Glycoprotein methods protocols - biotechnology
Assay for MHC-Unretricted MUC1-Specific CTL 46346338Analysis of the Frequency of MHC-UnrestrictedMUC1-Specific Cytotoxic T-Cells in Peripheral Bloodby Limiting Dilution AssayJohn R. McKolanis and Olivera J. Finn1. IntroductionMucins are highly glycosylated proteins present on the lumenal side of ductal epi-thelial cells. MUC1 is the only mucin with a transmembrane region anchoring it to thecell surface. The extracellular domain of MUC1 is composed of numerous tandemrepeats of a 20-amino acid sequence (1). Normal cells produce a highly glycosylatedand sialylated form of MUC1. The O-linked carbohydrate side chains of MUC1 ontumor cells are shorter and less abundant, exposing previously unrecognized antigenicsites on the polypeptide core of the molecule (2). This underglycosylated MUC1 mol-ecule on tumor cells can be recognized as a tumor-specific antigen by T-cells. Themajor cytotoxic T-lymphocyte (CTL) response to tumor-specific MUC1 is T-cellreceptor (TCR) mediated but major histocompatibility complex (MHC)-unrestricted.Owing to the high density of repeating antigenic epitopes extending along each MUC1molecule, it is postulated that a large number of TCRs can be triggered simultaneouslyto activate the CTLs to kill tumor cells or proliferate (3–5). Antibodies specific for theTCR, or for the defined MUC1 epitope recognized by the TCR, inhibit CTL recogni-tion. Antibodies against MHC molecules have no effect.In addition to the MHC-unrestricted CTLs, some MUC1-specific MHC-restrictedCTLs have also been identified. Patients who are HLA-A11 and HLA-A3 have CTLsthat recognize a nine amino acid peptide from the tandem repeat region bound to thesealleles (6). Another peptide has been reported that binds HLA-A2.1 allele (7).Although it is not yet clear whether these CTLs, MHC unrestricted or MHC restricted,have a therapeutic function, it has been considered important to measure their fre-quency pre- and post vaccination with MUC1-based immunogens used in various clini-cal trials (7–9) in order to be able to correlate this specific component of the antitumorresponse with the outcome of the trial.From:Methods in Molecular Biology, Vol. 125: Glycoprotein Methods and Protocols: The MucinsEdited by: A. Corfield © Humana Press Inc., Totowa, NJ 464 McKolanis and FinnWe have established a limiting dilution assay (LDA) to evaluate the frequency ofMHC-unrestricted MUC1-specific CTL. This assay is convenient because the sameconditions and reagents can be applied to all patients regardless of their HLA type. Webelieve that these CTLs will be an important corollary of a tumor-specific immuneresponse because they recognize an epitope that is expressed only on tumor cells andnot on normal epithelial cells. The same assay may be used to determine the frequencyof MHC-restricted CTL. However, the assay must be custom tailored to each patient,or at least to several specific HLA alleles.2. Materials1. RPMI-1640 tissue culture medium supplemented with 10% fetal calf serum (FCS) orhuman AB serum, 2 mM L-glutamine, 100 Units/mL penicillin, 100 µg/mL streptomycin.2. Human male AB serum purchased from Gemini Bio Products, Calabasas, CA.3. FCS purchased from Life Technologies, Grand Island, NY.4. Glutamine, purchased as 100X stock from Life Technologies, Grand Island, NY.5. Penicillin-streptomycin, purchased as 100x stock from Life Technologies.6. Hepes buffer, purchased as 100X stock from Life Technologies.7. Recombinant interleukin-2 (rIL-2), obtained from several sources.8. Phytohemagglutinin (PHA-P), purchased from Sigma, St. Louis, MO.9. Trypan blue, 0.4% stock solution, purchased from Life Technologies.10. Phosphate buffered physiologic saline (1.9 mM NaH2PO4, 8.1 mM NaHPO4, 154 mMNaCl, pH 7.2); may also be purchased as a 10X stock solution.11.51Sodium chromate and 3H thymidine, purchased from Amersham, Arlington Heights, IL.12. Ficoll density gradient, purchased as LSM from Organon Technica Corp, Durham, NC.13. DNAase, purchased from Calbiochem, La Jolla, CA. Filter-sterilized stock solution at10 mg/mL in RPMI-1640 containing HEPES buffer should be kept frozen at –80°C.14. Dimethyl sulfoxide (DMSO) purchased from Sigma.15. Tumor cells that express MUC1 can be obtained from American Type Culture Collection(ATCC), Rockville, MD, or as gifts from various investigators. Epstein-Barr virus (EBV)immortalized B-lymphoblastoid cell lines or other cells transfected with the MUC1 genecan be obtained from various investigators. Natural killer and lymphokine-activated killer(LAK) cell target K562 can be purchased from ATCC.16. Styrofoam boxes for slow freezing of cells, purchased from Sarstadt, Newton, NC.17. 