Correlations among measles virus-specific antibody, lymphoproliferation and Th1/Th2 cytokine responses following measles–mumps–rubella-II (MMR-II) vaccination

Abstract

Immunity to measles is conferred by the interplay of humoral and cellular immune responses, the latter being critical in maintaining long-term recall response. Therefore, it is important to evaluate measles-specific humoral and cellular immunity in populations several years after vaccination and understand the correlations among these measures of immunity. We examined measles-specific antibodies, lymphoproliferation and the Th1/Th2 signature cytokines, interferon (IFN)-γ and interleukin (IL)-4, in a population-based cohort of healthy children from Olmsted County, Minnesota after two doses of measles–mumps–rubella-II (MMR-II) vaccine. We detected positive measures of measles-specific cellular and humoral immunity in the majority of our study population. However, a small proportion of subjects demonstrated an immune response skewed towards the Th2 type, characterized by the presence of either IL-4 and/or measles-specific antibodies and a lack of IFN-γ production. Further, we observed a significant positive correlation between lymphoproliferation and secretion of IFN-γ (r = 0·20, P = 0·0002) and IL-4 (r = 0·15, P = 0·005). Measles antibody levels were correlated with lymphoproliferation (r = 0·12, P = 0·03), but lacked correlation to either cytokine type. In conclusion, we demonstrated the presence of both long-term cellular and humoral responses after MMR-II vaccination in a significant proportion of study subjects. Further, a positive correlation between lymphoproliferation and IL-4 and IFN-γ suggests that immunity to measles may be maintained by both Th1 and Th2 cells. We speculate that the Th2 biased response observed in a subset of our subjects may be insufficient to provide long-term immunity against measles. Further examination of the determinants of Th1 versus Th2 skewing of the immune response and long-term follow-up is needed.

Introduction

Measles leads to nearly 1 million deaths each year, due primarily to failure to vaccinate and due in part to vaccine failure [1–6]. Antibody responses to measles infection or vaccination have been studied extensively and are used as correlates of protection against measles. Vaccine failure has been defined in terms of antibody response alone, and the markers of measles virus-specific cellular immunity after vaccination have been poorly characterized. Although the role and relative contribution of cellular immunity in imparting protection against measles in response to vaccination is unknown, CD4 T cell responses are known to influence indirectly the antibody responses [7–10]. Measles infection is known to cause an acute and profound state of immune suppression in few individuals, characterized by polarization of effector CD4 T cells to produce T helper type 2 (Th2) cytokines [7–9,11]. Similarly, measles vaccination can also cause immune suppression and predominant IL-4 production in some cases [10,12]. This skewing of the cytokine response towards predominant Th2-type cytokines results in a deficient or low Th1 cellular immune response and increases susceptibility to co-infections with various other pathogens [13,14]. The preferential expansion of the CD4 T cell pool into Th2 cells is characterized by an increase in IgE, IgG1 and IgG4, diminished production of the Th1 signature cytokine interferon (IFN)-γ and heightened production of interleukin (IL)-4 (a cytokine critical for establishing the Th2 type response) [8,15]. Further, cellular immunity is known to play a critical role in controlling measles virus replication. Children with cellular immune deficiencies are known to develop severe complications associated with measles virus infection and recover poorly compared to their healthy counterparts [16,17]. Because cellular immunity plays a critical role in the maintenance of recall responses and control of measles virus infection, it is important to evaluate the presence of measles-specific cellular immunity in addition to antibody levels in populations several years after vaccination and understand the possible correlations among these measures of immunity.

In the current study, we examined the correlation among measles virus-specific antibodies, lymphoproliferation and Th1/Th2 cytokine responses following two doses of measles–mumps–rubella-II (MMR-II) vaccine. The purpose of this study was to identify markers of immune response to each effector arm of immunity and to assess if correlations exist among these measures of immunity.

