PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
 
Vaccine. Author manuscript; available in PMC 2017 March 19.
Published in final edited form as:
PMCID: PMC5357573
EMSID: EMS71520

Maternal BCG scar is associated with increased infant proinflammatory immune responses

Abstract

Introduction

Prenatal exposures such as infections and immunisation may influence infant responses. We had an opportunity to undertake an analysis of innate responses in infants within the context of a study investigating the effects of maternal mycobacterial exposures and infection on BCG vaccine-induced responses in Ugandan infants.

Material and methods

Maternal and cord blood samples from 29 mother-infant pairs were stimulated with innate stimuli for 24 h and cytokines and chemokines in supernatants were measured using the Luminex® assay. The associations between maternal latent Mycobacterium tuberculosis infection (LTBI), maternal BCG scar (adjusted for each other’s effect) and infant responses were examined using linear regression. Principal Component Analysis (PCA) was used to assess patterns of cytokine and chemokine responses. Gene expression profiles for pathways associated with maternal LTBI and with maternal BCG scar were examined using samples collected at one (n = 42) and six (n = 51) weeks after BCG immunisation using microarray.

Results

Maternal LTBI was positively associated with infant IP-10 responses with an adjusted geometric mean ratio (aGMR) [95% confidence interval (CI)] of 5.10 [1.21, 21.48]. Maternal BCG scar showed strong and consistent associations with IFN-γ (aGMR 2.69 [1.15, 6.17]), IL-12p70 (1.95 [1.10, 3.55]), IL-10 (1.82 [1.07, 3.09]), VEGF (3.55 [1.07, 11.48]) and IP-10 (6.76 [1.17, 38.02]). Further assessment of the associations using PCA showed no differences for maternal LTBI, but maternal BCG scar was associated with higher scores for principal component (PC) 1 (median level of scores: 1.44 in scar-positive versus −0.94 in scar-negative, p = 0.020) in the infants. PC1 represented a controlled proinflammatory response. Interferon and inflammation response pathways were up-regulated in infants of mothers with LTBI at six weeks, and in infants of mothers with a BCG scar at one and six weeks after BCG immunisation.

Conclusions

Maternal BCG scar had a stronger association with infant responses than maternal LTBI, with an increased proinflammatory immune profile.

Keywords: Maternal infections, Latent Mycobacterium tuberculosis infection, Maternal BCG scar, Infant innate responses, BCG immunisation, Tuberculosis, Heterologous effects

1. Introduction

The bacillus Calmette-Guérin (BCG) vaccine protects against tuberculous meningitis and miliary tuberculosis (TB) in the infant [13], and also protects against leprosy [4]. However, the protective efficacy of BCG against pulmonary TB varies between populations, with latitude highlighted as an important factor for responses in adolescents and adults [1,5,6]. We recently investigated the effect of maternal latent Mycobacterium tuberculosis infection (LTBI) on the infant response to BCG immunisation [7], with results suggesting that maternal M. tuberculosis infection may impair adaptive immune responses in the infants, although a study in South Africa showed no such effect [8]. The associations with innate immune responses were not assessed.

Evidence that BCG immunisation may influence innate responses includes findings in both observational studies and randomized controlled trials that have highlighted the heterologous effects of BCG on childhood survival in both low- and high-income countries [913]. This has been suggested to be due to BCG-induced increases in function of the innate immune system, a phenomenon termed ‘trained immunity’ [1418]. This is an observation of great global health significance, since mortality due to infectious agents other than TB is high in developing tropical countries [19].

One of the indicators of previous immunisation with BCG, in place of or in addition to vaccination records, is the presence or absence of a scar [2022]. It has been shown that 52–97% of newborns administered BCG vaccine develop a scar, with differences depending on the strain of BCG vaccine used, the administrator and age of administration [20,2326]. However, not all BCG vaccinated babies will scar. There are reports of a correlation between the presence of a scar and protection against TB [27,28], as well as studies showing better survival with fewer respiratory infections [24,29,30], fewer skin infections and sepsis [31] in infants with a BCG scar.

Little is known about the link between the development of a BCG scar in mothers and immune responses in infants. We have previously observed that maternal BCG scar was associated with lower T helper (Th) 2 responses to crude culture filtrate proteins of mycobacteria in the infants [32]. In the context of a study designed to investigate the effects of maternal infections, including LTBI, on infant immune responses [7], we had the opportunity to also evaluate associations between maternal BCG scar and immune response profiles in the offspring.

2. Materials And Methods

2.1. Study design, setting and ethical approval

The study design, settings, laboratory and clinical procedures have been described elsewhere [7]. Briefly, women residing within the study area and delivering in Entebbe General Hospital were eligible for inclusion. They were approached for consent, on admission in early labour, if they were willing to participate in the study, had a normal singleton pregnancy and were HIV negative. Cord blood was obtained at delivery, following consent. A questionnaire was completed to assess eligibility after delivery. The tuberculin skin test (TST, Statens Serum Institut, Copenhagen, Denmark) and T-SPOT.TB assay (Oxford Immunotec, Abingdon, UK) were used to test mothers for LTBI at approximately one week after delivery. Infants were then followed up to six weeks of life. This was an exploratory observational study in a relatively small number of subjects. The number of infants included in the study was chosen to be feasible within the time frame and resources available. The study was approved by the Uganda Virus Research Institute-Research and Ethics Committee, the Uganda National Council for Science and Technology and the London School of Hygiene & Tropical Medicine. Written, informed consent was obtained from participating women for themselves and their infant.

