Conceived and designed the experiments: ES TS MY. Performed the experiments: YD HW. Analyzed the data: YD.
The authors have declared that no competing interests exist.
BCG vaccine drives a strong T helper 1 cellular immunity which is essential for the protection against mycobacteria, however recent studies suggest that BCG vaccination can have non-specific beneficial effects unrelated to tuberculosis. In the present cohort study the development of cytokine profiles following BCG vaccination was investigated. Immune responses to PPD were assessed before vaccination and at ages of 5 months, 1 year, and 2 years, followed by BCG scar measurement at 4 years of age. BCG was shown to induce both Th1 and Th2 type responses against PPD at about 5 months of age after vaccination, and while Th1 response was sustained, Th2 responses declined over time. However, BCG scar size was strongly correlated with Th2 responses to PPD at 5 months of age. Importantly, we observed no clear effects of BCG vaccination on innate immune responses in terms of early IL-10 or TNF-α production whereas some alterations in general adaptive immune responses to PHA were observed.
Most immunological studies on the efficacy of Bacille Calmette-Guérin (BCG) vaccination focus on the production of IFN-γ as the main feature of Th1 response, which leads to the activation of important cells such as macrophages to contain mycobacteria. Besides the partial protection against TB through IFN-γ production
In the current longitudinal study following BCG vaccination of neonates in Indonesia, we have examined the production of Th1 (IFN-γ) and Th2 (IL-5, IL-13) cytokines in day 6 supernatants of whole blood in response to PPD and PHA to determine specific and non-specific adaptive immune responses. To assess the effect of BCG vaccination on the development of innate immune responses, IL-10 and TNF-α were measured in one day culture supernatants of whole blood stimulated with LPS and PPD. We also analyzed the relationship between cytokine responses to mycobacterial antigens and scar formation at an age when the scar formation has stabilized. In order to assess how external factors might influence responses to BCG vaccination, we studied the effect of maternal parasitic infection status on the profile of cytokine production over time.
This study was conducted according to the principles expressed in the Declaration of Helsinki. The study was approved by Ethics Committee of Faculty of Medicine, University of Indonesia. All mothers provided written informed consent for the collection of samples from their children and for subsequent analysis.
This study was performed as part of a birth cohort to examine the development of immune responses of children living in areas endemic for helminth infection. Maternal parasitological data such as filarial or intestinal parasite infection was obtained during the recruitment of pregnant women as described before
Children were recruited from two adjacent villages in a peri-urban area, after their mothers gave informed consents for participation in this study. Blood was withdrawn by venipuncturist and was collected in a heparinized tube at several time points: T0: before BCG vaccination, T5: 5 months of age, T12: 1 year of age, T24: 2 years of age. In Indonesia, BCG vaccination is the first vaccine to be given according to the national vaccination program, followed by three Hepatitis B vaccination, three diphtheria-pertusis-tetanus (DPT) that are given together with three oral polio vaccination, and finally measles vaccination before the child reaches 1 year of age. BCG vaccination program requires every infant to be vaccinated soon after birth by health staff from local primary health care center (PHC). BCG vaccine used here contains attenuated live
After BCG vaccination, the resulting BCG scars were measured at 4 years of age by the same person from the research team, and the mean diameter of scar size was calculated from diameters perpendicular to each other. The reason for measuring BCG scar at 4 years of age is because the scar formation is expected to have stabilized. A BCG scar was considered negative if the mean diameter was less than 2 mm.