Materials and methods

Human subjects

The details of our study sample recruitment have been described previously [18]. Briefly, we recruited a population-based, stratified random sample of 346 healthy children and young adults (12–18 years of age) from all socio-economic strata residing in Rochester, Minnesota. Study subjects had documentation in their medical records of receiving two doses of the MMR-II vaccine containing the Edmonston strain of measles virus (tissue culture infective dose TCID₅₀ ≥ 1000), the Jeryl Lynn B-strain of mump virus (≥ 20 000 TCID₅₀) and the Wistar RA 27/3-strain of rubella virus (≥ 1000 TCID₅₀) on or after the age of 12 months. There had been no circulating wild-type measles virus in this geographical area of residence during the subjects’ lifetimes. We determined the level of measles-specific antibodies, lymphoproliferation and cytokine production after ex vivo measles stimulation to characterize overall immunity to measles in our study population. We included 339 subjects in the final analysis for whom we had all four measures of immunity. The study was approved by the Mayo Clinic's Review Board. We obtained written, informed consent or assent (from younger subjects) from subjects and/or parents/guardians from all the subjects at the time of enrolment in the study.

Antibody assays

Quantitative levels of measles-specific IgG antibody titres for all serum specimens were determined by the Enzygnost (Dade Behring, Marburg, Germany) anti-masern-virus/IgG enzyme immunoassay (sensitivity = 99·6%; specificity = 100%), as described previously [18]. Briefly, sera were separated from whole blood by centrifugation. Sera were aliquoted and stored at −80°C until the time of the assay. Serum samples were added in duplicate to a microtitre plate, which contained two parallel wells coated with a test whole measles virus antigen or control antigen derived from non-infected cells. Optical densities were determined at 450 nm and corrected at 650 nm on a microplate reader (Molecular Devices Corporation, Sunnyvale, CA, USA). The difference in mean absorbance between the viral test antigen and the control antigen for each sample was calculated and multiplied by a correction factor (determined by dividing a kit-specific nominal value by the mean of the reference standard) to give the corrected change in absorbance (ΔA). Measles-specific IgG antibody levels (mIU/ml) were calculated from the antilog of the following formula: log₁₀ = α*▵A^β (where α and β are lot-dependent constants). Positive antibody response was defined as a change in absorbance (▵A) > 0·2. The coefficient of variation for this assay in our laboratory was 3·8%.

Preparation of peripheral blood mononuclear cells (PBMC)

PBMC were separated from heparinized venous blood by Ficoll-Hypaque (Sigma, St Louis, MO, USA) density gradient centrifugation and washed in RPMI-1640 medium (Celox Laboratories, Inc., St Paul, MN, USA) supplemented with 100 µg/ml streptomycin (Sigma), 100 U/ml penicillin (Sigma) and 8% heat-inactivated fetal calf serum (FCS) (Hyclone, Logan, UT, USA). Cells were then counted, resuspended in RPMI freezing media containing 10% dimethyl sulphoxide (Sigma), 20% FCS, frozen at −80°C overnight and stored in liquid nitrogen until cultured.

Lymphoproliferation assay

Lymphoproliferation to measles vaccine virus was assessed using an in vitro[³H]-thymidine incorporation assay as described previously [18]. Briefly, PBMC (2 × 10⁵cells/well) were incubated for 72 h in RPMI-1640 medium, supplemented with 5% autologous sera, in the presence of live attenuated measles vaccine virus [75 plaque-forming units (pfu)/well]. Plain RPMI media served as the unstimulated control, and phytohaemagglutinin (PHA, 5 µg/ml) was used as a positive control for lymphoproliferation. The PBMC were then pulsed with [³H]-thymidine for 18 h, harvested on glass fibre filters and counted for the amount of incorporated radioactivity. Results were then expressed as stimulation indices (SI), which are defined as the ratio of the median counts per minute (cpm) of antigen-stimulated wells to the median cpm of unstimulated control wells. Stimulation indices of three or higher were considered to be an indicator of a positive lymphoproliferation response [19].

