Human Virus-Specific CD8⁺ CTL Clones Revert from CD45RO^high to CD45RA^high In Vivo: CD45RA^highCD8⁺ T Cells Comprise Both Naive and Memory Cells

Donors

Five healthy HCMV-seropositive laboratory donors and two patients with primary HCMV were included in this study. All of the laboratory donors were HCMV seropositive, as determined by an IgG ELISA (Captia HCMV IgG immunoassay; Centocor, Malvern, PA). The patients with primary HCMV were confirmed by HCMV-specific IgM and subsequent seroconversion for HCMV-specific IgG (PHLS; Addenbrookes Hospital, Cambridge, U.K.). The MHC class I tissue type of each donor (Table I⇓) was determined by serologic typing (Lymphotype ABC-120; Biotest, Dreieich, Germany).

Viruses and cell lines

HCMV AD169 (ATCC VR-538) was grown in GMO5387 fibroblasts (Coriell Cell Repositories, Camden, NJ) infected at a multiplicity of infection (moi) of 0.01. Whole infected cultures were harvested 5 days after 100% cytopathic effect was evident, and were spun at 8000 rpm. Pellets were pooled, sonicated in an ice-cooled water sonication bath, aliquoted, and frozen at −70°C. Stocks were titered on 12-well plates of GMO5387s using 10-fold dilutions of virus; cytopathic effect was read after incubation for 10 days. Recombinant vaccinia viruses expressing the HCMV protein pp65 (vac pp65; gift of Dr. S. Riddel, Fred Hutchinson Cancer Research Centre, Seattle, WA), and a negative control expressing bacteriophage RNA polymerase T7 (vac T7) were grown in BHK cells infected at an moi of 0.1. After 48–72 h, the infected cells were harvested and subjected to three rounds of freeze-thaw, followed by sonication. The cell debris was removed by centrifugation, and supernatant-containing virus was aliquoted and stored at −70°C. A sample was subsequently titered on Vero cells; titers were normally between 1.5 and 10 × 10⁷ plaque-forming units/ml.

Peptides based on the lower matrix protein pp65

Minimal peptides corresponding to proposed MHC-restricted epitopes of HCMV pp65 were synthesized and purified (>95%) by HPLC (Affiniti Research Products (Exeter, U.K.) (16, 17). All peptides were dissolved in RPMI 1640 at 200 μg/ml and frozen in small aliquots at −70°C.

mAbs and surface phenotyping

PBMC were prepared from fresh heparinized venous blood samples by Ficoll-Hypaque (Lymphoprep; Nyegaard, Oslo, Norway) density-gradient centrifugation. The surface phenotype of fresh PBMC and cells purified by cell sorting was determined by flow cytometry. mAbs conjugated with FITC, PE, or peridinin chlorophyll protein (PerCP) used were anti-CD4 FITC (Leu3a), anti-CD8 FITC/PE/PerCP (Leu2a), anti-CD16 FITC (Leu11a), anti-CD19 FITC (SJ25C1), anti-CD45RA FITC (Leu18), anti-CD45RO PE (UCHL-1), anti-CD56 PE (Leu19). The proportion of cells using specific TCR Vβ-chains was determined by three-color immunofluorescence using a panel of TCR-specific FITC-conjugated mAbs (the panel included: Vβ 1, 2, 3, 5.1, 5.2, 6.7, 7, 8.1, 11, 12, 13.6, 14, 16, 17, 19, 20, 21.3, and 22 (Coulter, Palo Alto, CA); Vβ 13.1/13.3 and Vα2 (Serotech)) with PE-conjugated anti-CD8 and PerCP-conjugated CD3. Samples were analyzed on a FACSort flow cytometer (Becton Dickinson, Oxford, U.K.) and using WinMDI software (Joseph Trotter, Scripps, http://facs.scripps.edu).