Heparinized venous blood was processed for culture within 6 hours after venipuncture. As previously described
IL-10, TNF-α, IL-13 and IFN-γ levels were measured in supernatants by in-house multiplex bead based assay using Luminex IS 100 (Luminexcorp, Austin, TX, USA) while IL-5 level was measured by ELISA as described elsewhere
Cytokine levels were analyzed in two ways, either using raw cytokine levels when comparing cytokines between two groups of children or using net cytokine levels (substracted from the negative control) to obtain percentage of responders/non-responders (above or below zero for substracted values). Mann-Whitney non- parametric test was used to compare cytokine levels which were not normally distributed. Since the sample size in each time point was not equal, Mann-Whitney test was used to compare cytokine responses between two time points of measurement. The correlations between cytokine responses to PPD and PHA and between BCG scar size and cytokine responses were analyzed using Spearman's rank correlation. In order to adjust cytokine levels for other variables (gender, birth weight, birth season, age receiving BCG, place of residence, maternal intestinal helminth or protozoan infection), multiple linear regression was performed with diameter of BCG scar as the outcome. Age receiving BCG, diameter of BCG scars and cytokine levels were previously log-transformed to get a normal distribution. The statistical analyses were performed in SPSS version 16. A two-tailed
A total of 147 children were recruited into the study and blood samples were obtained before BCG vaccination was given (T0). From these children, 120 children could be followed up at T5, 105 children at T12 and 98 children at T24. Apart from 66 children in whom blood samples were available at all time points, there were 81 children in whom blood samples were not available at one or two follow up time points due to the refusal of parents at the particular time point to allow blood sampling of their child (
Child characteristics | All children (n = 147) | Children with complete cytokine data (n = 66) | Children with incomplete cytokine data (n = 81) |
Jati Sampurna | 69 (47%) | 34 (52%) | 35 (43%) |
Jati Karya | 78 (53%) | 32 (48%) | 46 (57%) |
3.2 (0.5) | 3.2 (0.4) | 3.2 (0.5) | |
<3 kg | 23/99 (23%) | 13/50 (26%) | 10/49 (20%) |
> = 3 kg | 76/99 (77%) | 37/50 (74%) | 39/49 (80%) |
dry season | 74 (50%) | 39 (59%) | 35 (43%) |
rainy season | 73 (50%) | 27 (41%) | 46 (57%) |
male | 70/141 (50%) | 36/66 (54%) | 34/75 (45%) |
female | 71/141 (50%) | 30/66 (46%) | 41/75 (55%) |
5 (2.0–8.5)(n = 141) | 4 (1.7–8.2)(n = 66) | 5 (2.0–9.0)(n = 75) | |
20.1 (5.8) (n = 119) | 20.7 (5.7)(n = 66) | 19.4 (5.9) (n = 53) | |
3.3 (2.0–4.5) (n = 96) | 3.3 (1.5–4.1) (n = 58) | 3.5 (2.4–4.8) (n = 38) | |
Filaria | 35/147 (24%) | 15/66 (23%) | 20/81 (25%) |
Intestinal helminth infection | 51/139 (37%) | 18/65 (28%) | 33/74 (45%) |
25/139 (18%) | 8/65 (12%) | 17/74 (23%) | |
21/139 (15%) | 9/65 (14%) | 12/74 (16%) | |
11/139 (8%) | 4/65 (6%) | 7/74 (9%) | |
Intestinal protozoan infection | 39/139 (28%) | 13/65 (20%) | 26/74 (35%) |
31/139 (22%) | 10/65 (15%) | 21/74 (28%) | |
15/139 (11%) | 3/65 (5%) | 12/74 (16%) | |
Co-infection of intestinal helminth and protozoan | 20/139 (14%) | 6/65 (9%) | 14/74 (19%) |
*other intestinal protozoa species:
In comparison with pre vaccination time point (T0), IFN-γ, IL-5 and IL-13 responses to PPD increased after BCG vaccination at 5 months of age (p<0.001 for all cytokines). In contrast to IFN-γ which remained high until the children reached 2 years of age, IL-5 and IL-13 responses showed a gradual decrease (
IFN-γ, IL-5, IL-13 responses to PPD (A) and PHA (B). Levels of cytokines at T0 (before vaccination, n = 143); T5 (5 months of age, n = 119), T12 (1 year of age, n = 103) and T24 (2 years of age, n = 97).The horizontal lines in the bars represent the median, with the lower 25% and upper 75% percentiles and extended whiskers the 10% and 90% percentiles. Mann-Whitney test: *0.05>
To determine whether BCG vaccination influenced the polyclonal non-specific responses as well, cytokine production to PHA was measured. As shown in
The relations between IFN-γ production in response to PPD and to PHA were analyzed next (
. r = Spearman's correlation coefficient, p =
We were interested in the pro and anti inflammatory innate immune responses and therefore measured IL-10 and TNF-α in day 1 supernatants following stimulation of whole blood with PPD (stimulates TLR2, unpublished data) and LPS (stimulates TLR4). Compared to T0, no increase and even a tendency for a gradual decrease over time was observed for IL-10 response to PPD (T5: p>0.05, T12: p>0.05, T24: p<0.001). Similarly, TNF-α response to PPD decreased over time up to 2 years of age (T5: p>0.05, T12: p = 0.001, T24: p<0.001) (
Levels of cytokines at T0 (before vaccination, n = 147 for PPD and n = 101 for LPS), T5 (5 months of age, n = 120 for PPD and n = 117 for LPS), T12 (1 year of age, n = 105 for PPD and n = 103 for LPS) and T24 (2 years of age, n = 98 for PPD and n = 97 for LPS).The horizontal lines in the bars represent the median, with the lower 25% and upper 75% percentiles and extended whiskers the 10% and 90% percentiles. Mann-Whitney test: *0.05>
Cytokine | Time point | n | Medium | n | PHA | n | PPD | n | LPS |
IL-10 | T0T5T12T24 | 14611910598 | 3.3(3.3–19.1)3.3(3.3–23.4)3.3(3.3–26.1)3.3(3.3–3.3) | 14712010598 | 84.8(42.6–144.4)56.4(31.5–117.2)58.3(30.4–103.4)64.7(40.4–97.5) | 14712010598 | 48.6(20.9–98.7)40.3(15.0–83.9)38.6(16.7–107.0)17.5(5.7–43.9) | 10111710397 | 348.0(195.0–664.7)325.4(161.5–551.9)357.8(159.2–742.1)159.2(22.5–283.9) |
TNF-α | T0T5T12T24 | 14611910598 | 0.84(0.84–9.7)0.84(0.84–4.7)0.84(0.84–5.2)5.4(0.84–24.0) | 14712010598 | 204.6(63.9–517.3)223.0(78.5–492.8)207.9(87.8–458.5)176.6(58.7–448.3) | 14712010598 | 308.9(122.2–925.6)243.9(126.1–587.8)193.5(98.5–356.2)125.8(53.3–484.5) | 10111710397 | 3528.8(1651.5–5808.7)3054.3(1821.2–5671.1)3553.2(1794.0–6075.0)1312.9(709.2–2173.2) |
IFN-γ | T0T5T12T24 | 14211910397 | 1.8(1.8–21.8)7.9(1.8–40.6)6.6(1.8–31.0)1.8(1.8–1.8) | 14311910397 | 92.2(35.2–239.3)110.0(41.2–265.6)116.9(41.3–258.7)283.7(89.4–840.8) | 14311910397 | 38.2(4.3–105.6)204.3(72.1–504.1)132.1(40.0–245.6)355.3(63.0–983.5) | ||
IL-5 | T0T5T12T24 | 14211910397 | 2.4(1.0–8.6)2.4(1.0–7.5)1.0(1.0–6.1)1.0(1.0–8.1) | 14311910397 | 713.0(223.1–1636.0)554.9(295.7–1421.4)825.4(430.3–2000.0)793.5(262.2–1575.9) | 14311910397 | 5.96(1.0–16.5)64.7(19.3–226.6)27.2(8.4–104.4)19.9(1.0–54.5) | ||
IL-13 | T0T5T12T24 | 14211910397 | 6.3(6.3–6.3)6.3(6.3–13.4)6.3(6.3–15.4)6.3(6.3–6.3) | 14311910397 | 497.5(254.6–1164.6)331.3(157.0–803.0)272.5(150.5–645.6)435.9(139.3–1033.4) | 14311910397 | 6.3(6.3–20.3)31.0(14.0–106.6)20.8(6.3–48.8)6.3(6.3–33.2) |
T0: before vaccination (2 months of age), T5: 5 months of age, T12: 1 year of age, T24: 2 years of age. All cytokine levels are expressed in median and interquartile range.
