveled microscope stage. Only non-apoptotic, rounded, non-adhered, isolated and individualized T cells were selected for study. Each sample was analyzed using identical Raman spectroscopic parameters. In order to assure that T cells were aligned in the laser target area and that the laser focus point was centered on the lymphocyte, an enhanced video imaging system was employed. Images were captured before and after each measurement to confirm that no cell shifting or destruction occurred. To provide for reference/control spectral peaks, both pure CD3/CD28 beads and T cells that were not subjected to magnetic sorter (T cells still bound with beads), were analyzed. Wire 2 software (Renishaw plc, Old Town, United Kingdom) was utilized in conjunction with the RS system for quantitative measurement.

RS measurements were conducted using a Renishaw InVia 2000 microscope-spectrometer with a Leica 63 X water immersion objective (NA = 1.20). T lymphocytes were analyzed using the 514.5 nm (green) Ar+ excitation wavelength. Laser power was set at 50% (8–12 mW) for a 2–4 µm laser spot size. Spectra were collected in the backscattering geometry with a 10 second integration time over a range of 500 to 1500 cm−1 with spectral resolution of 4 cm−1.

Statistical Analysis

Prior to statistical analysis, all data were checked for various types of noise. Spectral data demonstrated uniformly increasing fluorescence over the entire range of Raman shift. This was subtracted out using a modified cubic-spline algorithm requiring no a priori knowledge of the spectra. In addition, a median filter was applied to the raw data which eliminated cosmic ray and spikes. Principal component analysis was used to reduce dimensionality in the data. Chi square or Student’s T test was utilized as appropriate to further quantify differences in peak ratios and peak magnitudes with additional quantitative differences determined by discriminant function analysis (DFA). Graphical and quantitative analysis of data were accomplished using SPSS version 15 software (Statistical Software, Chicago, IL).

Results

Figure 1 summarizes Raman spectra for inactivated, resting, alloreactive, and CD3/CD28-activated T cell groups.

Figure 1
Mean Raman spectra of alloreactive, inactivated, resting, and CD3/CD28-activated T lymphocytes at the 514.5nm wavelength.

CD3/CD28-activated versus Inactivated T Lymphocytes

A total of 75 CD3/CD28-activated and 75 inactivated T cells were analyzed using the 514.5 nm excitation wavelength. Systematic review of samples revealed no cellular disruption following laser exposure. Qualitative comparisons of activated and inactivated spectra demonstrated differences at 903 cm−1, 1031 cm−1, 1069 cm−1, 1093 cm−1, 1155 cm−1, 1326 cm−1, and 1449 cm−1 Raman shifts (figure 2). Quantitative analysis of these shifts demonstrated differences in peak magnitudes at all positions except for 628cm−1, 788 cm−1, and 1002 cm−1 (Table I, figure 3A-E), which have been shown in previous studies to correlate with cellular viability (18). Moreover, the CD3/CD28-activated and inactivated groups showed significant differences in peak ratios at the 1182:1195 cm−1 position (Table I, figure 4).

Figure 2
Superimposed mean Raman spectra from inactivated (red) and CD3/CD28-activated T lymphocytes annotated with (*) foci of significant differences and (^) markers of cellular viability (18).
Figure 3
Raman spectra of inactivated (red) and CD3/CD28-activated T lymphocytes demonstrating differences in peak magnitudes (insets A-E). Annotations within insets: Inset A- (*) 903 cm−1; Inset B- (^) 788 cm−1; Inset C- (*) 1002 cm−1 ...
Figure 4
Raman spectra of inactivated (red) and CD3/CD28-activated T lymphocytes demonstrating differences in (*) 1182 cm−1 and (^) 1195 cm−1 peaks.
Table I
Comparative summary of relevant Raman shift of CD3/CD28-activated, inactivated, alloreactive, and resting T lymphocytes.

CD3/CD28-activated versus Alloantigen-activated T Lymphocytes

A total of 46 CD3/CD28-activated and 40 alloantigen-activated T cells were analyzed with summary comparison spectra shown in figure 5. When comparing these groups, they differed significantly at the 903 cm−1, 1031 cm−1, 1093 cm−1, 1155 cm−1, 1326 cm−1, and 1449 cm−1 positions. However, there was no difference when analyzing the peak ratio at the 1182:1195 cm−1 position (Table I). Discriminant function analysis of the Raman spectral data for CD3/CD28-activated and alloantigen-activated T cell groups is summarized in figure 6. There was 100% accuracy when using the activation signature to separate these groups.

Figure 5
Superimposed mean Raman spectra from CD3/CD28-activated (red) and alloreactive (blue) T lymphocytes annotated with (*) foci of significant differences and (^) markers of cellular viability (18).
Figure 6
Discriminate function analysis of alloreactive versus CD3/28-activated T lymphocytes.

