Restricted Isotypic Antibody Reactivity to Hepatitis C Virus Synthetic Peptides in Immunocompromised Patients

Abstract

An enzyme immunoassay based on three synthetic peptides from the core, NS4, and NS5 regions of hepatitis C virus allowed the detection of antibodies in 100% of immunocompetent infected patients and in 91% of immunocompromised patients (hemodialysis and hemophiliac patients). Immune impairment seemed to restrict the spectrum of antibody isotypes reacting to the core peptide.

Synthetic peptides have proven to be valuable tools for immunodiagnosis of hepatitis C virus (HCV) infection (8, 10, 18). Most of the available serological tests for HCV are based on the combined use of synthetic peptides and recombinant proteins. In a previous study, restricted antibody reactivity to HCV peptides was observed in HCV-infected hemodialysis patients, although some immunodominant epitopes in the core and NS4 regions were still consistently recognized (5). The aim of this study was to design an immunodiagnostic system for HCV based exclusively on synthetic peptides and suitable for epidemiological studies. In addition, isotypic antibody reactivity to these peptides was evaluated and compared between immunocompromised and immunocompetent HCV-infected patients.

A total of 105 sera from HCV-infected hemodialysis patients previously tested for viral hepatitis serological markers (16) were used in this study. Hemophiliac patient sera (24 hemophilia A patients) were also tested. Both hemophiliac and hemodialysis patients were defined as immunocompromised HCV-infected patients in this study. HCV-positive blood donors and other HCV-infected patients (n = 66) without any evident cause of immunosuppression were also included. Information about the genotypes of HCV circulating in 51 of these patients (17) was available: genotypes 1a, 1b, 2a, 2b, 3a, and 4 (14, 15, 8, 6, 1, and 1 patients, respectively) and six mixed infections were present in the patients analyzed. All patients were human immunodeficiency virus (HIV) negative except for one hemophiliac patient. A total of 96 sera from blood donors, negative for hepatitis B virus (HBV), HCV, and HIV markers, were also tested.

Core peptides (716 [20-mer], RNTYRRPQDVKFPGGGQIVG [amino acids {aa} 13 to 32]; 286 [40-mer], STNPKPQRKTKRNTNRRPQDVKFPGGGQIVGGVYLLPRRG [aa 2 to 41]), NS4 peptides (59 [20-mer], AFASRGNHVSPTHYVPESDA [aa 1920 to 1941]), and NS5 peptides (290 [35-mer], RKFPPALPIWARPDYNPPLLESWKDPDYVPPVVHG [aa 2279 to 2313]) were manually synthesized (9, 13) and were based on the consensus sequence of the HCV database (14). The immunodominant antigenic region NS3f (2) was modeled with long peptides (aa 1359 to 1408, 50-mer; aa 1409 to 1448, 40-mer; aa 1449 to 1488, 40-mer), but no significant reactivity was obtained (data not shown). All peptides were synthesized as monomers and polymers, as previously described (15).

Antibody reactivity to the peptides was tested by enzyme immunoassay, as previously described (5). Whether tested singly or as a mixture, 1 μg of each peptide was adsorbed to wells. Human sera were tested at a 1/40 dilution in 1% skim milk in phosphate-buffered saline by using for detection a 10⁻⁵ dilution of peroxidase-labelled goat anti-human immunoglobulin G (IgG) (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) in the same buffer. Subclass-specific antibodies to each peptide were tested at the same serum dilution by using anti-isotype-specific monoclonal antibodies conjugated to biotin (Sigma Chemical Co., St. Louis, Mo.), each at a pretested optimal dilution, and streptavidin-peroxidase (1/1,000) (Sigma). TMB (3,3′,5,5′-tetramethylbenzidine) (Bio-Rad Laboratories) was used as a substrate. Cutoff values were determined as means of negative serum values plus 3 standard deviations. Statistical differences were evaluated by the chi square and/or Fisher exact test with Yates’ correction, according to a computerized Epi Info program, version 5.01b (Centers for Disease Control and Prevention, Atlanta, Ga.). Average statistical differences were evaluated by the Student t test.

The antibody reactivity to a 40-mer peptide (peptide 286) designed from the N-terminal region of the HCV core protein was significantly higher than that directed to a 20-mer peptide (peptide 716) derived from a region located inside the 40-mer peptide and previously found to be the most frequently recognized in this region (5). Of 35 hemodialysis patient sera tested, 31 (89%) recognized peptide 286 while only 24 (69%) reacted with peptide 716, and a significantly higher optical density (OD) was observed for 14 patients with peptide 286 (Table 1). The use of longer peptides in the core region has previously been suggested by others to be more effective (21). In the region shared by the two peptides, an amino acid change was introduced in the 40-mer peptide, 286, since it was more frequently found in the sequences of the different HCV isolates; this may have contributed to the increased reactivity observed with peptide 286. The length and conservation of peptide 286 ensured appropriate reactivity of sera of patients infected with all of the genotypes tested in this study (data not shown).

TABLE 1.

Enhanced reactivity to a 40-mer core peptide by HCV-infected hemodialysis patient sera (n = 35)

Parameter	Sera nonreacting with peptides	Sera reacting only with 286	Higher-OD recognition of 286a	Similar recognition of both peptides
No. (%) of patients	4 (11)	7 (20)	7 (20)	17 (49)
Mean ODb
Peptide 716		0.155	0.505	0.905
Peptide 286		0.720	1.155	1.044

^aMeans of triplicate OD values for each sample were compared with Student’s t test. 