96-well, U-bottomed microtiter plates, purchased from Becton Dickinson, Franklin Lakes, NJ.18. Cytotoxicity assay harvesting materials, purchased from Scatron Instruments, Sterling, VA.3. Methods3.1. Isolation of Mononuclear Cells from Peripheral BloodBlood samples are obtained from patients at specific time intervals, depending onthe goals of the study. Peripheral blood mononuclear cells are separated from the restof the blood using cenrifugation through Ficoll density gradient (LSM). This involveslayering 10 mL of blood diluted 1:2 in RPMI medium without serum, over 3 mL ofLSM in a 15-mL conical centrifuge tube. Following centrifugation at 400g for 20 min,mononuclear cells can be removed from the interface between the two layers. Thecells are washed three times in phosphate-buffered saline (PBS) to remove contami-nating LSM. From 1 mL of blood we obtain 1–2 × 106cells. Assay for MHC-Unretricted MUC1-Specific CTL 4653.2. Cryopreservation of CellsMononuclear cells are frozen in 90% FCS and 10% DMSO at a concentration of5–10 × 106cells/mL. The cells are first resuspended in cold FCS at one-half of thefinal volume. The remaining volume of cold FCS with 20% DMSO is then addeddropwise while swirling the tube. A controlled rate of freezing can be simulated bysimply placing the vials in a styrofoam box in a –80°C freezer overnight. The vials arethen transferred to liquid nitrogen for long-term storage.At the time of the assay, the vials are thawed rapidly in a 37°C water bath. Thecontents are transferred to 10 mL of RPMI-1640 containing 30 µg/mL of DNAase.Following a 30-min incubation at 37°C, the cells are washed three times in PBS. Viablecells are counted by diluting the sample 1:2 in trypan blue and counting in a hemocy-tometer the cells that have not taken up the color blue.3.3. Setting up the LDA CulturesIn the LDA culture assay, hundreds of peripheral blood lymphocyte microculturesare stimulated once in vitro with the appropriate antigen. The cells are grown in 96-wellU bottomed plates.1. At each dilution, seed peripheral blood lymphocytes in a 100-µL volume into 24 wells ofa round-bottomed 96-well microtiter plate (see Fig. 1). We routinely use twofold dilu-tions of cells starting with 2 × 105cells/well and ending with 6.25 × 103cells/well. (seeNote 1). Culture medium is RPMI-1640 supplemented with 10% human AB serum, peni-cillin, and streptomycin.2. Add irradiated (6000 RADS) stimulator cells expressing MUC-1 in a 50-µL volume at aconcentration of 2 × 104/well.3. Use irradiated (3000 rad) autologous peripheral blood lymphocytes (PBLs) as “feeder”cells, and add at a concentration of 2 × 104/well in a 50-µL volume. Suspend the “feeder”cells in culture medium containing one half of the optimal concentration of IL-2 (seeNotes 1–4).4. Following 7 d, in culture, remove 100 µL of culture medium from each well and replacewith fresh medium containing IL-2. Perform the CTL assay on d 10.3.4. Setting up the CTL Assay1. If adherent, trypsinize target cells and wash 1 × 106cells three times in PBS. After the lastwash, carefully aspirate the supernatant so that the smallest possible volume remains.2. Resuspend the cell pellet in 100 µL of fresh 51Cr, specific activity 1 mCi/mL, and incu-bate at 37°C for 1–2 h. Shake the tube periodically to keep the cells in suspension.3. Following three washes, place 1 × 103 cells in a gamma counter to determine the uptakeof radioactivity and thus ensure proper labeling. If less than 500 counts are obtained withtumor cells or 300 counts with EBV-transformed B-cells, a new batch of targets should beprepared. The new targets should be labeled for a longer period of time or newer chro-mium should be used.4. While the targets are being prepared, the rest of the assay can be set up. Each well of theoriginal plates should be carefully mixed with a micropipettor and transferred into twowells of a 96-well microtiter V-bottomed plate (“split-well” assay). Thus, row A of theoriginal 96-well plate becomes rows A and B of the assay plate (see Fig. 1). Each newwell contains 100 µL of the original 200-µL cell cuture. 