Preparation of cell populations

Total CD8⁺, CD8⁺CD45RO^high, and CD8⁺CD45RA^high cells were prepared from PBMC using negative cell sorting to minimize any effects of ligation of CD8, CD45RA, or CD45RO on the cell surface. It is not possible to purify intermediate CD45RA⁺RO⁺ cells by negative selection. PBMC were stained with FITC-conjugated anti-CD4, anti-CD19, anti-CD16, and anti-CD56 mAbs (to remove CD4⁺ T cells, B cells, and NK cells, respectively) and sorted for nonstained cells using a FACSVantage cell sorter (Becton Dickinson). CD8⁺CD45RO^high cells were isolated by staining another aliquot of PBMC with anti-CD4, anti-CD19, anti-CD16, anti-CD56, and anti-CD45RA, and sorting for nonstained cells. CD8⁺CD45RA^high cells were isolated by staining another aliquot of PBMC with anti-CD4, anti-CD19, anti-CD16, anti-CD56, and anti-CD45RO, and sorting for unstained cells. The purity of the CD45RA^high cells was >99% (CD45RO^high cell contamination 0.1–0.65%); the purity of the CD45RO^high cells was >99% (CD45RA high cell contamination 0.1–0.27%) (Fig. 1⇓, C and F).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Separation of total CD8⁺, CD8⁺CD45RA^high, and CD8⁺CD45RO^high T cells from PBMC by negative cell sorting and quantitation of pp65-specific CTLp frequencies within each population by LDA. Total CD8⁺ cells were enriched from PBMC of donor 001 by staining with mAbs to CD4, CD16, CD19, and CD56 and sorting for negative cells. CD45RA^high cells were derived as for total CD8 cells with the addition of a mAb to CD45RO-PE and sorting for negative cells. CD45RO^high cells were derived as for total CD8 cells with the addition of a mAb to CD45RA-FITC and sorting for negative cells. A, C, and E, Aliquots of all sorted cells were restained with CD8-PerCP, CD45RA-FITC, and CD45RO-PE and gated on CD8⁺ cells to assess the sorting purity. B, D, and F, All three populations were stimulated with autologous PBMC pulsed with HCMV, and CTLp frequencies were determined by LDA against target cells expressing vaccinia pp65 (▪) and as a control vaccinia T7 (▴).

Generation of HCMV-specific CTL in limiting dilution assays (LDA)

Replicate microcultures (n = 27) of purified CD8⁺ T cell populations were set up in 96-well round-bottom plates in which the number of responder T cells per well was progressively reduced over an appropriate range of dilutions in RPMI 1640 supplemented with 10% human AB serum (HCMV seronegative; Blood Transfusion Service, Addenbrookes Hospital, Cambridge, U.K.), 2 mM l-glutamine, 10⁵ IU/L penicillin, and 100 mg/L streptomycin (referred to as RPMI-HuAB). As stimulator cells, autologous PBMC were pulsed for 1 h with HCMV (moi 0.01), irradiated (2400 rad), and added at 5 × 10⁴ cells/well. Although HCMV does not appear to establish a productive infection of unstimulated PBMC, exogenous pp65 protein is able to enter the MHC class I processing pathway (15) for stimulation of pp65-specific CD8⁺ T cells. The medium was further supplemented with human rIL-2 to give a final concentration of 5 IU/ml. LDA cultures were incubated at 37°C in 5% CO₂ and refed with RPMI-HuAB supplemented with 5 IU/ml IL-2 on days 5 and 10. On day 14, using split-well analysis, the cells in each individual well were resuspended and divided into aliquots that were assayed simultaneously for cytotoxicity against radiolabeled target cells in 4-h ⁵¹Cr release assays. Target cells comprised 4 × 10³ cells/well autologous and MHC-mismatched lymphoblastoid B cell lines that were pulsed or unpulsed with pp65 peptide (40 μg/ml) for 1 h, or infected for 18 h with vac pp65 or control vac T7 (moi = 10). Target cells were washed three times and counted before use in chromium release assays. The LDA results were analyzed as described previously (16).