2.2. Immunological assays

Innate immune responses were measured in 29 mother-infant pairs using a whole blood assay (WBA) with supernatant analytes measured by Luminex®, and gene expression profiles were measured in infant samples obtained at one (n = 42) and six (n = 51) weeks after BCG immunisation using microarray.

2.3. Innate stimulation and measurement of responses using luminex® assay

Heparinized maternal and cord blood samples were diluted 1:1 with RPMI 1640 medium (Life Technologies Corporation, NY, USA) and stimulated with lipopolysaccharide (LPS) (toll-like receptor (TLR) 4 agonist, 100 ng/ml), FSL-1 (TLR2/6 agonist, 50 ng/ml), CpG-ODN2006 (TLR9 agonist, 5 μg/ml), CL097 (TLR7/8 agonist, 1 μg/ml) all from InvivoGen, San Diego, CA, USA, PAM3Cys-Ser (TLR1/2 agonist; ECM Microcollections GmbH, Tubingen, Germany; 100 ng/ml), Mannan (DC-SIGN agonist; Sigma-Aldrich; 100 μg/ml) and Curdlan (Dectin-1 agonist; Wako Chemicals GmbH, Neuss, Germany; 100 μg/ml). An unstimulated well was included to act as a negative control. After 24 h of incubation at 37 °C in 5% CO2, culture supernatants were harvested and stored at -80 °C for analysis of cytokines and chemokines. The concentrations of analytes in the culture supernatants were measured using a Bioplex multiplex cytokine assay system (Bio-Rad Laboratories, Hercules, CA, USA), following instructions from the manufacturer. A Bio-Plex 200 System (Bio-Rad Laboratories, Hercules, CA, USA) and the Bio-Plex Manager software (version 6.0; Bio-Rad Laboratories, Hercules, CA, USA) were used to run the samples. According to the manufacturer’s instructions, a curve fit was applied to standard curves, which were then used to extract sample concentrations. Limits of the assay working range (lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ)) quoted by the manufacturer for each cytokine/chemokine were used to clean the data. For values below the acceptable range, half of the LLOQ was used and for values above the ULOQ, the ULOQ value for that particular analyte was used. The cytokines and chemokines analysed were IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, IL-17A, IFN-γ, IP-10, MCP-1, MIP-1α, MIP-1β, RANTES, TNF-α, GM-CSF and VEGF.

2.4. RNA amplification and microarray

Gene expression microarrays were undertaken using unstimulated whole blood samples obtained from 42 and 51 infants at one and six weeks, respectively, to assess gene expression profiles after BCG immunisation. The Illumina RNA Amplification Kit (Ambion, Austin, TX, USA) was used to amplify a median of 124 ng (range 63–174 ng) of the extracted RNA. A Biotin-16-UTP label was incorporated into amplified RNA during the in vitro transcription process (Perkin Elmer Life and Analytical Sciences, Woodbridge, Ontario, Canada). Amplification gave yields ranging from 1 μg to 25 μg. Amplified RNA (1000 ng per array) was hybridized to the IlluminaHumanHT-12_V4 BeadChip according to the manufacturer’s instructions (Illumina, San Diego, CA, USA). The IlluminaHumanHT-12_V4 bead chip comprises 42,000 sequences representing 31,000 annotated genes from the curated portion of the NIH Reference Sequence Database (http://www.ncbi.nlm.nih.gov/RefSeq/). Each sequence is represented at least 30 times on the array. Arrays were scanned with an Illumina bead array confocal scanner, according to the manufacturer’s instructions. Array data processing and analysis was performed using Illumina BeadStudio software.

2.5. Statistical analysis

The objective of this analysis was to investigate the effects of maternal latent TB and helminth infection on infant innate immune responses. In the event, helminth infections were rare in this study group [7], so the principal exposures considered were maternal LTBI and maternal BCG scar. In the multivariate analysis, the effects of maternal LTBI and maternal BCG scar were adjusted for. Maternal and infant factors such as maternal age, gravidity status, infant birth weight and gender were not crudely associated with infant responses and were not adjusted for, and the numbers involved were generally small.