Discussion

The immune system reflects a duality of requirements dictated by the need to remain relatively conserved in many of its activation pathways, while maintaining a high level of specificity for a particular antigen. This notion has been thoroughly demonstrated through studies involving the major histocompatibility complex, the T cell receptor, and other cell surface receptors involved in the activation process (19). Our data suggest that when examining the array of receptors expressed during activation by two different methodologies using the RS system, this duality of conservation and specificity is articulated via unique peak differences. When analyzing the ratios of the 1182:1195 peaks, we found similar patterns in both the alloreactive and the CD3/CD28-activated samples that were not present in the inactivated or resting T cell samples. This Raman shift likely represents a conserved change in cell surface molecules shared by activated T cells regardless of methodology of activation. When applying established peak assignment data (2023), these peaks are representative of tyrosine/phenylalanine and adenine/thymine changes. These conserved peak changes can be contrasted with those shifts occurring at 903 cm−1, 1031 cm−1, 1093 cm−1, 1155 cm−1, 1326 cm−1, and 1449 cm−1 which were observed in the CD3/CD28-activated but not in the alloantigen-activated T cell samples. These foci of shifts likely represent a specific change in cell surface molecules reflecting a response to the particular CD3/CD28 stimulus. When cross referencing these foci with established peak assignment data (2023), they represent changes in nucleic acids (903 cm−1,1093 cm−1, and 1449 cm−1), amino acids (1031 cm−1, and 1155 cm−1), or both (1326 cm−1)- changes which are consistent with molecular processes responsible for the transcription, translation, and expression of cell surface receptors. When further analyzing these peak differences that were unique to the CD3/CD28-activated T cells, there was a reduction in peak magnitudes observed in all six of the aforementioned peak positions. Although the primary etiology of this pattern remains cryptic, it most likely represents a reshuffling rather than a down-regulation of cell surface bio-molecular material.

Based upon the aforementioned observations, it is possible to detect T cell activation and characterize the particular mode of this activation using RS. This represents a foundational step toward the development of a RS-based system for the noninvasive detection of AR. Other proposed noninvasive detection techniques which have shown efficacy, such as metabolomics (24), proteomics (25, 26), and DNA microarray profiling (27,28), are costly, time intensive, computationally expensive, and require extensive infrastructure to carry out analysis. A RS-based system offers a decided advantage by focusing on determined foci of spectral differences thus providing a low-cost, rapid analysis which can be outfitted for portable use in the clinical setting. The RS system can detect T cell activation within both blood and urine biologic matrices. This flexibility is afforded by the component subtraction of Raman spectral contributions made by other cells, fluids, and free floating macromolecules. The system can use “learned” T cell activation signatures to select T cells that match the stored cellular profiles even from within a multi-cellular fluid. Moreover, the ability of RS to differentiate between the signatures of T cells activated by different means affords a potential for high specificity. The value of this high specificity is three-fold. First, it provides an avenue to move away from the paradigm of needing an invasive biopsy to confirm the diagnosis of AR and also provides an alternative to protocol biopsies to detect SCAR. Second, RS detection of AR offers a modality that could significantly reduce the delay in AR diagnosis. Under the current algorithm, delays in treatment are created due to the need to eliminate other potential causes of SCE, thus potentially allowing further histologic damage to the nephrons of transplanted graft. Third, a comprehensive panel of clinically relevant signatures based upon viral (CMV), fungal (Candida), and bacterial (E. coli) stimuli could be investigated which will further improve the rapid diagnosis and management of infections in the post-transplant period.

Conclusion

Based upon RS analysis of the cell surfaces of alloreactive and CD3/CD28-activated T lymphocytes, we conclude that the receptor expression and resulting spectral signatures of these two activated T cell populations differ significantly. These unique RS signatures will allow for a more refined approach toward the development of a noninvasive AR detection system that has a high specificity and sensitivity.

Footnotes

2This work was supported by the National Institutes of Health Grant 5R01EB000741-05, the National Institutes of Health Grant 2T32 GM008420-14, and the David Fromm Research Award, Wayne State University Department of Surgery.