Neither filarial infection (data not shown) nor intestinal helminth infection status of mothers had a clear effect on the response of children to PPD in terms of Th1, Th2, pro or anti inflammatory cytokines either at pre or at different ages after vaccination (
Cytokine | Time | n | IH+ | n | IH- | n | IP+ | n | IP- | ||
IL-10 | T0T5T12T24 | 51393435 | 39.0(14.9–83.6)30.6(13.3–66.3)34.7(14.5–124.4)22.1(10.8–57.9) | 88756462 | 50.6(25.4–103.7)48.3(15.2–93.6)45.2(20.4–120.8)15.2(3.3–37.5) | 0.1040.3030.4850.102 | 39322327 | 25.6(12.5–49.4)17.2(10.1–45.7)27.8(12.1–81.3)20.4(3.3–45.1) | 100827570 | 62.5(25.6–116.3)52.8(19.2–95.4)49.3(22.5–124.9)17.4(6.5–44.0) | 0.0020.0020.0710.942 |
TNF-α | T0T5T12T24 | 51393435 | 200.5(84.8–885.5)184.0(119.1–537.8)237.4(126.3–379.0)160.2(40.3–826.4) | 88756462 | 335.7(149.0–930.1)287.6(126.1–682.1)177.1(86.3–383.8)119.7(55.2–401.9) | 0.2280.3480.4380.583 | 39322327 | 199.5(77.0–502.5)158.9(78.4–393.3)124.1(42.3–195.3)60.2(30.8–139.0) | 100827570 | 359.4(148.6–1093.7)282.2(136.6–784.3)243.0(119.8–484.8)163.8(80.0–540.6) | 0.0050.0200.0110.002 |
IFN-γ | T0T5T12T24 | 51383435 | 21.3(1.8–93.5)159.5(69.9–379.1)114.5(40.2–263.1)563.9(128.4–1143.0) | 85756361 | 50.2(7.4–111.4)210.3(71.3–601.0)158.1(42.5–253.9)273.8(51.1–900.0) | 0.0840.4700.7280.104 | 39312227 | 18.3(1.8–102.6)140.0(52.2–319.2)163.9(58.0–288.0)171.0(49.6–1057.5) | 97827569 | 48.4(7.1–116.0)234.8(87.9–685.9)132.1(34.1–243.8)360.8(77.6–983.6) | 0.1310.0360.4530.483 |
IL-5 | T0T5T12T24 | 51383435 | 5.1(1.0–16.4)61.5(12.4–276.8)57.9(15.8–161.6)19.9(1.0–63.6) | 85756361 | 6.5(1–16.8)66.5(22.9–213.6)22.3(7.3–72.1)18.9(1–54.5) | 0.6810.7840.0260.618 | 39312227 | 6.6(1.0–17.1)47.9(16.9–242.7)28.3(10.9–81.9)13.5(1.0–46.1) | 97827569 | 6.1(1.0–16.5)69.2(20.6–228.7)28.2(8.4–133.0)22.1(1.0–66.3) | 0.7260.8900.9900.185 |
IL-13 | T0T5T12T24 | 51383435 | 6.3(6.3–18.9)37.7(6.3–113.6)22.1(6.3–58.7)6.3(6.3–42.6) | 85756361 | 6.3(6.3–20.4)30.5(14.1–104.9)20.8(6.3–48.8)6.3(6.3–31.7) | 0.9800.6120.5120.863 | 39312227 | 6.3(6.3–15.3)25.3(6.3–86.9)9.8(6.3–46.8)6.3(6.3–22.3) | 97827569 | 6.3(6.3–22.8)33.0(16.8–116.7)24.1(6.3–51.2)6.3(6.3–34.2) | 0.0320.1900.2630.256 |
IH: intestinal helminth, IP: intestinal protozoa. T0: before vaccination (2 months of age), T5: 5 months of age, T12: 1 year of age, T24: 2 years of age.
