Evaluation of Alloreactivity in Responder-Stimulator Pairs by Determination of Gamma Interferon-Producing Cells and Cytotoxic-T-Lymphocyte Precursor Frequencies

Abstract

We used the enzyme-linked immunospot (ELISPOT) assay and the cytotoxic-T-lymphocyte precursor frequency assay to evaluate alloreactivity in responder-stimulator pairs. High frequencies of responder cells among cells from HLA-mismatched pairs and low frequencies among cells from pairs of siblings with identical HLA types were detected by both assays. The ELISPOT assay thus illustrated the helper and cytotoxic-T-cell response to allogeneic HLA antigens.

Histocompatibility before transplantation is determined by serological or molecular methods. These techniques define HLA antigens and alleles shared between the donor and the recipient but do not provide any functional information from which complications after allografting can be predicted. Perfect HLA matching, in fact, does not ensure the absence of rejection or graft-versus-host disease (1, 5), while certain HLA mismatches do not seem to preclude successful transplantation (2, 4). In vitro functional methods, namely, the helper precursor frequency and cytotoxic-T-lymphocyte precursor frequency (CTL-p-f) assays, have been used to measure the strengths of alloreactions promoted by T cells. Yet these assays are difficult to perform and require a large number of cells and radionuclides. New methods for determining T-cell functions include the enzyme-linked immunospot (ELISPOT) assay, which measures cytokine-producing cells. We have evaluated the impact of HLA matching on the immune response with the gamma interferon (IFN-γ) ELISPOT assay by using peripheral blood mononuclear cells (PBMC) isolated from pairs of healthy donors designated as responders (R) and stimulators (S). The alloreactivities of cells from four HLA-mismatched (HLA-MM) pairs, four pairs of siblings with identical HLA types (HLA-IS), and two monozygotic twin (MZ) pairs were assessed. The CTL-p-f assay, already set up and used in the past in our laboratory, was performed in parallel. Healthy donors were typed for HLA-A, HLA-B, and HLA-DR antigens by serology. HLA-A*, HLA-B*, and HLA-DRB1* loci were typed by PCR with a sequence-specific primer. PBMC for CTL-p-f and ELISPOT assays were separated from heparinized peripheral blood by using Histopaque-1077 (Sigma-Aldrich, St. Louis, MO) density gradient centrifugation and immediately used as fresh cells.

ELISPOT assay.

R PBMC were seeded in U-bottom, 96-well plates in a limiting dilution in triplicate and cocultured with fixed doses of irradiated S PBMC. After 30 h (or, in some experiments, 54, 78, 102, or 126 h), cells were transferred into ELISPOT plates coated with an anti-IFN-γ monoclonal antibody (mAb) for 18 h at 37°C with 5% CO₂. Cells were then discarded, and a biotinylated anti-IFN-γ mAb was added. Then streptavidin-peroxidase was added and spots were revealed with a chromogenic substrate. All reagents were purchased from Mabtech (Stockholm, Sweden). Plates were examined with an automatic reader, and results were shown as the number of IFN-γ-producing cells (IFN-γ-pc)/10⁶ R cells. Experiments were performed with RPMI 1640 containing fetal calf serum (FCS) or human AB serum. In ELISPOT HLA-blocking experiments, S cells from two HLA-MM pairs were incubated with 40 μg/ml of mouse anti-human HLA class I and II mAb for 15 min at 4°C and then incubated with R cells.

CTL-p-f assay.