Cytokine and chemokine concentrations showed skewed distributions. Results were transformed to log10 (cytokine concentration + 1) for graphical representation using GraphPad Prism v6.0c (GraphPad software, Inc., La Jolla, CA, USA) and for analysis by linear regression using bootstrapping [33] using STATA v. 13.1 (College Station, TX, USA). Results from regression analyses are presented as adjusted geometric mean ratios (aGMR) [95% confidence interval (CI)]. Multiplex data values below the lowest concentration were assigned as 1.6 pg/ml. Unstimulated responses were subtracted from antigen-stimulated results and negative values were set to zero. The Mann–Whitney U test was used to compare responses between infants of mothers with and without LTBI and those with and without a BCG scar and correlation between two continuous variables was assessed using the spearman rho test. For the different stimuli, the median maternal and cord blood responses, as well as the associations of infant responses with maternal LTBI and maternal BCG scar were analysed. In addition to looking at single cytokines and chemokines, Principal Component Analysis (PCA) [34] was performed on the cytokine and chemokine variables to summarize them. For this, an average cytokine or chemokine response was worked out for each infant by calculating the mean concentration obtained from the seven different stimuli (after subtracting unstimulated responses). The R programme (v3.2.2. R Foundation for Statistical Computing, Vienna, Austria) was used for further assessment of the associations.

For microarray, raw Illumina probe data were exported from BeadStudio and screened for quality. Pre-processing and statistical analysis was conducted using the R statistical language and various software packages from Bioconductor [35]. Quantile normalization was applied, followed by a log2 transformation. The LIMMA package was used to fit a linear model to each probe and (moderated) t tests or F tests were performed on the groups being compared. To control the expected proportions of false positives, the FDR for each unadjusted p value was calculated using the Benjamini and Hockberg method implemented in LIMMA. The microarray data are available through the National Center for Biotechnology Information Gene Expression Omnibus (GSE87801). Pathway analysis was performed using Gene Set Enrichment Analysis (GSEA), a non-parametric annotation-driven statistical analysis method [36], to assess which biological processes are associated with the different LTBI and BCG scar groups. We tested gene sets from the Molecular signature Database (MsigDB, http://www.broad.mit.edu/gsea/msigdb Hallmark collection (h.all.v5.0.symbols.gmt) which summarize and represent specific well-defined biological states or processes displaying coherent expression. Statistical significance was set for p value below 0.05.

3. Results

3.1. Participant characteristics

The flow of the participants through the study and recruitment details have been described elsewhere [7]. Of the twenty-nine mothers considered for the WBA/Luminex analysis, 12 had a LTBI and 16 had a BCG scar. Three mothers had missing information on BCG scar and were not included in the analysis. Mothers with and without a BCG scar were comparable in terms of age (25 years versus 26 years, p = 0.78), LTBI (31% versus 50%, p = 0.42) and gravidity status (37% versus 50% primigravida, p = 0.70). Their infants were comparable in terms of birth weight (3.09 versus 3.22, p = 0.47) and gender (19% versus 40% male, p = 0.38). Ninety-three mothers were considered for the gene expression microarray, and of these, 21 had a LTBI and 38 had a BCG scar. Mothers with and without a BCG scar were comparable in terms of age (24 years versus 25 years, p = 0.34), LTBI (26% versus 41%, p = 0.26), gravidity status (39% versus 45% primigravida, p = 0.78). Their infants were comparable in terms of birth weight (3.24 versus 3.21, p = 0.77) and gender (40% versus 47% male, p = 0.77) (Table 1).

Table 1
Characteristics of participants by maternal BCG scar status. The figures are given as numbers with percentage (%) in brackets, or as mean values. P value is based on unmatched t test for differences in maternal age and infant birth weight, and a two-sided ...

3.2. The innate immune responses to the different stimuli

The median cytokine and chemokine responses to the different stimuli were analysed. Supplementary Tables 1A and 1B illustrate these for mothers and infants, respectively. There were overall low to moderate concentrations of cytokines, chemokines and growth factors in both maternal and cord blood samples, except for IL-6, IL-8, MCP-1, MIP-1α, MIP-1β and IP-10 (to TLR 7/8 agonist) where concentrations were high across the different stimuli.

3.3. The association between maternal LTBI, maternal BCG scar and innate immune responses in mothers and their offspring

Cytokine and chemokine responses were analysed for associations with maternal LTBI and maternal BCG scar.

For the combined results, maternal responses were not associated with their own BCG scar, except for VEGF where mothers without a BCG scar, compared to those with, had higher concentrations (p = 0.031, Fig. 1A). For IL-4, mothers with a BCG scar, compared to those without, had higher responses (p = 0.012, Supplementary Table 1). Maternal LTBI was positively associated with cord blood IP-10 responses, with an aGMR [95% CI] of 5.10 [1.21, 21.48], p = 0.026 (data not shown).

Fig. 1
The association between maternal BCG scar and infant innate responses. Combined median cytokine or chemokine production following overnight stimulation with lipopolysaccharide (LPS) (toll-like receptor (TLR) 4 agonist), FSL-1 (TLR2/6 agonist), CpG-ODN2006 ...