PCR amplification of TCR Vβ regions

Total RNA was extracted from 1–10 × 10⁵ cells using an RNA extraction kit (Qiagen, West Sussex, U.K.). First strand cDNA was reverse transcribed with an oligo(dT) primer and avian myeloblastosis virus reverse transcriptase using a reverse-transcription kit (Promega, Madison, WI), according to the manufacturer’s instructions. PCR was performed using a panel of 36 TCR Vβ family-specific primers (17) together with the corresponding C region-specific primer (synthesized by Genosys Biotechnologies, Cambridge, U.K.). Each reaction was conducted in a total volume of 50 μl containing 1 mM of each dNTP (Boehringer Mannheim, Indianapolis, IN), 3.75 mM MgCl₂, each primer at a final concentration of 1 mM, and 1 U of Taq polymerase (Promega), in buffer supplied by the manufacturer. A total of 2 μl of cDNA was used for PCR amplification in each case. The reaction was overlaid with mineral oil, and amplification was performed for 45 cycles of PCR. Conditions were 1-min denaturation at 94°C, 30 s of annealing at 60°C, and 30 s of extension at 72°C on a DNA thermal cycler (Perkin-Elmer Cetus 9600 Instruments, Norwalk, CT). A total of 45 μl of each PCR product was separated on a 1.3% agarose gel. Expression of Vβ genes was considered positive where an approximately 300–400 nt rearranged band could be visualized with ethidium bromide staining.

Quantitative clonotypic analysis

We previously generated peptide-specific CTL clones in vitro and sequenced the TCR β-chain (17). To quantify a defined peptide-specific CTL clone in PBMC by molecular assay, we designed a complementary oligonucleotide 15–20 nt long based on the TCR β-chain hypervariable n-D-n region of that immunodominant CTL clone (referred to as the clonotypic probe (Table I⇑)). We also designed a conserved TCR C region probe to quantify total amplified TCR sequences. Vβ family-specific PCR was performed in duplicate on cDNA from total CD8⁺, CD8⁺CD45RA^high, and CD8⁺CD45RO^high cells from each donor. PCR products were also produced from the defined peptide-specific CTL clone (positive control) and from a mixed population of PBMC derived from four HCMV-seronegative donors (negative control). The duplicate PCR samples were separated on a 1.3% agarose gel, and the DNA was transferred onto a filter (Bio-Rad, Richmond, CA) using standard techniques in blotting buffer (0.6 M NaOH, 0.2 M NaCl) overnight. Filters were washed in 2× SSC and baked at 80°C for 30 min. Filters were prehybridized in 15 ml of 7% SDS, 0.25 M Na₂HPO₄ for 1 h at 10–15°C below Tm of the oligonucleotide probe. The oligonucleotide probes were end labeled with [α-³²P]dATP using T4 polynucleotide kinase; unincorporated label was separated from the probe using a G-25 Sephadex spin column (Amersham, Arlington Heights, IL). Filters were first hybridized with the clonotypic probe overnight at 10–15°C below the Tm of the clonotypic probe in hybridization buffer (7% SDS, 0.25 M Na₂HPO₄, 10% polyethelyne glycol 8000). After hybridization, the filters were washed three times (5% SDS, 20 mM Na₂HPO₄) for 15 min each. Binding of the probe was imaged and quantitated using a phosphor imager (Beckman, Fullerton, CA); the filters were also exposed to x-ray film. Bound clonotypic probe was removed from the filters by soaking in 0.4 M NaOH for 30 min, followed by one wash in 2× SSC. The filter was then rehybridized with the TCR β-chain C region probe (AGATCTCTGCTTCTGATG) following the same protocol as for the clonotypic probe. In each experiment, it was possible to correct for differences in radioactive labeling and binding of the clonotypic and C region probes by measuring the radioactivity of each probe bound to the positive control sample as a reference standard (TCR β-chain sequences amplified from the biologically derived CTL clone contain the unique hypervariable n-D-n region and C region in equal amounts). The proportion of clonotype sequence within the total amplified TCR Vβ sequence for a given population of CD8⁺ cells was calculated as: test sample cpm (probed with clonotypic probe)/test sample cpm (probed with constant probe) × 100; positive control cpm (probed with clonotypic probe)/positive control cpm (probed with constant probe).