466 McKolanis and Finn5. To each new well, add 2.5 × 104unlabeled LAK target cells in a volume of 50 µL. ForCTL assays with tumor cells as labeled target cells, we use unlabeled K562 cells. WhenMUC1 transfected EBV B-cells are used as targets, we use unlabeled parental EBV cellsto compete out the LAK lysis.6. At each dilution, add 1 × 103of the labeled specific targets to one of the split wells and tothe other well 1 × 103labeled control targets, in 50-µL volumes. Include several wellscontaining only target cells for spontaneous and maximal release values7. Centrifuge the plates at 150g for 5 min and incubate at 37°C for 4 h.8. Harvest supernatant from the wells and count the radioactive chromium released fromlysed cells. Obtain maximal release values following lysis of the cells in designated wellswith acid or hypertonic saline.9. Calculate specific release using the following formula: 100 × (sample release – spontane-ous release/maximal release – spontaneous release).10. Calculate percentage of specific release for each well containing the specific target andcompare this with the corresponding well containing the control target. Wells are scoredpositive if specific target cell lysis is 15% greater than the control (see Note 7).Fig. 1. An example of an experiment utilizing limiting dilution culture, showing the processof setting up the original culture plates and the final assay plates. In the culture plates, each cellconcentration is represented by two rows (24 wells). Each culture well is split into two assaywells. The black wells in the assay plates were scored positive for CTL activity. Rows A con-tain the MUC1+ target, and rows B contain the negative control target. Assay for MHC-Unretricted MUC1-Specific CTL 4673.5. Calculation of the CTL FrequencyInasmuch as positive wells may contain more than one responding T-cell, the fre-quency of antigen-specific cells is determined by the number of negative wells in eachdilution. The number of negative wells is divided by the total number of wells at eachdilution. This fraction is converted to its base 10 log and is plotted on a log scale againstthe number of cells added to each well. The Maximal Likelihood method is one of thebest statistical estimators used to find the best fitting straight line going through theorigin (10). Linear regression can also be utilized. It is assumed in these assays that allparameters for T-cell growth are optimal and that any well containing antigen-specificcells will be positive in the assay. The single hit model of Poisson distribution predictsthat when 37% of the test cultures are negative, there is an average of one precursor cellper well (11). Several statistical computer programs are available to simply evaluate thedata. Figure 2 shows graphic data from the assay represented by Fig. 1 on an individualwith slightly elevated frequency of MUC1-specific cells.3.6. LDA Special Considerations3.6.1. Time of AssayLDA assays can be used to quantitate specific CTL precursors that need in vitro reactiva-tion, or recently activated specific effector CTL. Precursor CTL are determined following 10or 11 d in culture with antigen. Activated CTL could be assayed after 3–5 d of culture.Fig. 2. MUC1-specific CTL frequency derived from the results depicted in Fig. 1. The solidline represents CTL frequency derived in the absence, and the broken line represents CTLfrequency derived in the presence of anti-HLA class I antibody W632, illustrating the MHC-unrestricted nature of the CTL. The log of the fraction of negative cultures is plotted on a logscale against the number of cells per well. Linear regression analysis was used to calculate theslope. The wells containing 200,000 cells had no negative wells; thus, the value is 0 and is notplotted. A frequency of 1/65,949 for untreated cultures and 1/57,542 for antibody-treated cul-tures is determined when 37% of the cultures are negative. This