Cord blood samples obtained from infants of mothers with a BCG scar, compared to those without BCG scar, had overall higher responses to innate stimuli for the following analytes: IFN-γ (aGMR 2.69 [1.15, 6.17]), IL-12p70 (1.95 [1.10, 3.55]), IL-10 (1.82 [1.07, 3.09]), VEGF (3.55 [1.07, 11.48]) and IP-10 (6.76 [1.17, 38.02] There was a similar, but weaker, trend for the proinflammatory cytokines TNF-α (aGMR 1.99 [0.69, 5.89]) and IL-1β (1.55 [0.37, 6.61]). (Fig. 1B, and Supplementary Tables 2 and 3).

The associations between infant responses to the different stimuli and maternal LTBI (Supplementary Figs. 1A and 1B) and maternal BCG scar (Supplementary Figs. 2A and 2B) were analysed. The following CpG-specific cytokine and chemokines were positively associated with maternal LTBI: IL-12p70 (p = 0.014), MCP-1 (p = 0.011) and MIP-1β (p = 0.007) (Supplementary Fig. 1B). Cytokines and chemokines that were positively associated with maternal BCG scar included: IL-10 (p = 0.017) and GM-CSF (p = 0.042) to PAM3Cys-Ser; TNF-α (p = 0.044), IL-2 (p = 0.019), IL-1β (0.005), IL-6 (p = 0.017), IL-10 (p = 0.001), GM-CSF (p = 0.014) and VEGF (p = 0.048) to FSL-1; TNF-α (0.017) to LPS; IFN-γ (p = 0.018), IL-12p70 (p = 0.023), GM-CSF (p = 0.047) to CL097; IL-2 (p = 0.048), IL-1β (0.017), IL-10 (p = 0.040), IL-8 (p = 0.011), GM-CSF (p = 0.027) to Mannan; TNF-α (p = 0.027), IL-12p70 (P = 0.012) and VEGF (P = 0.003) to Curdlan (Supplementary Figs. 2A and 2B).

3.4. Principle component analysis of infant innate immune responses

We observed correlations among the cytokines and chemokines measured and this was summarized using PCA. For the mothers, two principle components (PCs) were identified, which together, accounted for 43% of the variance in the dataset. The first PC explained 25% of the total variance and was characterized by IFN-γ, TNF-α, IL-12p70, IL-1β, IL-6, IL-4, IL-10, IL-13 and the second PC explained a further 18% of the total variance and was characterized by MCP-1, MIP-1α, MIP-1β, IL-8, and IL-17A based on factor loadings > 0.1 (Fig. 2A). Neither Maternal LTBI (data not shown) nor maternal BCG scar (Fig. 2B) was associated with the mothers’ own PC scores.

Fig. 2
Scatterplots of first and second factor loadings for maternal and cord blood, derived from Principal Component Analysis of 17 analytes, showing cytokines and chemokines (A and C), individual mothers (B) and neonates (D). For mothers, the first principal ...

For the infants, two PCs identified accounted for 53% of the variance in the dataset. The first PC explained 39% of the total variance and was characterized by most of the cytokines and growth factors measured (IFN-γ, TNF-α, IL2, IL-12p70, IL-4, IL-13, IL-10, IL-1β, IL-6, IL-8, VEGF and GM-CSF) (Fig. 2C). The second PC explained a further 14% of the total variance and was characterized by MCP-1 and MIP-1β. Infants with a high response in PC1 were born to mothers with a BCG scar (Fig. 2D).

These results are illustrated in Fig. 3. There were no associations between maternal LTBI and levels of PCs in the infants (Fig. 3A and B), and no associations between maternal BCG scar and levels of PCs in the mothers (Fig. 3C and D). Maternal BCG scar was associated with high levels of PC1 in the infants (median level of scores: 1.44 in scar-positive versus −0.94 in scar-negative, p = 0.020, Fig. 3E). There was no association between maternal BCG scar and levels of PC2 in the infants (median level of scores: −0.002 in scar-positive versus 0.754 in scar-negative, p = 0.065, Fig. 3F).

Fig. 3
The association between maternal LTBI, maternal BCG scar and the innate immune responses in mothers and neonates. PCA was used to assess the association between maternal LTBI, maternal BCG scar and infant responses. The association between maternal LTBI ...

The correlations among the cytokines and chemokines measured are shown in Supplementary Table 4.

3.5. Analyses of clusters of innate cytokines and chemokines

In addition to the PCA, we performed a hierarchical bicluster analysis of the innate responses to further identify sets of cytokines and chemokines that might be coordinately expressed in infants of mothers with and without a BCG scar using R programming. Three clusters (C) of cytokines were identified (illustrated in Fig. 4): MCP-1, MIP-1α, MIP-1β, IL-17A (C1), VEGF, GM-CSF, IL-12p70 (C2) and IL-1β, IL-8, TNFα, IFN-γ, IL-2, IL-4, IL-10 (C3). Eleven cytokines formed an additional cluster (C4) that contained high concentrations of the proinflammatory cytokines produced by infants of mothers with a BCG scar.

Fig. 4
Cluster analysis of the innate cytokines and chemokines using the average linkage distance between clusters using R. Clusters go from root to leaf node for each cytokine and for the individual infants. Clusters in between are based on their agglomerative ...