Limiting numbers of R PBMC (5 × 10⁴ to 3.125 × 10³) were cocultured in 96-well, U-bottom plates with 5 × 10⁴ irradiated (30 Gy) S PBMC. The medium was RPMI 1640 with human AB serum. Twenty replicates for each R cell dilution were set up. As a negative control, irradiated S cells were incubated without R cells. On day 3, cultures were fed with interleukin-2. On day 7, OKT3 mAb-induced blasts obtained from the original S cells were incubated with ⁵¹Cr for 1 h and added to the cultures. Plates were centrifuged and incubated for 4 h. Then, 50 μl of supernatants was harvested and transferred onto solid-scintillator plates. ⁵¹Cr in supernatants was evaluated by a β-radiation counter. Wells were scored as positive when the count was higher than the mean plus 3 standard deviations of the negative control. All data were processed using software that gave the CTL-p-f. Results were shown as the number of cytotoxic-T-lymphocyte precursors/10⁶ R cells.

In preliminary experiments performed with cells from an HLA-MM pair, increasing irradiation doses (30, 45, and 60 Gy) inhibited cytokine secretion but reduced the stimulatory capability of S cells (data not shown). To reveal cytokines secreted mainly from R cells, 30-Gy-irradiated S cells were seeded at 2 × 10⁵, 1 × 10⁵, and 5 × 10⁴ in the same final volume (200 μl) with 2 × 10⁵ R cells in a limiting dilution. The most appropriate concentration of S cells to obtain a surrogate one-way reaction proved to be 5 × 10⁴ cells/200 μl. The number of spots, in fact, titrated linearly (r² = 0.99) with the R cell concentration, suggesting a limited contribution from S cells in secretion. Spot values corresponding to the first R cell concentration (2 × 10⁵ R cells/200 μl) did not differ significantly (Grubb's test; P > 0.05) among the three R/S cell ratios (962 ± 236, 985 ± 146, and 858 ± 104). This result suggests that the stimulatory capability was maintained.

An ELISPOT assay of three R-S pairs was then performed for 48 h with these parameter values. Results showed strong alloreactivity between the cells from an HLA-MM pair and few spots corresponding to the HLA-IS and MZ pairs (data not shown). The kinetics of cytokine secretion were determined, and cells from two HLA-IS pairs (pairs 6 and 8) were cultured for 48, 72, 96, 120, and 144 h to show alloreactivity. FCS elicited IFN-γ secretion from unstimulated R cells after 48 h, whereas no secretion from cells in human AB serum was observed. Human serum was then used for all experiments, and ELISPOT and CTL-p-f assays of 10 R-S pairs were performed. In ELISPOT assays, HLA-MM R and S cells were incubated for 48 h and HLA-IS and MZ cells were incubated for 144 h. Results showed high frequencies of IFN-γ-pc among cells from HLA-MM pairs (mean for the four pairs, 371 IFN-γ-pc/10⁶ R cells), and these frequencies decreased when anti-HLA mAbs were added to the S cells (Fig. 1). Among cells from HLA-IS (mean, 8 IFN-γ-pc/10⁶ R cells) and MZ (mean, 3.5 IFN-γ-pc/10⁶ R cells) pairs, low frequencies of IFN-γ-pc were found. High CTL-p-f among cells from HLA-MM pairs (mean, 61 cytotoxic-T-lymphocyte precursors/10⁶ R cells) and low frequencies among cells from HLA-IS (mean, 6 cytotoxic-T-lymphocyte precursors/10⁶ R cells) and MZ (mean, 2 cytotoxic-T-lymphocyte precursors/10⁶ R cells) pairs were found. Results are summarized in Table 1.

FIG. 1.

IFN-γ-pc among cells from pairs 1 and 2 as assessed by using S cells treated with (+) anti-HLA (α-HLA) class I and II mAbs.

TABLE 1.