Interestingly, children born to mothers infected with intestinal protozoa in which
Of sixty six children with complete cytokine data at all time points, there were 58 children with BCG scar measurement at 4 years of age. Since BCG vaccination increased the IFN- γ, IL-5 and IL-13 responses to PPD but not the innate responses as shown in IL-10 or TNF-α responses to PPD, we show the production of IFN- γ and IL-5 in 15 vaccinated children with no BCG scar (<2 mm) and 43 children with a positive BCG scar (≥2 mm) (
Open dots represent individuals with BCG scar <2 mm, closed dots represent those with BCG scar > = 2 mm. N = 58 children. r = Spearman's correlation coefficient, p =
The cytokine responses to PPD at older age, thus longer after BCG vaccination were weakly associated with BCG scar size at 4 years of age (data not shown).
The present study shows that BCG not only increases specific adaptive responses in terms of Th1 and Th2 cytokines in response to mycobacterial antigens but it also affects non-specific polyclonal responses. However, we did not find any evidence for its ability to influence pro and anti inflammatory innate immune responses as assessed by early IL-10 and TNF-α production in response to PPD or by LPS.
BCG vaccination enhanced Th1 and Th2 cytokine responses to mycobacterial antigen at 5 months of age (around 20 weeks after vaccination) and although the elevated Th1 responses to PPD were maintained up to 2 years of age, the Th2 responses to PPD waned. In line with our results, several studies in Gambian and United Kingdom infants
With respect to adaptive immune responses, it was interesting to note that there was some enhancement of polyclonal responses as assessed by IFN-γ and IL-5 production stimulated by PHA. The correlations between responses to PHA and PPD were strongest at 5 months and 1 year of age but not at later age, which might suggest that BCG can have an effect on cellular immunity beyond that to mycobacterial antigen.
We found no significant increase in IL-10 and TNF-α responses to PPD following BCG vaccination. It should be noted that IL-10 and TNF-α were measured at day 1 post stimulation to give us an insight into early innate responses, which are highest in day 1 supernatants. It has been shown that mycobacteria and its components
In contrast to the study of Malhotra and coworkers
In our study,
Finally, many studies examining the mechanisms behind tissue fibrosis and remodeling have indicated the involvement of Th2 responses in pathogen or chemical induced injury [30,31], which by stimulating collagen formation might initiate the repair processes [32]. However the role of Th2 responses in scar formation induced by BCG has not been studied before. Here we show a strong association between IL-5 or IL-13 responses to PPD early after BCG vaccination and scar formation at 4 years of age when scar formation is thought to be stabilized. A study done by Elliott and colleagues found that IL-5 response to culture filtrate protein of
In summary, this study has demonstrated the induction of both antigen specific Th1 (long term) and Th2 (short term) cytokine responses, with BCG scar formation associated more strongly with a Th2 cytokine response early after vaccination. Although, the innate IL10 and TNF-α responses were not affected by BCG vaccination, there was some indication of enhancement of adaptive responses beyond PPD after vaccination. These studies need to be taken further by a more in- depth analysis of the immunological changes at the innate and the adaptive immune system in order to be able to understand the non-specific effects of BCG on mortality and morbidity found in epidemiological studies.
Cytokine responses to PPD in children born to mothers either co-infected with intestinal helminths and protozoa or negative for both.
(0.06 MB DOC)
We are grateful to all participants and health staff at Puskesmas Jati Sampurna who were involved in this study. We thank Sudirman for his assistance in blood collection and whole blood assay. We thank Y.C.M Hoeksma-Kruize for her expertise in the whole blood assay technique. We thank Purnomo for the examination of stool samples.