References

1. Flechner S, Goldfarb D, Solez K, Modlin C, Mastroianni B, Savas K, Babineau D, Kurian S, Salomon D, Novick A, Cook D. Kidney transplantation with kidney transplantation with sirolimus and mycophenolate mofetil-based immunosuppression: 5-year results of a randomized prospective trial compared to calcineurin inhibitor drugs. Transplantation. 2007;83:883–892. [PubMed]
2. Foster C, Philosophe B, Schweitzer E, Colonna J, Farney A, Jarrell B, Anderson L, Bartlett S. A decade of experience with renal transplantation in African-Americans. Annals of Surgery. 2002;236:794–805. [PubMed]
3. Spring 2007 Regional Meeting Data of the U.S. Organ Procurement and Transplantation Network and the Scientific Registry of Transplant Recipients. http://www.optn.org/latestData.
4. Meier-Kriesche H, Ojo A, Magee J, Cibrik D, Hanson J, Leichtman A, Kaplan Bruce. African-American renal transplant experience decreased risk of death due to infection: possible implications for immunosuppressive strategies. Transplantation. 2000;70:375–379. [PubMed]
5. Campbell S, Hothersall E, Preston J, Brown A, Hawley C, Wall D, Griffin A, Isbel N, Nicol D, Johnson D. Frequency and severity of acute rejection in live-versus cadaveric-donor renal transplants. Transplantation. 2003;76:1452–1457. [PubMed]
6. Nett P, Heisey D, Shames B, Fernandez L, Pirsch J, Sollinger H. Influence of kidney function to the impact of acute rejection on long-term kidney transplant survival. Transpl Int. 2005;18:385–389. [PubMed]
7. Matas A, Simmons R, Kjellstrand C, Najarian J. Pseudorejection: factors mimicking rejection in renal allograft recipients. Ann Surg. 1977;186:51–59. [PubMed]
8. Nankivell BJ, Borrows RJ, Fung CLS, et al. Natural history, risk factors and impact of subclinical rejection in kidney transplantation. Transplantation. 2004;78 242.4. [PubMed]
9. Nankivell BJ, Fenton-Lee CA, Kuypers DRJ, et al. Effect of histological damage on long term kidney transplant outcome. Transplantation. 2001;71 515.5. [PubMed]
10. Nankivell BJ, Borrows RJ, Fung CLS, et al. The natural history of chronic allograft nephropathy. N Eng J Med. 2003;349:2326. [PubMed]
11. Brown KL, Palyvoda OY, Thakur JS, Nehlsen-Cannarella SL, Fagoaga OR, Gruber SA, Auner GW. Raman spectroscopic differentiation of activated versus non-activated T lymphocytes: An in vitro study of an acute allograft rejection model. J Immunol Methods. 2009;340:48–54. [PMC free article] [PubMed]
12. Mannie MD, McConnell TJ, Xie C, Li Yo. Activation-dependent phases of T cells distinguished by use of optical tweezers and near infrared Raman spectroscopy. J Immunol Methods. 2005;297:53–60. [PubMed]
13. Boyum A. Isolation of leukocytes from human blood. Further observations methyl, cellulose, dextran, and ficoll as erythrocyte aggregating agents. Scand J Clin Invest. 97 Suppl:1968–1931. [PubMed]
14. Dupont B, Hansen JA, Yunis EJ. Advances in immunology. New York: Academic Press; 1976. Human mixed-lymphocyte culture reaction: Genetic, specificity and biological implications; p. 107. [PubMed]
15. Melamed MR, Mullaney PF, Mendelsohn ML. Flow Cytometry and Sorting. 2nd ed. New York: Wiley-Liss; 1990.
16. Fleisher TA, Marti GE, Hagengruber C. Two-color flow cytometric analysis of monocyte depleted human blood lymphocyte subsets. Cytometry. 1988;9:309–315. [PubMed]
17. Schwarting R, Biedobitek G, Stein H. Cluster report: CD69. In: Knapp W, Dörken B, Gilks WR, et al., editors. Leucocyte Typing IV: White Cell Differentiation Antigens. New York, NY: Oxford University Press; 1989. pp. 428–432.
18. Notingher I, Verrier S, Haque S, Polak JM, Hench LL. Spectroscopic study of human lung epithelial cells (A549) in culture: living cells versus dead cells. Biopolymers. 2003;72(4):230–240. [PubMed]
19. Janeway C, Travers P, Walport M, Capra J. The major histocompatibility complex of genes: organization and polymorphism. In: Austin P, Lawrence E, editors. Immunobiology: The immune system in health and disease. 4th edition. NY: Garland Publishing; 1999. p. 135.p. 148.
20. Deng H, Bloomfield V, Benevides J, Thomas G. Dependence of the Raman signature of genomic B-DNA on nucleotide base sequence. Biopolymers. 1999;50:656. [PubMed]
21. Erfurth S, Peticolas W. Melting and premelting phenomenon in DNA by laser Raman-scattering. Biopolymers. 1975;14:247. [PubMed]
22. Benevides J, Thomas G. Characterization of DNA structures by Raman spectroscopy: high-salt and low-salt forms of double helical poly (dG-dC) in H2O and D2O solutions and application to B, Z and A-DNA. Nucleic Acids Res. 1983;11:5747. [PMC free article] [PubMed]
23. Puppels G, Demul F, Otto C, Greve J, Robertnicoud M, Arndtjovin D, Jovin T. Studying single living cells and chromosomes by confocal Raman microspectroscopy. Nature. 1990;347:301. [PubMed]
24. Wishart D. Metabolomics: The principles and potential applications to transplantation. American Journal Transplant. 2005;5(12):2814–2820. [PubMed]
25. Clarke W, Silverman C. Characterization of Renal Allograft Rejection by Urinary Proteomic Analysis. Annals of Surgery. 2003;237(5):660–665. [PubMed]
26. Mann M, Kelleher N. Precision proteomics: The case for high resolution and high mass accuracy. PNAS. 2008;105(47):18132–18138. [PubMed]
27. Sarwal M, Chua M, Kambham N, Hsieh S, Satterwhite T, Masek M, Salvatierra O. Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. N Engl J Med. 2003;349:125–138. [PubMed]
28. Akalin E, Hendrix R, Polavarapu R, Pearson T, Neylan J, Larsen C, Lakkis F. Gene expression analysis in human renal allograft biopsy samples using high-density oligoarray technology. Transplantation. 2001;72(5):948–953. [PubMed]