In long-term healthy HCMV carriers, many of the pp65 peptide-specific CD8⁺ CTLp in PBMC are CD45RA^high

CD8⁺ T cells purified from PBMC of HCMV carriers by negative cell sorting were a mixture of CD45RA^high, CD45RO^high, and CD45RA⁺RO⁺ intermediates (Fig. 1⇑A). Because CD8⁺CD45RA^high and CD8⁺CD45RO^high T cells are mutually exclusive populations, high purity CD8⁺CD45RA^high cells were obtained by sorting, using staining with anti-CD45RO and collecting negative cells (Fig. 1⇑C); high purity CD8⁺CD45RO^high cells were obtained using staining with anti-CD45RA and collecting the negative cells (Fig. 1⇑E). Each of the populations was stimulated in LDA with autologous irradiated HCMV-infected PBMC. Ag-specific cytotoxicity was assessed using target cells that expressed whole pp65 (vaccinia recombinant) and target cells that were pulsed with the immunodominant pp65 peptide appropriate for donor HLA type (Table I⇑). In five long-term HCMV carriers, the CD8⁺ cell population as a whole contained high frequencies of pp65-specific CTLp (Fig. 1⇑B; Table II⇓). Both the CD8⁺CD45RA^high and CD8⁺CD45RO^high subpopulations contained pp65-specific CTLp, including CTLp specific to the same immunodominant pp65 peptide (Fig. 1⇑, D and F; Table II⇓).

In healthy HCMV carriers, CD8⁺ T cells in PBMC contained a higher percentage of CD45RA^high cells than CD45RO^high cells. In each HCMV carrier, the contribution of CD45RA^high and CD45RO^high cells to the total pool of pp65 peptide-specific memory cells in PBMC was determined by multiplying the peptide-specific CTLp frequencies within each T cell subpopulation by the percentage of CD45RA^high or CD45RO^high cells within CD8⁺ T cells in PBMC (Table II⇑). In two healthy HCMV carriers (001 and 009), 10-fold more of the pp65 peptide-specific CTLp in PBMC were CD45RA^high than were CD45RO^highCD45RA⁻; in two other HCMV carriers (005 and 002), equivalent numbers of the pp65 peptide-specific CTLp in PBMC were CD45RA^high and CD45RO^high.

Large clone sizes of peptide-specific CTL clonotypes are present in CD45RA^high cells in long-term HCMV carriers

In each HCMV carrier, most of the independently derived pp65 peptide-specific CTL clones use the same TCR β-chain at the level of nucleotide sequence, indicating a high degree of clonal focusing (17). To quantify an individual peptide-specific CTL clone, we designed an oligonucleotide probe complementary to the hypervariable β-chain sequence of the CTL clone (Table I⇑). In order first to determine the relationship between the amount of clonotypic DNA and the amount of bound radiolabeled probe in the blotting assay (see Materials and Methods), cDNA from a biological pp65 peptide-specific CTL clone was PCR amplified and added in increasing amounts to TCRs derived from a pool of cDNA from PBMC of four HCMV-seronegative donors. Each of the mixtures of TCR DNAs was probed first with the clonotypic probe (Fig. 2⇓B) and, after stripping the filter, with the C region probe that binds to all TCR β-chain DNA sequences (Fig. 2⇓C). The results show a close correlation between the actual amount of clonotypic sequence added into the test samples and the experimentally derived values using clonotypic probing: the relationship is linear over a wide range (Fig. 2⇓D) that includes the experimental values observed in subsequent experiments (Table III⇓). The technique has the sensitivity to detect approximately 1% clonotype sequence against the background of 99% different n-D-n sequences within the corresponding Vβ family.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Determination of the accuracy and sensitivity of a TCR Vβ clonotyping technique. A, An increasing percentage of TCR Vβ PCR product from a pp65 peptide-specific CTL clone was added to TCR Vβ PCR product derived from pooled cDNA from four HCMV-seronegative donors; a sample of the clonal PCR product (positive control) and of the PCR product derived from the seronegative donors (negative control) was also included. B, After resolution on an agarose gel and transfer of the DNA to a nylon fiter, the filter was probed with a radiolabeled clonotypic probe. C, The filter was stripped of bound clonotypic probe and reprobed with a radiolabeled probe to the TCR C region. D, Percentage of clonotype derived by clonotypic probing technique plotted against percentage of clonotypic PCR product added to each test sample.