Measurement of IFN-γ and IL-4 cytokines

The details of the IFN-γ and IL-4 cytokine assays are described in detail elsewhere [20]. Briefly, for IFN-γ determination, thawed PBMC were cultured at a concentration of 2 × 10⁵ in RPMI 1640 media containing 10% normal human serum (NHS), 100 U/ml penicillin and 100 µg/ml streptomycin with or without measles (MV) virus (Edmonston B vaccine strain of MV cultured in African green monkey kidney cells) at a multiplicity of infection (moi) of 0·5 for 6 days. For IL-4 determination, we cultured thawed PBMC at a concentration of 4 × 10⁵ in RPMI-1640 media, supplemented with 10% NHS, 100 U/ml penicillin and 100 µg/ml streptomycin. Cells were cultured in the presence of 2 µg/ml of monoclonal IL-4 receptor antibody (R&D Systems, Minneapolis, MN, USA) with measles virus at a moi of 0·1 for 6 days as previously described [21,22]. These doses were chosen carefully to operate in approximately 70% measles-virus stimulation dose for immune activation (and not too high or too low a dose) to avoid skewing the immune response to either direction due to overloading or under-presentation of the antigen. Stimulation with PHA, 5 µg/ml, was used as a positive control. Vero cell lysate was used as unstimulated control. Cell culture supernatants were collected and assayed for IFN-γ and IL-4 using a standard enzyme-linked immunosorbent assay (ELISA) protocol (OptiEIA Human IFN-γ and IL-4) (PharMingen, San Diego, CA, USA). The IFN-γ and IL-4 concentration of each test sample was calculated by reference to the standard curve. Median background levels of IFN-γ and IL-4 cytokine production in unstimulated wells were subtracted from the median measles-induced responses to produce ‘corrected’ secretion values. Any value above the unstimulated control was considered positive. Negative corrected values indicate that the unstimulated secretion levels were, on average, higher than the stimulated secretion levels.

Statistical analysis

Data were summarized descriptively using frequencies and percentages for all categorical variables and medians and interquartile ranges (IQR) for all continuous variables. Associations of immune response with demographic and clinical variables were assessed using analysis of variance methods. Due to data skewness, all P-values were calculated using rank-transformed values.

An intraclass correlation coefficient (ICC) and corresponding 95% confidence interval was calculated to assess the magnitude of variance across the four different immune response measures relative to error variance [23]. Rank-transformed values were used to calculate the ICC due to data skewness in the immunity variables. Spearman's correlation coefficients were then used to assess all pairwise associations between the immunity measures. All statistical tests were two-sided, and all analyses were carried out using SAS software system (SAS Institute, Inc., Cary, NC, USA).

Results

Demographic variables

Our study population had a median age of 15 years at the time of recruitment with near equal gender representation (53% males). The majority of subjects were Caucasians (93%) and the median age at the first and second immunization was 15·6 months and 12·1 years, respectively. Measles virus-specific immune responses were measured after a median time period of 4·8 years after the last immunization (Table 1).

Table 1. Demographic variables of the study population.

Variable	n = 339
Age, years (median, IQR*)	15 (13,17)
Gender (n, %)
Female	159, 47%
Male	180, 53%
Race (n, %)
Caucasians	316, 93%
Others	23, 7%
Age, months at first MMR-II immunization (median, IQR)	15·6 (15·1, 17·5)
Age, years at second MMR-II immunization (median, IQR)	12·1 (11·5, 12·6)
Time, years since last MMR-II immunization to blood draw (median, IQR)	4·8 (2·8, 6·1)

^*IQR, interquartile range; MMR-II, measles–mumps–rubella-II.

Immunological variables

The median value for measles-specific IgG was 1553 mIU/ml, and the median SI for the lymphoproliferation was 3·8 (Table 2). We observed an increase in the antibody levels to measles with increasing age at the first MMR-II vaccination (P = 0·05). Similarly, an increase in age at the second MMR-II vaccination showed an increasing trend in antibody levels to measles (P = 0·04) (data not shown). No association was found between the lymphoproliferative response and any of the demographic variables (data not shown). Median values for measles-specific IFN-γ and IL-4 cytokines were 40·7 pg/ml and 9·7 pg/ml, respectively (Table 2). The IFN-γ and IL-4 responses to measles vaccination were found to be independent of gender, age or time lapsed since last immunization in our study cohort (data not shown).

Table 2. Characterization of measles-specific cellular and humoral positive responses in study subjects.

