Detection and Quantification of CD4⁺ T Cells with Specificity for a New Major Histocompatibility Complex Class II-Restricted Influenza A Virus Matrix Protein Epitope in Peripheral Blood of Influenza Patients

Abstract

FVFTLTVPS was identified as the core sequence of a new major histocompatibility complex class II-restricted T-cell epitope of influenza virus matrix protein. Epitope-specific CD4⁺ T cells were detected in the peripheral blood of patients with frequencies of up to 0.94%, depending on the number of additional terminal amino acids.

PM19, a CD4⁺ T-cell clone established from an influenza patient, was found to respond to a peptide corresponding to the amino acid positions 55 through 73 of the matrix protein of influenza viruses. To establish the exact sequence of the epitope, synthetic peptides scanning the entire MP_55–73 region were synthesized and tested for their capacity to induce cytolysis of autologous Epstein-Barr virus (EBV)-transformed B-lymphoblastoid target cells by PM19 in a standard ⁵¹Cr release assay with PEW-EBV as targets and an effector-to-target ratio of 5:1. The minimal peptides that triggered some, albeit low, response were the nonapeptides MP_62–70 (FVFTLTVPS) and MP_63–71 (VFTLTVPSE) (Table 1), of which MP_62–70 conforms better to the sequence motif for peptides presented by HLA-DR4 (25), which had been identified as a restricting major histocompatibility complex (MHC) allomorph (Table 2). Only HLA-DR4-expressing B-lymphoblastoid cell lines were lysed. This cytolysis was inhibited by 200 μg of the monoclonal antibody L243 specific for HLA-DR/ml (15). Maximal responses were obtained with MP_62–72 (FVFTLTVPSER), MP_61–72 (GFVFTLTVPSER), and MP_60–73 (LGFVFTLTVPSERG). MP_62–70 appears to be the core sequence of the new epitope, fitting exactly into the peptide binding groove of the MHC molecule with the additional amino acids extruding from the ends of the groove (Table 3). Further extensions resulted in reduced activity, and truncation of the minimal peptides resulted in a complete loss of activity. MP_55–73, which had been used to establish the clone, induced a weak but reproducible response.

TABLE 1.

Peptides used to identify the influenza virus matrix protein epitope MP_62–70 and the reaction profiles of the CD4⁺ T-cell clone PM19 and of peripheral blood CD4⁺ T cells of the HLA-DR4⁺ patients ANP and KIM

Positions	Peptide (MP)a																			Responseb
																				PM19c	T cells of patientd:
	Sequence																				KIM	ANP
	55	56	57	58	59	60	61	62	63	64	65	66	67	68	69	70	71	72	73
55–73	L	T	K	G	I	L	G	F	V	F	T	L	T	V	P	S	E	R	G	+	ND	ND
56–66		T	K	G	I	L	G	F	V	F	T	L								−	ND	ND
56–67		T	K	G	I	L	G	F	V	F	T	L	T							−	ND	ND
56–68		T	K	G	I	L	G	F	V	F	T	L	T	V						−	ND	ND
56–69		T	K	G	I	L	G	F	V	F	T	L	T	V	P					−	ND	ND
56–70		T	K	G	I	L	G	F	V	F	T	L	T	V	P	S				++	ND	ND
57–66			K	G	I	L	G	F	V	F	T	L								−	ND	ND
57–68			K	G	I	L	G	F	V	F	T	L	T							−	ND	ND
57–68			K	G	I	L	G	F	V	F	T	L	T	V						−	ND	ND
57–69			K	G	I	L	G	F	V	F	T	L	T	V	P					−	ND	ND
58–66*				G	I	L	G	F	V	F	T	L								−	ND	ND
58–67				G	I	L	G	F	V	F	T	L	T							−	ND	ND
58–68				G	I	L	G	F	V	F	T	L	T	V						−	ND	ND
58–69				G	I	L	G	F	V	F	T	L	T	V	P					−	ND	ND
59–68					I	L	G	F	V	F	T	L	T	V						−	ND	ND
59–69					I	L	G	F	V	F	T	L	T	V	P					++	ND	ND
59–70					I	L	G	F	V	F	T	L	T	V	P	S				++	ND	ND
59–71					I	L	G	F	V	F	T	L	T	V	P	S	E			+	ND	ND
59–72					I	L	G	F	V	F	T	L	T	V	P	S	E	R		++	ND	ND
59–73					I	L	G	F	V	F	T	L	T	V	P	S	E	R	G	++	ND	ND
60–70						L	G	F	V	F	T	L	T	V	P	S				++	++	+
60–71						L	G	F	V	F	T	L	T	V	P	S	E			++	+	+
60–72						L	G	F	V	F	T	L	T	V	P	S	E	R		+	+++	+
60–73						L	G	F	V	F	T	L	T	V	P	S	E	R	G	+++	+	+
61–70							G	F	V	F	T	L	T	V	P	S				+	+	−
61–71							G	F	V	F	T	L	T	V	P	S	E			+	+++	+
61–72							G	F	V	F	T	L	T	V	P	S	E	R		+++	++++	−
61–73							G	F	V	F	T	L	T	V	P	S	E	R	G	++	+	−
62–70								F	V	F	T	L	T	V	P	S				+	+	−
62–71								F	V	F	T	L	T	V	P	S	E			++	+	+
62–72								F	V	F	T	L	T	V	P	S	E	R		+++	++	++
62–73								F	V	F	T	L	T	V	P	S	E	R	G	++	+	+
63–70									V	F	T	L	T	V	P	S				−	+	+
63–71									V	F	T	L	T	V	P	S	E			+	+	−
63–72									V	F	T	L	T	V	P	S	E	R		+	−	+
63–73									V	F	T	L	T	V	P	S	E	R	G	+	++	+