We used this probing technique to quantify the size of individual CTL clones in CD8⁺ T cell populations purified directly from PBMC of long-term HCMV carriers. We found that in a given carrier, the same pp65-specific CTL clonotype could be detected in both the CD45RA^high and CD45RO^high CD8⁺ T cells. In some subjects, the clonotype constituted one-half of all of the TCR sequences in the relevant Vβ⁺ population of CD8⁺ cells (Fig. 3⇓, Table III⇑). The absolute clone sizes of individual peptide-specific CTL clones in the total CD8⁺ population were very large, up to 0.8% of CD8⁺ PBMC. In all donors tested, we found that the CD45RA^high population contributed 6- to 10-fold more than the CD45RO^high population to the total clone size in CD8⁺ PBMC (Table III⇑).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Quantitation of an individual peptide 56-specific clone within the total CD8⁺ population and the CD45RA^high and CD45RO^high subpopulations of donor 002. A, Duplicate TCR Vβ PCR products were derived from each of the T cell populations; PCR product derived from the biological peptide 56-specific CTL clone (positive control) and PCR product derived from pooled cDNA from PBMC of four HCMV-seronegative donors (negative control) were resolved on a 1.5% agarose gel, transfered to a filter, and probed with a radiolabeled clonotypic probe. B, The filter was stripped of bound clonotypic probe and reprobed with a radiolabeled probe to the TCR C region. The binding of the probes was quantitated using a phosphor imager, and for each population the clonotype sequence as a percentage of all TCR Vβ sequences was calculated according to the formula in Materials and Methods. Clonotypic analysis for all donors is shown in Table III⇑.

Massive expansion of CD45RO⁺CD8⁺ T cells, associated with fresh ex vivo pp65-specific cytotoxicity, during primary HCMV infection

The large clone sizes of pp65-specific memory CTL in CD45RA^highCD8⁺ cells of long-term virus carriers suggest that following primary HCMV infection, some of the clonally expanded activated CD45RO^high pp65-specific CTL may have reverted to less activated memory CTL that reexpress the CD45RA^high phenotype. To address this possibility directly, we studied two patients during and after symptomatic primary HCMV infection.

Fresh PBMC were obtained from patient CMV104 at 3, 4, and 8 wk after the onset of symptoms (fever, anorexia, headache associated with atypical lymphocytosis, and abnormal liver function tests). PBMC were stained with anti-CD45RA, anti-CD45RO, and anti-CD8, and analyzed by flow cytometry (Fig. 4⇓, A, D, G, B, E, and H); at each time point, PBL counts were determined in the clinical haematology laboratory, from which the absolute number of CD8⁺ cells and the CD8⁺CD45RA^high cells were calculated. Aliquots of unstimulated PBMC were depleted of CD16⁺ NK cells and used as effector cells in direct ex vivo cytotoxicity assays against HLA-matched target cells pulsed with the appropriate pp65 peptides (Fig. 4⇓, C, F, and I). At 3 wk, there was a substantial CD8⁺ lymphocytosis of 4.9 × 10⁹/L (normal range of 0.4–1 × 10⁹/L), the large majority of which were CD45RO^high (4.7 × 10⁹/L). At this time, there was a very high level of ex vivo cytotoxicity against two different HLA-B7-restricted pp65 peptides, p31 and p56. Similar results were obtained 1 wk later. At 8 wk after the onset of symptoms, the absolute CD8 count had decreased to 0.7 × 10⁹ cells/L (within normal range) and there was a substantial reduction in the CD8⁺CD45RO^high population. The absolute number of CD8⁺CD45RA^high cells remained relatively stable at 3, 4, and 8 wk (0.2 × 10⁹ cells/L, 0.5 × 10⁹ cells/L, and 0.4 × 10⁹ cells/L, respectivly). Thirty-eight weeks after onset of symptoms, CD8⁺CD45RA^high cells were purified by negative sorting from PBMC and stimulated with autologous irradiated PBMC pulsed with peptide 56 in LDA. The frequency of peptide 56-specific CTLp in CD8⁺CD45RA^high cells was 1240 per 10⁶ CD8⁺ T cells (frequency against control target <20 per 10⁶ CD8⁺CD45RA^high T cells).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