is shown at the intersection ofthe two plotted lines with the horizontal dashed line. 468 McKolanis and Finn3.6.2. Number of CellsThe frequency of MUC1-specific precursor cells in most healthy individuals isabout 1 in 106cells. This frequency is increased to varying degrees in cancer patientsand is in the range of 1 in 150,000 to 1 in 300,000. Following immunization of cancerpatients in phase I trials, we have observed that the frequency in some patients mayincrease between two- to 11-fold (8). These are, nevertheless, relatively low frequen-cies that require cell concentrations for the assays to start at 2 × 105cells/well. Wehave found, however, that greater than 4 × 105cells/well can result in overcrowding ofthe well and inconsistent results. Optimal statistical analysis requires at least 24 repli-cates of each dilution. Thus, the minimum number of cells required is 1.3 × 107foreach patient to be tested.3.6.3. LAK vs Specific CTL ActivityOwing to the presence of IL-2 in the culture medium and additional IL-2 secretedfollowing T-cell activation, some nonspecific cytokine-activated cell lysis can occur,which increases the background noise. We use unlabeled K562 cells as cold-targetcompetitors with labeled targets to minimize the background. Other unlabeled (cold)targets could be used to compete out possible alloreactive responses or EBV-specificresponses (see Subheading 3.6.4.).3.6.4. Target CellsEBV-transformed MUC1-transfected B-cells are excellent stimulator cells because theyalso express costimulation molecules necessary for optimal T-cell activation. However,unless these cell lines are derived from each patient and only autologous cells are used inthe assay, it is possible to generate allospecific CTL during the culture period. If autolo-gous cells are used, often EBV-specific CTL are activated together with the MUC1-spe-cific CTLs. One way to overcome this problem is to use the parental untransfected B-cellline as the cold-target inhibitor. Appropriate control target cell in the split-well assayswould be the same untransfected EBV-transformed cell, or a tumor cell target that expressesMUC1 but does not share MHC alleles with the stimulator cells.The MHC-unrestricted nature of the MUC1-specific CTL allows us to use a morestandardized and less labor-intensive approach. We stimulate the PBL in vitro withone MUC1+tumor cell line and test CTL function against another MUC1+tumor cellline that shares no MHC alleles with the stimulator tumor. Allospecific cytotoxicity, ifany generated in the cultures, would be ineffective against the second target.4. Notes1. It is most convenient to prepare dilutions of cells directly in the microtiter plates. Thefirst two rows receive 4 × 105cells in 200 µL. Using a 12 -well multichannel pipettor, 12serial dilutions can then be performed simultaneously by transferring 100-µL volumesinto lower wells containing 100 µL of medium alone.2. Sufficient IL-2 is added to the culture to initiate the growth of antigen stimulated T-cellsbut not the growth of nonspecific lymphokine activated T-cells. Thus, if 20 U/mL is theoptimal concentration of IL-2, the stimulator cells are resuspended in 10 U/mL of IL-2.Thus, in this example, each well receives 0.5 U of the cytokine. Assay for MHC-Unretricted MUC1-Specific CTL 4693. It is also possible to culture the cells for 3 d before adding IL-2. This may reduce nonspe-cific background from cytokine-activated T-cells.4. Depending on the source, IL-2 is purchased in units/milliliter or micrograms/milliliterquantities. Each new batch should be evaluated to determine the concentration that pro-motes optimal growth. Dilutions of IL-2 from 0.1 to 100 U/mL or 0.1–100 ng/mL areprepared in RPMI-1640 containing 10% human AB serum, penicillin/streptomycin,glutamine, and HEPES. Using 96-well microtiter U-bottomed plates, triplicate 100-µLaliquots of each dilution are added to 100 µL of human PHA blasts at a concentration of 1× 105/mL. PHA blasts are prepared by culturing 5 × 105Ficoll-separated peripheral bloodmononuclear cells in 5 mL of 10% human AB serum with 10 µg/mL PHA for 3 d at 37°Cin a humidified CO2incubator. The cells can be frozen in 10% DMSO and 90% serum,and stored in small aliquots indefinitely