^bMean OD of serum samples. 

No significant increase in reactivity was detected by using polymeric peptides instead of monomers (data not shown). When a mixture of peptides 286, 59, and 290, derived from the core, NS4, and NS5, respectively, was evaluated as a potential diagnostic antigen for HCV infection, high sensitivity and specificity were obtained with immunocompetent patients (66 of 66 patients recognized [100%] and 1 of 96 reactive among negative patient sera [99% specificity]). The mixture, however, was recognized by only 91% of the immunocompromised HCV-infected patients (118 of 129). The presence of an NS3 antigen in the cocktail used may be critical for adequate recognition of this particular group of patients (3–5). Nevertheless, the peptide cocktail analyzed in this study seems to be useful for epidemiological studies in the general population.

The NS5 region was included, as it has been suggested that the use of NS5 antigens may add sensitivity to diagnostic tests (20). In fact, some HCV-infected hemodialysis patient sera from the group studied by Devesa et al. (5) reacted only with peptide 232 from the NS5 region, which is comprised in peptide 290 used in this study (data not shown). A small portion of the 290 peptide is well conserved among different HCV isolates (aa 2288 to 2297). This might explain the frequent recognition of peptide 290 by the sera of patients tested in this study, irrespective of the infecting genotype (data not shown).

The 20-mer peptide derived from NS4 (peptide 59) and included in the cocktail contains a sequence highly conserved among different HCV isolates, which might be one of the reasons for its high frequency of recognition (5, 10, 11). Antibodies against peptide 59 seem to be raised at early stages of infection, with some degree of antigenic cross-reactivity with proteins from different origins (24). In our study, the use of peptide 59 in the cocktail did not produce nonspecific reactions. The only false-positive result was due to recognition of the core peptide, where some nonspecific reaction has also been described (23).

The antibody subclass reactivity was assayed for each of the three peptides from the core, NS4, and NS5. For the latter two peptides, the isotypic reactivity was mostly of the IgG1 subtype, a reactivity similar to that observed in region NS4a (20, 25). No significant differences were found in the isotypic reactivities between immunocompetent and immunocompromised patients (data not shown). Only in two hemodialysis patient sera was the highest reactivity, to both NS4 and NS5 peptides, of the IgG3 subclass (data not shown).

For the core peptide, IgG1 reactivity was observed in all patients. Antibodies of IgG3, IgG4, and/or IgA were found in 44% of immunocompetent HCV-infected patients but in significantly fewer immunocompromised patients, particularly in HBV-infected patients (Table 2). Reactivity to the core region has previously been shown to be predominantly by IgG1 antibodies, although all other isotypes, particularly IgG3, have been detected (6, 19, 22). A similar pattern was found for the 40-mer peptide 286, although IgG3 and IgG4 were present at similar frequencies (Table 2).

TABLE 2.

Isotype reactivity to HCV peptides

Peptide	HCV-infected patient characteristic (n)	No. (%) of sera containing antibody of isotype:
		IgG1	IgG2	IgG3	IgG4	IgM	IgA	Non-IgG1/IgM
59	HCV+ (70)a	65 (93)	10 (14)	5 (7)	0	25 (36)	8 (11)	20 (29)
290	HCV+ (102)b	99 (97)	8 (8)	8 (8)	3 (3)	1 (1)	3 (3)	17 (17)
286	Immunocompromised HCV+ (52)	52 (100)	2 (4)	6 (12)	1 (2)	5 (10)	0	9 (17)
	Hemophiliac (15)	15 (100)	0	0	0	2 (13)	0	0
	Hemodialysis HBsAg+ (10)c	10 (100)	1 (10)	0	0	1 (10)	0	1 (10)
	Hemodialysis HBsAg− (27)	27 (100)	1 (4)	6 (22)	1 (4)	2 (7)	0	8 (30)
	Immunocompetent HCV+ (59)	59 (100)	3 (5)	11 (19)	12 (20)	13 (22)	9 (15)	26 (44)
Significanced		NS	NS	NS	<0.01	NS	<0.01	<0.01

^aA total of 28 HCV⁺ immunocompetent patients and 42 HCV⁺ immunocompromised patients (hemophiliac and hemodialyzed patients) were evaluated. 

^bA total of 52 immunocompetent and 50 immunocompromised patients were evaluated. 

^cHBV-coinfected hemodialyzed patients. HBsAg, hepatitis B surface antigen. 

^dComparison of reactivities to peptide 286 between all immunocompromised and immunocompetent HCV-infected patients. NS, not significant (P > 0.05). 

The intrinsic immune deficit in hemodialysis patients has already been described (7), and acquired immune impairment has been reported for hemophiliac patients (1, 12). The restricted isotypic response to peptide 286 in both hemodialysis and hemophiliac patients may be due to some degree of impairment of the immune function. Among hemodialysis patients, this effect was more pronounced in HBV-coinfected patients, who were indeed the group with more-reduced reactivity to HCV peptides (5). These results stress the need for highly sensitive diagnostic tools for immunocompromised patients.