Variable (n = 339)	Median (IQR)	n, % Subjects with positive response*
IgG (mIU/ml)	1553 (737, 2637)	290, 86%
Stimulation index (SI)	3·8 (1·9, 6·6)	195, 58%
IFN-γ (pg/ml)	40·7 (8·1, 176·7)	281, 83%
IL-4 (pg/ml)	9·7 (2·8, 24·3)	288, 85%

^*Positive antibody response = change in absorbance (ΔA) > 0·2. Measles-specific IgG antibody levels (mIU/ml) are calculated from the antilog of the following formula: log₁₀ = α*ΔAβ(where α and β are lot-dependent constants). * Positive stimulation index (SI) response ≥ 3. * Positive cytokine response is any value above background. Median background levels of interferon (IFN)-γ and interleukin (IL)-4 cytokine production in unstimulated wells were subtracted from the median measles-induced responses to produce ‘corrected’ secretion values. The uncorrected median (IQR) for IFN-γ (pg/ml) was 40·7 (3·7, 196·3) and −2·9 (−9·9, 6·85) for measles virus (MV) stimulated and unstimulated control, respectively. The uncorrected median (IQR) for IL-4 (pg/ml) was 15·5 (5·0, 30·8) and 3·46 (−1·42, 10·48) for MV stimulated and unstimulated control, respectively.

The proportion of individuals that had a positive response for measles specific antibody, lymphoproliferation, IFN-γ and IL-4 as per our cut-off criteria was 0·86, 0·58, 0·83 and 0·85, respectively (Table 2). Overall, 37·5% of our study subjects had a positive response for all four measures of immunity, 98·8% for at least one measure of immunity, while only 1·2% lacked positivity to any of the measured variables (Table 3). From the 98·8% of subjects that measured positive for at least one variable, 7·4% of the subjects demonstrated undetectable levels of IFN-γ production and negative measles-specific lymphoproliferation (Table 3).

Table 3. Characterization of measles-specific humoral and cellular immunity in study subjects.

Immune measure	Variable positive	Frequency*	Percentage*	% Cumulative
Cellular and humoral	IgG, SI, IFN-γ, IL-4	127	37·5	90·2
	IgG, SI, IFN-γ	20	5·9
	IgG, SI, IL-4	21	6·2
	IgG, IFN-γ, IL-4	82	24·2
	SI, IFN-γ, IL-4	17	5·0
	IgG, SI	4	1·2
	IFN-γ, IL-4	17	5·0
	IgG, IFN-γ	15	4·4
	SI, IL-4	3	0·9
	Th1-biased	SI, IFN-γ	2	0·6	1·2
	SI	1	0·3
	IFN-γ	1	0·3
Th2-biased	IgG, IL-4	17	5·0	7·4
	IL-4	4	1·2
	IgG	4	1·2
None	–	4	1·2	1·2

^*Frequency and percentage for individual immune measure are based on the cut-off criteria mentioned in the text.

Correlation between immunological variables

Table 4 shows the correlation among the pairs of measures of measles-specific antibody, lymphoproliferation, IFN-γ and IL-4 as represented by a Spearman's correlation matrix. The antibody levels were correlated to lymphoproliferation (r = 0·12, P = 0·03), but lacked any correlation with the measures of IFN-γ and IL-4 secretion. Alternatively, we found a significant correlation between lymphoproliferation and IFN-γ (r = 0·20, P = 0·0002) as well as lymphoproliferation and IL-4 (r = 0·15, P = 0·005) secretion. Further, measles-specific IFN-γ had a significant positive correlation with measles-specific IL-4 secretion (r = 0·29, P < 0·0001) (Table 4). The overall value of the intraclass correlation for all four measures of immunity for measles was 0·14 (95% CI = 0·09–0·19). However, as these correlations are weak in magnitude (r < 0·3), they should be interpreted with caution.

Table 4. Correlation between measles-specific antibody, lymphoproliferation, interferon (IFN)-γ and interleukin (IL)-4 responses in study subjects as represented by a Spearman's correlation matrix.

Variable	IgG level (mIU/ml)	Stimulation index (SI)	IFN-γ (pg/ml)	IL-4 (pg/ml)
IgG level (mIU/ml)	1·00	0·12 (0·03)	0·03 (0·56)	0·04 (0·45)
Stimulation index (SI)	–	1·00	0·20* (0·0002)	0·15* (0·005)
IFN-γ (pg/ml)	–	–	1·00	0·29* (< 0·0001)
IL-4 (pg/ml)	–	–	–	1·00

The values in parentheses represent the corresponding P-values.