in liquid nitrogen. Before use, the PHA blasts arethawed and rested by overnight incubation in 10% FCS at 37°C to obtain an accurateviable count. Following incubation in different IL-2 dilutions for 3 d at 37°, 1 µCi of 3Hthymidine is added. After 18–20 h at 37°C the cells are harvested and counted. The dilu-tion of IL-2 that gives 50% of maximal growth is considered the optimal concentration.5. Successful culture of T-cells requires enriched media such as RPMI-1640 or AIM-5,human male AB serum, and IL-2. Owing to lot-to-lot variation, AB serum should also betested for supporting optimal cell growth. Testing of human serum is done with PHAblasts prepared as described above. Vendors will supply small samples of several serumlots on request. Cells (1 × 105) are thawed and cultured for 4 d in RPMI-1640 with 10%test serum, and the optimal concentration of IL-2 determined as described in Note 4. Aviable cell count is a reliable way to determine proliferation. A second method is to assessincorporation of 3H thymidine by mitotic cells. This is done in triplicate using 96-wellmicrotiter plates. Viable cells (3 × 104) are cultured overnight with 1 µCi of 3H thymidine.The cells are then harvested and counted for uptake of radionuclide by dividing cells.6. It may be possible to assay simultaneously more than two target cells by splitting theoriginal well into a number of test wells, depending on how vigorously the T-cells haveproliferated during the 10-d culture period.7. The arbitrary discrimination between positive and negative cultures is usually a 15% dif-ference in specific release. Some patients have low levels of specific release and little orno observed background LAK activity. It may be possible in these patients to considerlower specific lysis as positive.References1. Hilkins, J. and Bujis, F. (1988) Biosynthesis of MAM-6, an epithelial sialomucin. Evi-dence for involvement of a rare proteolytic cleavage step in the endoplasmic reticulum.J Biol. Chem. 263, 4214–4222.2. Girling, A., Bartkova, J., Burchell, J., Gendler, S., Gillet, C., and Taylor-Papadimitriou, J.(1989) A core protein epitope of the polymorphic epithelial mucin detected by mono-clonal antibody SM-3 is selectively exposed in a range of primary carcinomas. Int.J. Cancer 43, 1072–1076.3. Finn, O. J. (1992) Antigen-specific, MHC-unrestricted T cells. Biotherapy 4, 239–249.4. Magarian-Blander, J., Ciborowski, P., Hsia, S., Watkins, S. C., and Finn, O. J. (1998) Inter-cellular and intracellular events following the MHC-unrestricted TCR recognition of a tumor-specific peptide epitope on the epithelial antigen MUC-1. J. Immunol. 160, 3111–3120.5. Magarian-Blander, J., Hughey, R. P., Kinlough, C., Poland, P. A., and Finn, O. J. (1996)Differential expression of MUC1 on transfected cell lines influences its recognition byMUC1 specific T cells. Glycoconjugate J. 13, 749–756. 470 McKolanis and Finn6. Domenech, N., Henderson, R. A., and Finn, O. J.(1995) Identification of an HLA-A11-Restricted Epitope from the tandem repeat domain of the epithelial tumor antigen mucin.J. Immunol. 155, 4766–4774.7. Apostolopulos, V., Karanikas, V., Haurum, J. S., and McKenziee, I. F .C. (1997) Induc-tion of HLA-A2-restricted CTLs to mucin 1 human breast cancer antigen J. Immunol. 159,5211–5218.8. Goydos, J. S., Elder, E., Whiteside, T. L., Finn, O. J., and Lotze, M. T. (1996) A phase Itrial of a synthetic mucin peptide vaccine. J. of Surg. Res. 63, 298–304.9. Karanikas,V., Hwang, L., Pearson, J., and Ong, V., (1998) Antibody and T cell responsesof patients with adenocarcinoma immunized with mannan-MUC1 fusion protein. J. Clin.Invest. 100, 2783–2793.10. Taswell, C. (1981) Limiting dilution assays for the determination of immunocompetentcell frequencies. J. Immunol. 126, 1614-1619.11. Sharrock, C. E. M., Kaminski, E., and Man, S. (1990) Limiting dilution analysis of humanT cells a useful clinical tool. Immunol. Today 11, 281–285. . Successful culture of T-cells requires enriched media such as RPMI-1640 or AIM-5,human male AB serum, and IL-2. Owing to lot-to-lot variation, AB serum. MUC1 mol-ecule on tumor cells can be recognized as a tumor-specific antigen by T-cells. Themajor cytotoxic T-lymphocyte (CTL) response to tumor-specific