^aThe asterisk indicates the HLA-A*0201-restricted CTL epitope. Boldfacing indicates the core sequences of the T-cell epitopes. 

^bThe clone PM19 was tested in a standard ⁵¹Cr release assay (see Table 2); the peripheral blood T cells were monitored using flow cytometry for detection of intracellular cytokines. 

^c−, none; +, 0 to 9% net lysis; ++, 10 to 19% net lysis; +++, 20 to 30% net lysis at a peptide concentration of 1 ng/ml. 

^d−, none; +, 0 to 0.25%; ++, 0.26 to 0.50%; +++, 0.51 to 0.75%; ++++, 0.76 to 1.00% peptide-reactive cells among the CD4⁺ cells. 

TABLE 2.

Identification of HLA-DR4 as a restriction element for the recognition of matrix protein-derived peptides by the CD4⁺ T-cell clone PM19^a

Cell line or donor	HLA haplotype					% specific 51Cr release
	A	B	C	DR	DQ
PEW-EBV	2.1; 32	62; 70	w3	4	3	23.5
JY	2.1	7.1		4; 6		29.5
T2	2.1	5				2.3
JY + L243						5.8
Patients
ANP	2; 24	27; 44	w2; w5	2; 4; 53	1; 3
KIM	11	35; 51	w4	1; 4; 53	1; 3

^aThe cytolysis of the target cells PEW-EBV (genetically identical with PM19), JY, and T2 by PM19 was analyzed in a standard ⁵¹Cr release assay using MP_60–73 as the antigen. The HLA alleles were determined serologically. HLA-DR4 of PEW, cells from ANP, and cells from KIM were retyped by PCR analysis and sequencing and were found to be DRB1*0401 in the case of PEW and ANP and DRB1*0404 in the case of KIM. 

TABLE 3.

Anchor residues for the HLA-A0201- and DR4-restricted T-cell epitopes

Subject	Sequence positions and corresponding amino acid residuesa
Influenza virus matrix protein sequence position	58	59	60	61	62	63	64	65	66	67	68	69	70
DR4-restricted matrix protein T-cell epitope
Relative epitope positions					1	2	3	4	5	6	7	8	9
Core sequence					F	V	F	T	L	T	V	P	S
Anchor or preferred residues for DRB1*0401b					F, Y, W, I, L, V, M			F, W, I, L, V, A, D, E; no R, no K		N, S, T, Q, H, R	pol.cchg.cali.c		pol., ali., K
Anchor or preferred residues for DRB1*0404b					V, I, L, M			F, Y, W, I, L, V, M, A, D, E; no R, no K		N, T, S, Q, R	pol., chg., ali		pol., ali, K
HLA-A*0201-restricted T-cell epitope
Relative epitope positions	1	2	3	4	5	6d	7	8	9
Core sequence	G	I	L	G	F	V	F	T	L
Anchor or preferred residues for HLA-A*0201b		L, M				V			V, L

^aBoldfacing of numbers and letters indicates primary anchor positions and amino acids. 

^bAccording to Rammensee and colleagues (25). 

^cpol., polar; chg., charged; ali., aliphatic. 

^dSecondary anchor residues. 