^*P-value ≤ 0·005 are shown in bold type.

Discussion

In the current study, we demonstrate the presence of long-term cellular and/or humoral immunity after vaccination with two doses of the currently recommended MMR-II vaccine in the majority of our study population by measuring measles-specific lymphoproliferation, antibodies and Th1/Th2 signature cytokines, IFN-γ and IL-4. However, a small fraction of individuals (7·4%) demonstrated a Th2-biased immune response characterized by the presence of antibodies and IL-4 and undetectable IFN-γ and lymphoproliferative response.

Although protective efficacy of measles vaccine is defined in terms of antibodies alone, several ‘experiments of nature’ demonstrate that cellular immunity is critical and sufficient to recover from measles infection. Children with agammaglobulinaemia have a normal course of clinical measles and recover uneventfully from measles infection, while those with defective cellular immunity (e.g. leukaemia, HIV) develop progressive illnesses and have high mortality from measles infection [14,24]. Previous studies demonstrate that measles infection can result in an acute and profound state of immune suppression and skew the immune response towards the Th2 phenotype, characterized by high production of IL-4 and increased immunoglobulin levels [7–11]. In addition, there is substantial evidence for a biased Th2 response after measles vaccine administration as documented by decreased responsiveness to mitogens [12] and predominant IL-4 production [10]. However, we have limited understanding of these measures of immunity to measles in populations several years after vaccination. In the current cohort, we demonstrate that a majority (90%) of the subjects had evidence for measles-specific immune markers for both the humoral and cellular effector arms years after measles vaccination. However, a small yet significant proportion (7·4%) of subjects did show a preferential bias towards the Th2 type response. Further, an even smaller fraction (1%) demonstrated no evidence of any of the four measures of immunity to measles. Although there is no critical requirement of vaccine-induced cellular response in imparting protection against infection, the individuals lacking cellular immunity may be at a disadvantage for virus clearance. No data are available on evaluation of the potential roles of Th1 and Th2 cells in measles viral clearance in humans. However, a recent study using the rhesus monkey model of measles infection demonstrates clearly that CD8 lymphocytes are major mediators of protection, and antibodies have a limited role in the control of measles replication and clearance [25]. The lack of correlation between measles-specific antibodies and prolonged measles virus shedding in HIV-infected Zambian children also re-emphasizes the importance of cellular immunity in recovery from measles infection [26]. In contrast, antibodies alone can certainly mediate protection against measles virus infection as infants with immature cellular immunity can be protected by maternally acquired antibodies [27], or neutralizing antibodies from vaccination [28]. Similarly, a formalin-inactivated vaccine incapable of priming CD8 T cells is protective if antibody titres are high [29,30]. However, as antibodies wane over time, cellular immunity may be more important for maintaining long-term immunity and recovery from the disease. Similar discordance between the contribution of humoral immunity in viral clearance and protection has been established for other viruses, such as herpes simplex virus and hepatitis B virus [31,32].

The large sample size from our current study allowed us to determine correlations between measures of humoral and cellular immunity after vaccination, compared to solely looking at antibodies as markers for vaccine-induced immunity. We demonstrate a significant positive correlation between lymphoproliferation and two cytokines, IL-4 and IFN-γ, which play a crucial role in both the generation and regulation of immune responses. Further, the levels of IL-4 and IFN-γ were correlated positively in the study subjects. The results suggest that recall immunity to measles is maintained as a fine-tuned and tightly regulated immunological balance, where both IL-4 and IFN-γ play a key role.