IFN-γ-pc and CTL-p-f in 10 R-S pairs^a

Pair no.	Pair type	Cell type	Results of HLA typing	Gender	No. of IFN-γ-pc/106 R cells ± SD	No. of CTL precursors/106 R cells
1	HLA-MM	R	A2,3; B13,50(21); DRB1* 03,14	F	172 ± 57	83
		S	A29(19); B44(12),63(15); DRB1* 07,11	F
2	HLA-MM	R	A2,3; B44(12),49(21); DRB1* 02,11	F	670 ± 78	71
		S	A2,24(9); B8,51(5); DRB1* 03,07	F
3	HLA-MM	R	A24(9),30(19); B35,49(21); DRB1* 15,01	M	275 ± 73	20
		S	A1,24(9); B61(40),63(15); DRB1* 13,14	M
4	HLA-MM	R	A29(19),30(19); B53,39(16); DRB1* 11,13	M	365 ± 61	71
		S	A1,24(9); B61(40),63(15); DRB1* 13,14	M
5	HLA-IS	R	A*24,25; B*35,13; DRB1* 07,11	F	7 ± 7.6	6
		S	A*24,25; B*35,13; DRB1* 07,11	M
6	HLA-IS	R	A*01,02; B*07,50; DRB1* 1101,07	M	20 ± 18	10
		S	A*01,02; B*07,50; DRB1* 1101,07	M
7	HLA-IS	R	A*29,68; B*39,44; DRB1* 07,07	F	4 ± 2.9	4
		S	A*29,68; B*39,44; DRB1* 07,07	F
8	HLA-IS	R	A*01,24; B*15,40; DRB1* 13,14	F	0	3
		S	A*01,24; B*15,40; DRB1* 13,14	M
9	MZ	R	A*02,30; B*13,18; DRB1* 07,11	F	4 ± 20	2
		S	A*02,30; B*13,18; DRB1* 07,11	F
10	MZ	R	A*02,30; B*18; DRB1* 01,03	F	3 ± 5.8	2
		S	A*02,30; B*18; DRB1* 01,03	F

^aAsterisks indicate HLA alleles. F, female; M, male; CTL, cytotoxic T lymphocyte.

The ELISPOT assay measures T-cell function and may be a useful aid in assessing the prognosis of a transplant. To illustrate the immune response mainly on the part of R cells, this being closer to the T-cell response after hematopoietic stem cell or solid organ allografting, S cells were irradiated with 30 Gy and the R/S cell ratio was increased. More precise methods employ purified non-T cells as S cells (3), but these cells are often difficult to obtain in sufficient numbers, especially from oncohematological patients. Our approach is a trade-off between the maintenance of stimulation capability and the conditions for a one-way reaction and the collection and holding of a larger quantity of cells for repeated determinations, such as immunological follow-up.

The determination of the kinetics of IFN-γ secretion disclosed a nonspecific production of cytokine probably due to factors in the FCS. Human AB serum did not elicit this nonspecific increase of IFN-γ-pc and would seem preferable in prolonged cultures before ELISPOT analysis. High frequencies of IFN-γ-pc among cells from HLA-MM pairs and lower frequencies among cells from HLA-IS and MZ pairs were found. The CTL-p-f assay showed a similar pattern of alloreactivity detection. The determined R cell frequencies, especially among cells from HLA-MM pairs, were lower in the CTL-p-f assay than in the ELISPOT assay. This result could be due to different sensitivity thresholds as well as the kinds of responses detected by the two techniques (direct IFN-γ secretion by R cells as detected by the ELISPOT assay versus the cytotoxic response after the activation and expansion of cytotoxic-T-lymphocyte precursors present at the beginning of the culture). ELISPOT analysis may also reveal different T-cell populations (helper and cytotoxic) secreting IFN-γ. The incubation of S cells with anti-HLA class I or II mAb decreased the number of IFN-γ-pc. Hence, the spots detected may represent both CD8 and CD4 IFN-γ-secreting T lymphocytes recognizing allogeneic HLA molecules. In conclusion, the ELISPOT assay may be a useful method for detecting alloreactivity elicited by HLA mismatches in R-S pairs. IFN-γ ELISPOT analysis is simpler than the CTL-p-f assay and may reveal both the cytotoxic and the T-helper response. It can thus be proposed for the detection of alloreactivity before and after allografting as a means of predicting immunological complications.