CD45RO and RA expression on CD8⁺ T cells and fresh ex vivo peptide-specific cytotoxicity in primary HCMV infection in donor 104. A, D, and G, The percentage of CD8⁺ T cells in fresh PBMC at 3, 4, and 8 wk after onset of symptoms. B, E, and H, The distribution of CD45RA and CD45RO on CD8⁺ T cells in fresh PBMC. C, F, and I, Fresh ex vivo cytotoxicity in PBMC at an E:T ratio of 40:1 against autologous target cells pulsed with pp65 peptides 69, 31, and 56, and unpulsed.

A second patient (102) with primary HCMV infection was studied 3 wk after the onset of symptoms. At this time point, there was a large CD8⁺ lymphocytosis, with a high percentage of these cells being CD45RO⁺. There was also fresh ex vivo cytotoxicity against pp65 peptide-pulsed target cells. Eighteen weeks after onset of symptoms, PBMC obtained from patient 102 were sorted into total CD8, CD8⁺CD45RA^high, and CD8⁺CD45RO^high T cell populations, and the peptide-specific CTLp frequencies were determined for each population. Eighteen weeks after onset of symptoms, pp65 peptide-specific CTL were detected in both the CD45RA^highCD45RO⁻ and CD45RO^highCD8⁺ T cell subpopulations (data not shown).

Longitudinal analysis of pp65-specific clonotypes following primary HCMV infection shows clonal reversion from CD45RO^high to CD45RA^high T cells

Multiple independent pp65 peptide 56-specific CTL clones were derived from the PBMC of patient CMV104 4 wk after onset of symptoms, and the Vβ TCR usage of these clones was determined by PCR analysis with a panel of Vβ TCR primers, as described (16). A large proportion of the clones was found to utilize TCR Vβ 6.4. A clonotypic probe specific to the n-D-n region from one of these clones was used to quantitate the size of this pp65 peptide-specific CTL clone within the CD45RA^high and CD45RO^high CD8⁺ T cell populations at 3, 4, 8, and 37 wk after onset of symptoms. The results show that at 3 wk after onset of symptoms, there was very little of the clonotype within the Vβ 6.4⁺ CD45RA^high cell population as compared with the Vβ 6.4⁺ CD45RO^high population (Fig. 5⇓). Four weeks after onset of symptoms there was still a pronounced difference in the amount of clone detected in the CD45RA^high as compared with the CD45RO^high populations (Fig. 5⇓). At 8 wk after onset, similar or higher levels of the clonotype were detected in the Vβ 6.4⁺ CD45RA^high as compared with the Vβ 6.4⁺ CD45RO^high population. At 37 wk after onset of symptoms, the clonotype was distributed in both the CD45RA^high and CD45RO^high T cell populations (Fig. 5⇓), as seen previously in the long-term virus carriers (Table III⇑, Fig. 3⇑). As a mAb to TCR Vβ 6.4 is not commercially available, we were unable to calculate the absolute clone size within the CD45RA^high and CD45RO^high populations at each of the time points sampled.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Quantitation of an individual peptide 56-specific clone within CD45RA^high and CD45RO^high CD8⁺ T cell subpopulations derived from PBMC of donor 104 during and after primary HCMV infection. A, Duplicate TCR Vβ 6.4 PCR products were amplified from each of the T cell populations taken at 3, 4, 8, and 37 wk after onset of symptoms. PCR product derived from the biological pp65-specific clone (positive control) and pooled PCR product derived from PBMC of four HCMV-seronegative donors (negative control) were probed with a radiolabeled clonotypic probe. B, The filter was stripped of bound clonotypic probe and reprobed with a radiolabeled probe specific to the TCR C region. For each subpopulation, the clonotype sequence as a percentage of all TCR Vβ 6.4⁺ sequences is shown for two independent amplifications and probing experiments; % in bold represents the gel shown. ∗, Duplicate PCR sample failed.