IFN-γ is critical for macrophage activation, T cell proliferation and differentiation and up-regulation of human leucocyte antigens [33–35]. IFN-γ exerts an antiviral effect by promoting the lysis and clearance of measles-infected cells and by inhibiting viral gene expression and replication [25, 36, 37]. In comparison, a Th2 response is characterized by IL-4 production and high levels of measles-specific antibodies after measles infection [8, 10, 38], which is used traditionally as the measure of protective immunity against subsequent infection. Because we found a positive correlation between lymphoproliferation and IL-4 and IFN-γ, our results suggest that measles immunity is maintained by both Th1/Th2 cells (or Th0-type T cells that are producers of both IL-4 and IFN-γ) [39]. Similar Th0 responses to wild-type measles virus-infected dendritic cells have also been observed [40]. However, measles vaccine-infected dendritic cells or subjects immunized with the measles vaccine predominantly produce a Th1 cytokine response [40,41]. This implies that to mount an immune response at various stages of viral invasion, the host cells maintain a functional pool of mixed Th1/Th2 cell populations. The Th1 response may help to clear intracellular viruses from infected host cells, and the Th2 response may help to clear extracellular viruses, which can be neutralized by antibodies and inflammatory cytokines. For example, a high intact rate of humoral and cellular immunity and the presence of both Th1/Th2 cytokines are observed in influenza vaccinated subjects, which suggests the presence of Th0 cells [42]. In addition, we did not find any correlation between antibody levels to measles and measures of Th1/Th2 cytokines. This agrees with our previous findings demonstrating that correlations between measles-specific antibodies and IFN-γ and IL-4-secreting cell frequencies did not reach statistical significance [43]. Similar disagreement in quantitative correlation between antibody levels to measles and measures of T cell immunity can also be found in other studies [44–46]. Further work on the determinants of Th1 versus Th2 skewing of the immune response is needed.

The major strength of our study is the stratified random sample of our population. The resulting sample well represents the US Caucasian population with high vaccination rates. An advantage of an ethnically homogeneous population is decreased confounding variables (ethnicity, previous viral exposure history, etc.), increasing the confidence in our findings. Further, the immunity measures were true responses to vaccination, as these subjects had not encountered wild measles virus in their lifetimes. We understand that an ideal control to evaluate the specificity of measles responses would be testing for immune measures to measles stimulation in naive cells. However, we are confident that the IFN-γ and IL-4 levels detected in our study are measles-specific. For example, in our previous studies we were not able to detect any IL-4 and only very low levels of IFN-γ in naive cells after strong mitogen stimulation such as PHA [41]. However, we could detect significant IL-4 and IFN-γ secretion in response to ex vivo PHA stimulation in PBMC from these subjects after MMR-II vaccination [41]. This suggests that when we did observe IL-4 and IFN-γ secretion with measles stimulation in vaccinated individuals, it is represented as measles-specific response. A limitation of our study may be that we used different stimulation doses for IL-4 and IFN-γ secretion assays which may influence the differentiation to either Th1 or Th2 response. However, our selected doses had given us approximately 70% immune activation in our optimization studies, which is not too high or too low a dose to skew the immune response to either direction due to overloading or under-presentation of the antigen. Another limitation of our study is measurement of whole measles antibodies instead of neutralizing antibodies. Neutralizing antibodies would specifically measure protective antibody titres [28]. We were prompted to measure whole antibodies due to the correlation of neutralizing antibody titres with whole antibody enzyme immunoassays [47] and the ease of performing whole antibody immunoassays on a large sample size. Further, we measured Th1 and Th2 cytokines in a whole lymphocyte population instead of isolated CD4 and CD8 populations due to a limitation in cell numbers available. However, we believe that CD4 cells are the predominant producers of the Th1/Th2 mixed response to measles based on our previous studies characterizing cytokine secreting cell types after measles stimulation [22,39].

In conclusion, we demonstrate that the majority of our study population had detectable measures of measles-specific cellular and humoral immunity following a median of 4·8 years after receipt of the second dose of MMR-II vaccine. However, 1·2% lacked a positive response to any measures of measles vaccine induced immunity, and 7·4% demonstrated a Th2-biased immune response. Our study provides insights into the mechanisms of measles vaccine-induced immune responses. It may be concluded that, in most cases, the immune system mounts a balanced Th1/Th2 response to vaccination. However, in a small fraction of individuals, a bias towards either type may be induced. However, it is speculated that individuals with humoral immunity alone may be at a disadvantage for decreased viral clearance and long-term protection due to waning of antibodies over time. Additional doses of vaccines or improved vaccination strategies such as peptide-based vaccines may need to be developed to overcome the lack of cellular immunity in those individuals who do not respond to the current vaccine. This is critical for recall immunity and lifelong protection, where humoral immunity alone may be insufficient to recover from infection.