3.6. Gene expression profiles in infants of mothers with and without LTBI, and in the infants of mothers with and without a BCG scar

In order to further examine the associations we found with the innate responses using the Luminex® assay, gene expression microarray analysis was performed using blood obtained from 42 and 51 infants at one and six weeks post-BCG, respectively, using RNA extracted from unstimulated whole blood. Gene expression from infants of mothers with and without LTBI and those with and without a BCG scar were compared. Infants of mothers with LTBI, compared to those of mothers without LTBI, had downregulated interferon and inflammation pathways one week after BCG immunisation (Fig. 5A), but up-regulated interferon and inflammation pathways at six weeks post immunisation (Fig. 5B). In contrast, the interferon and inflammation pathways were both up regulated in infants of mothers with a BCG scar at one (Fig. 6A and Supplementary Fig. 3A) and six (Fig. 6B and Supplementary Fig. 3B) weeks after BCG immunisation.

Fig. 5
Gene Set Enrichment Analysis for the comparison of infants of mothers with and without LTBI. A checkerboard map showing top enriched pathways on y-axis and top leading edge genes (gene members contributing most to the enrichment score) on the x-axis. ...
Fig. 6
Gene Set Enrichment Analysis for the comparison of infants of mothers with and without a BCG scar. A checkerboard map is presented showing top enriched pathways on y-axis and top leading edge genes (gene members contributing most to the enrichment score) ...

4. Discussion

This study reports an unexpected finding about the association between maternal BCG scar and infant responses in a birth cohort. We have shown that infants of mothers with a BCG scar have enhanced proinflammatory responses. The concentrations of proinflammatory cytokines measured in cord blood in response to stimulation with innate stimuli using the Luminex® assay were increased in infants of mothers with a BCG scar. The expression of genes in the interferon and inflammation responses pathways measured using gene transcription microarray was also increased in infants of mothers with LTBI at six week post BCG immunisation, and in infants of mothers with a BCG scar at one and six weeks after BCG immunisation.

Innate immune responses may determine the effectiveness of adaptive responses [37] and lead to either biased [38] or regulatory immune profiles [3941]. The increased responses reported here may therefore impact on immune responses to vaccines administered at birth and on the course of infections and disease in childhood. Further studies of human innate immune profiles in response to immunisation, and during infections and disease, are needed.

There were no associations between maternal BCG scar and the mothers’ own innate immune responses: associations were manifested only in the infants. The presence of a maternal BCG scar was taken to indicate BCG immunisation of a mother during infancy. There are suggestions of positive associations between IFN-γ responses and reactions at the site of BCG immunisation [42,43], and presence of a scar has been associated (in other studies) with protection against LTBI [27,28]. Scar might therefore be a good measure of protective immune responses. However, it is difficult to reconcile how a response to a vaccine administered to mothers in their infancy would exert its effects several years later in the offspring. It is possible that there may be common genetic factors between the mothers and their infants that determine scar formation and subsequent responses in the infants, or that the factors associated with scar formation in the mothers are transmitted to the infants. The lack of association between maternal BCG scar and the mother’s own responses could be attributed to cumulative life-time exposures that alter the initial maternal innate immune responses after BCG immunisation. We did not collect data on scarring in these infants, but an ongoing larger study with a longer follow up will provide the opportunity to assess relationships between scarring and immune responses in mothers and their infants.

The development of a scar is also dependent upon the strain, dose and method of administration of the BCG vaccine [44]. The Danish strain of BCG vaccine, compared to BCG Russia, has been shown to elicit stronger responses in infants one year later and to cause more scarring [2325,45,46], and the intradermal route of administration is associated with the formation of distinctive scars [47,48]. We were unable to ascertain the strain, the dose and the method of administration of BCG vaccine in these women, although the most common strain and the method used in this setting are BCG Russia and the intradermal method, respectively. Since BCG immunisation is administered in the neonatal period, it is difficult to obtain information about BCG immunisation status of adults in a country where hospitals do not routinely record vaccine strain. There is therefore the possibility of misclassification of women based on the presence or absence of a scar. It is possible that the scar-negative women may have been BCG vaccinated without developing a scar, or that scars were lost with time. Our observed differences in infant response may therefore relate either to the mother’s BCG immunisation status or to the quality of the mother’s response to BCG immunisation.

Previous studies have reported the presence [49,50] or absence [51] of maternal cells in cord blood samples. It is therefore possible that the high proinflammatory response observed in cord blood could be due to responses from maternal cells in cord blood, but the method we used for collecting cord blood (by needle and syringe, with no “milking” of the cord, coupled with the use of trained midwives) minimized contamination. Previous tests carried out on maternal and cord blood samples in our studies (comparing levels of β-human chorionic gonadotropin) showed that contamination of cord blood by maternal blood was rare (unpublished data).