Depending on its length, the sequence of the MHC class II-restricted epitope overlaps by five to seven amino acids the MHC class I-restricted MP_58–66 sequence (GILGVFVTL) (8, 19) (Table 3). This is one of only four cases of overlapping MHC class I- and class II-restricted T-cell epitopes reported so far. The others were found in the human immunodeficiency virus (HIV) glycoprotein 160 (26), in influenza A virus nucleoprotein (6), and for a mutation of the ras p21 oncogene (1). The low number of overlapping MHC class I- and class II-restricted epitopes can be attributed to the different processing pathways of antigens restricted by the two MHC classes (3). In the case of virus infections, both pathways can participate in the processing, since viral antigens are both internally expressed and external in the form of virus particles. In addition, it has been suggested that long-lived cytosolic antigens can be processed for presentation by MHC class II molecules through an alternative pathway (4, 10, 13, 17, 22). Brefeldin A rapidly inhibits this route (11), so that binding by recycled HLA-DR molecules, as for other exogenous antigens, can be excluded (18, 21, 23).

MP_62–70 is derived from a particularly conserved region of influenza virus matrix protein. Only eight mutations are documented for this epitope with the published 188 sequences, which cover more than 60 years of molecular evolution of this highly variable virus (http://www.flu.lanl.gov). These exchanges are valine to cysteine at position 68 (four times), proline to threonine at position 69 (once), serine to glycine at position 70 (once), and arginine to glutamine at position 72 (twice). They are less conservative than those found for the MHC class I-restricted MP_58–66, for which eight mutations are also described: isoleucine to valine (seven times) and isoleucine to leucine (once) at position 59. The high degree of conservation in this region of the matrix protein might reflect constraints associated with the structural function of the molecule in the influenza virus particle and thus might indicate a suitable target structure for the development of synthetic vaccines.

The epitope was verified by analyzing the response of peripheral blood CD4⁺ T cells to the peptides delineated in Table 1 by flow cytometric detection of cells induced to produce interleukin-2 (14) using the phycoerythrin-labeled monoclonal antibody 5344.111 (Becton Dickinson, Heidelberg, Germany). Peripheral white blood cells of two influenza patients, KIM and ANP, were isolated 1 week and 4 weeks, respectively, after the onset of influenza symptoms. Cells from both donors expressed DR4: HLA-DRB1*0401 in the case of ANP as clone PM19 and HLA-DRB1*0404 in the case of KIM. Epitope-specific CD4⁺ T cells were detected for both donors (Table 1). Depending on the peptides, their frequencies ranged from 0.12% for peptide MP_63–70 to 0.94% for MP_61–72 in donor KIM and between 0.03% for MP_60–70 and MP_61–71 and 0.32% for MP_62–72 in ANP. The most potent peptide for ANP cells was MP_62–72, which is also optimal for clone PM19. T cells from KIM responded best to MP_61–72, which was not recognized by cells from ANP at all (Table 1). These differences might indicate T-cell receptor repertoire differences between the two individuals. Figure 1 presents the histograms which show the repertoire differences between the two donors. The frequencies detected for ANP were lower overall than those for KIM, which might reflect different immune response capacities but which could also be due to the different time periods that had passed since the onset of disease. The specificity differences showed no correlation with the different DR4 β-chains expressed by the donors, although the binding motifs for the two DR4 allomorphs differ slightly (25), but this seems to be caused by differences in the individual T-cell receptor repertoires. Obviously, the amino acids extruding from the ends of the peptide binding groove influence the magnitude of the response of a single T-cell clone as well as the number of responding T-cell clones.

FIG. 1.

Responses of peripheral blood CD4⁺ T cells of the influenza patients KIM and ANP to peptides that include the core sequence of the HLA-DR4-restricted T-cell epitope MP_62–70. The specific T cells were identified by intracellular staining for interleukin-2 and were quantified by flow cytometry. Numbers on the abscissa indicate fluorescence intensity; those on the ordinate indicate the number of events.

The frequencies reported here for MP-specific CD4⁺ T cells are two to three orders of magnitude higher than the frequencies reported for CD4⁺ T cells specific for the influenza virus epitope HA_307–319 (20) but are in a range similar to that described for HIV or cytomegalovirus (24, 28, 29). MHC class I-restricted T cells specific for single viral epitopes, on the other hand, were found with frequencies ranging up to 2% for HIV infections (7, 9), 15% for cytomegalovirus (14; Sherev, unpublished observations), 1.2% for hepatitis C virus in peripheral blood or 2% in liver (12), and an extreme of 44% for a case of acute EBV infection (2, 5, 16, 27).