Interferon and inflammatory pathways were down-regulated in infants of mothers with LTBI at one week, but up-regulated at six weeks after BCG immunisation; this offers some support to the hypothesis that prenatal exposure to maternal LTBI modifies the infant response to BCG, but the change in direction of effect as the immune response matured was unexpected, and these findings would need to be confirmed in a larger study.

Limitations of the study were its observational and explorative nature, its small sample size relative to the many outcomes assessed. Maternal and infant factors such as maternal age, gravidity status, infant birth weight and gender were not adjusted for since these were not crudely associated with infant responses, and the numbers involved were generally small.

In summary, maternal BCG scar had a stronger association with infant responses than maternal LTBI, with an increased proinflammatory profile of immune responses. The mechanisms that underlie this association need to be further examined in a larger study.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.vaccine.2016.11.079.

Figure s1

Figure s2

Figure s3

Legend, Suppl figures and tables

Table s1

Table s2

Table s3

Table s4

Acknowledgements

We thank the participants and the members of the teams that were involved in this study: the Co-infection Studies Programme (CiSP) and Entebbe General Hospital. In particular, we are grateful to the CiSP field, clinic and laboratory staff and the midwives of Entebbe Hospital.

Funding statement

This work was supported by the European Community’s Seventh Framework Programme (FP7/2007-2013) under EC-GA no. 241642. PAM was also supported by a Commonwealth PhD Fellowship, AME by a Welcome Trust Senior Fellowship (grant number 095778), SC by Wellcome Trust funding (grant number 084344) and PAM and SC by an MRC project grant (MR/K019708).

Footnotes

Contributed by

Author contributions

A.M.E. conceived the study. P.A.M., S.C. and A.M.E. designed the study. P.A.M. coordinated the study (together with S.A.L.), carried out the immunoassays under the supervision of S.C., performed the data analysis under the supervision of E.L.W. and S.N, drafted the manuscript and coordinated the writing of the manuscript. A. F. conducted the microarray assays. G.N. performed the T-SPOT. TB assays. P.A.M., E.L.W., H.M.D., A.M.E., S.C., A.F., G.N., R.P.S., S.P., S.N., and S.C. contributed to discussion and interpretation of the results. P.A.M., A.F., H.M.D., A.M.E. and S.C. participated in writing the manuscript. All the authors read and approved the final manuscript.

Conflict of interest statement

The authors have no associations that might pose a conflict of interest.

References

[1] Colditz GA, Brewer TF, Berkey CS, Wilson ME, Burdick E, Fineberg HV, et al. Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the published literature. Jama. 1994;271:698–702. [PubMed]
[2] ten Dam HG, Hitze KL. Does BCG vaccination protect the newborn and young infants? Bull World Health Organ. 1980;58:37–41. [PubMed]
[3] Trunz BB, Fine P, Dye C. Effect of BCG vaccination on childhood tuberculous meningitis and miliary tuberculosis worldwide: a meta-analysis and assessment of cost-effectiveness. Lancet. 2006;367:1173–80. [PubMed]
[4] Group KPT. Randomised controlled trial of single BCG, repeated BCG, or combined BCG and killed Mycobacterium leprae vaccine for prevention of leprosy and tuberculosis in Malawi. Karonga Prevention Trial Group. Lancet. 1996;348:17–24. [PubMed]
[5] Fine PE. Variation in protection by BCG: implications of and for heterologous immunity. Lancet. 1995;346:1339–45. [PubMed]
[6] Mangtani P, Abubakar I, Ariti C, Beynon R, Pimpin L, Fine PE, et al. Protection by BCG vaccine against tuberculosis: a systematic review of randomized controlled trials. Clin Infect Dis Off Pub Infect Dis Soc Am. 2014;58:470–80. [PubMed]
[7] Mawa PA, Nkurunungi G, Egesa M, Webb EL, Smith SG, Kizindo R, et al. The impact of maternal infection with Mycobacterium tuberculosis on the infant response to bacille Calmette-Guerin immunization. Philos Trans R Soc London Ser B Biol Sci. 2015;370 [PMC free article] [PubMed]
[8] Jones CE, Hesseling AC, Tena-Coki NG, Scriba TJ, Chegou NN, Kidd M, et al. The impact of HIV exposure and maternal Mycobacterium tuberculosis infection on infant immune responses to bacille Calmette-Guerin vaccination. Aids. 2015;29:155–65. [PMC free article] [PubMed]
[9] Kristensen I, Aaby P, Jensen H. Routine vaccinations and child survival: follow up study in Guinea-Bissau, West Africa. Bmj. 2000;321:1435–8. [PMC free article] [PubMed]
[10] Nankabirwa V, Tumwine JK, Mugaba PM, Tylleskar T, Sommerfelt H. Group P-ES. Child survival and BCG vaccination: a community based prospective cohort study in Uganda. BMC Public Health. 2015;15:175. [PMC free article] [PubMed]
[11] Aaby P, Kollmann TR, Benn CS. Nonspecific effects of neonatal and infant vaccination: public-health, immunological and conceptual challenges. Nat Immunol. 2014;15:895–9. [PubMed]
[12] de Castro MJ, Pardo-Seco J, Martinon-Torres F. Nonspecific (heterologous) protection of neonatal BCG vaccination against hospitalization due to respiratory infection and sepsis. Clin Infect Dis Official Pub Infect Dis Soc Am. 2015;60:1611–9. [PubMed]
[13] Aaby P, Roth A, Ravn H, Napirna BM, Rodrigues A, Lisse IM, et al. Randomized trial of BCG vaccination at birth to low-birth-weight children: beneficial nonspecific effects in the neonatal period? J Infect Dis. 2011;204:245–52. [PubMed]
[14] Netea MG, Quintin J, van der Meer JW. Trained immunity: a memory for innate host defense. Cell Host Microbe. 2011;9:355–61. [PubMed]
[15] Benn CS, Netea MG, Selin LK, Aaby P. A small jab – a big effect: nonspecific immunomodulation by vaccines. Trends Immunol. 2013;34:431–9. [PubMed]
[16] Kleinnijenhuis J, Quintin J, Preijers F, Joosten LA, Ifrim DC, Saeed S, et al. Bacille Calmette-Guerin induces NOD2-dependent nonspecific protection from reinfection via epigenetic reprogramming of monocytes. Proc Natl Acad Sci USA. 2012;109:17537–42. [PubMed]
[17] Kleinnijenhuis J, Quintin J, Preijers F, Benn CS, Joosten LA, Jacobs C, et al. Long-lasting effects of BCG vaccination on both heterologous Th1/Th17 responses and innate trained immunity. J Innate Immun. 2014;6:152–8. [PMC free article] [PubMed]
[18] Kleinnijenhuis J, Quintin J, Preijers F, Joosten LA, Jacobs C, Xavier RJ, et al. BCG-induced trained immunity in NK cells: Role for non-specific protection to infection. Clin Immunol. 2014;155:213–9. [PMC free article] [PubMed]
[19] Black RE, Morris SS, Bryce J. Where and why are 10 million children dying every year? Lancet. 2003;361:2226–34. [PubMed]
[20] Floyd S, Ponnighaus JM, Bliss L, Warndorff DK, Kasunga A, Mogha P, et al. BCG scars in northern Malawi: sensitivity and repeatability of scar reading, and factors affecting scar size. Int J Tuberc Lung Dis Official J Int Union Against Tuberc Lung Dis. 2000;4:1133–42. [PubMed]
[21] WHO. Weekly epidemiological record. 2004
[22] Santiago EM, Lawson E, Gillenwater K, Kalangi S, Lescano AG, Du Quella G, et al. A prospective study of bacillus Calmette-Guerin scar formation and tuberculin skin test reactivity in infants in Lima. Peru Pediatrics. 2003;112:e298. [PubMed]
[23] Anderson EJ, Webb EL, Mawa PA, Kizza M, Lyadda N, Nampijja M, et al. The influence of BCG Vaccine strain on mycobacteria-specific and non-specific immune responses in a prospective cohort of infants in Uganda. Vaccine. 2012;30:2083–9. [PMC free article] [PubMed]
[24] Garly ML, Martins CL, Bale C, Balde MA, Hedegaard KL, Gustafson P, et al. BCG scar and positive tuberculin reaction associated with reduced child mortality in West Africa. A non-specific beneficial effect of BCG? Vaccine. 2003;21:2782–90. [PubMed]
[25] Roth A, Gustafson P, Nhaga A, Djana Q, Poulsen A, Garly ML, et al. BCG vaccination scar associated with better childhood survival in Guinea-Bissau. Int J Epidemiol. 2005;34:540–7. [PubMed]
[26] Frankel H, Byberg S, Bjerregaard-Andersen M, Martins CL, Aaby P, Benn CS, et al. Different effects of BCG strains – a natural experiment evaluating the impact of the Danish and the Russian BCG strains on morbidity and scar formation in Guinea-Bissau. Vaccine. 2016;34:4586–93. [PubMed]
[27] He G, Li Y, Zhao F, Wang L, Cheng S, Guo H, et al. The prevalence and incidence of latent tuberculosis infection and its associated factors among village doctors in China. PLoS ONE. 2015;10:e0124097. [PMC free article] [PubMed]
[28] Soysal A, Millington KA, Bakir M, Dosanjh D, Aslan Y, Deeks JJ, et al. Effect of BCG vaccination on risk of Mycobacterium tuberculosis infection in children with household tuberculosis contact: a prospective community-based study. Lancet. 2005;366:1443–51. [PubMed]
[29] Stensballe LG, Nante E, Jensen IP, Kofoed PE, Poulsen A, Jensen H, et al. Acute lower respiratory tract infections and respiratory syncytial virus in infants in Guinea-Bissau: a beneficial effect of BCG vaccination for girls community based case-control study. Vaccine. 2005;23:1251–7. [PubMed]
[30] Roth A, Garly ML, Jensen H, Nielsen J, Aaby P. Bacillus Calmette-Guerin vaccination and infant mortality. Expert Rev Vaccines. 2006;5:277–93. [PubMed]
[31] Jason J, Archibald LK, Nwanyanwu OC, Kazembe PN, Chatt JA, Norton E, et al. Clinical and immune impact of Mycobacterium bovis BCG vaccination scarring. Infect Immun. 2002;70:6188–95. [PMC free article] [PubMed]
[32] Elliott AM, Mawa PA, Webb EL, Nampijja M, Lyadda N, Bukusuba J, et al. Effects of maternal and infant co-infections, and of maternal immunisation, on the infant response to BCG and tetanus immunisation. Vaccine. 2010;29:247–55. [PMC free article] [PubMed]
[33] McGuinness D, Bennett S, Riley E. Statistical analysis of highly skewed immune response data. J Immunol Methods. 1997;201:99–114. [PubMed]
[34] Jolliffe IT. Principle component analysis
[35] Huber W, Carey VJ, Gentleman R, Anders S, Carlson M, Carvalho BS, et al. Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015;12:115–21. [PMC free article] [PubMed]
[36] Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005;102:15545–50. [PubMed]
[37] Medzhitov R, Janeway C., Jr Innate immunity. New Engl J Med. 2000;343:338–44. [PubMed]
[38] Brown J, Baisley K, Kavishe B, Changalucha J, Andreasen A, Mayaud P, et al. Impact of malaria and helminth infections on immunogenicity of the human papillomavirus-16/18 AS04-adjuvanted Vaccine in Tanzania. Vaccine. 2014;32:611–7. [PMC free article] [PubMed]
[39] Webb EL, Mawa PA, Ndibazza J, Kizito D, Namatovu A, Kyosiimire-Lugemwa J, et al. Effect of single-dose anthelmintic treatment during pregnancy on an infant’s response to immunisation and on susceptibility to infectious diseases in infancy: a randomised, double-blind, placebo-controlled trial. Lancet. 2011;377:52–62. [PMC free article] [PubMed]
[40] Malhotra I, Mungai P, Wamachi A, Kioko J, Ouma JH, Kazura JW, et al. Helminth- and Bacillus Calmette-Guerin-induced immunity in children sensitized in utero to filariasis and schistosomiasis. J Immunol. 1999;162:6843–8. [PubMed]
[41] Nookala S, Srinivasan S, Kaliraj P, Narayanan RB, Nutman TB. Impairment of tetanus-specific cellular and humoral responses following tetanus vaccination in human lymphatic filariasis. Infect Immun. 2004;72:2598–604. [PMC free article] [PubMed]
[42] Kemp EB, Belshe RB, Hoft DF. Immune responses stimulated by percutaneous and intradermal bacille Calmette-Guerin. J Infect Dis. 1996;174:113–9. [PubMed]
[43] Hoft DF, Leonardi C, Milligan T, Nahass GT, Kemp B, Cook S, et al. Clinical reactogenicity of intradermal bacille Calmette-Guerin vaccination. Clin Infect Dis Official Pub Infect Dis Soc Am. 1999;28:785–90. [PubMed]
[44] ten Dam HG, Fillastre C, Conge G, Orssaud E, Gateff C, Tanaka A, et al. The use of jet-injectors in BCG vaccination. Bull World Health Organ. 1970;43:707–20. [PubMed]
[45] Roth A, Sodemann M, Jensen H, Poulsen A, Gustafson P, Weise C, et al. Tuberculin reaction, BCG scar, and lower female mortality. Epidemiology. 2006;17:562–8. [PubMed]
[46] Sartono E, Lisse IM, Terveer EM, van de Sande PJ, Whittle H, Fisker AB, et al. Oral polio Vaccine influences the immune response to BCG vaccination. A Nat Exp PloS ONE. 2010;5:e10328. [PMC free article] [PubMed]
[47] ten Dam HG. Research on BCG vaccination. Advances in tuberculosis research Fortschritte der Tuberkuloseforschung Progres de l’exploration de la tuberculose. 1984;21:79–106. [PubMed]
[48] Jin BW, Hong YP, Kim SJ. A contact study to evaluate the BCG vaccination programme in Seoul. Tubercle. 1989;70:241–8. [PubMed]
[49] Hall JM, Lingenfelter P, Adams SL, Lasser D, Hansen JA, Bean MA. Detection of maternal cells in human umbilical cord blood using fluorescence in situ hybridization. Blood. 1995;86:2829–32. [PubMed]
[50] Socie G, Gluckman E, Carosella E, Brossard Y, Lafon C, Brison O. Search for maternal cells in human umbilical cord blood by polymerase chain reaction amplification of two minisatellite sequences. Blood. 1994;83:340–4. [PubMed]
[51] Kogler G, Gobel U, Somville T, Enczmann J, Arkesteijn G, Wernet P. Simultaneous genotypic and immunophenotypic analysis of interphase cells for the detection of contaminating maternal cells in cord blood and their respective CFU-GM and BFU-E. J Hematother. 1993;2:235–9. [PubMed]