1887

Abstract

Background

Our immune system is composed of an innate (germ-line encoded) and adaptive (acquired) arm. It involves specialized hematopoietic cells but also cell-intrinsic (non-hematopoietic) mechanisms, antigen-specific receptors

(T and B cell receptors/immunoglobulins), microbial sensors (pattern recognition receptors), and a complex network of signaling molecules that cooperatively work together to prevent or control infection by invading microorganisms via a variety of effector mechanisms. Both innate and adaptive immune defense mechanisms are tightly regulated, to avoid potentially harmful consequences such as tissue damage to the host, and to maintain tolerance to self as well as to harmless foreign antigens. An imbalance or insufficiency of these immune defense and regulatory mechanisms becomes apparent as clinical disease, either in the form of symptomatic infections (in which case the host immune defense was insufficient to prevent or control infection, dysregulated, or even deficient), or during autoimmune disease (indicating an overly and inappropriate reaction of the immune system to harmless or self antigens). The immune defense and regulatory mechanisms are highly age-dependent, reflecting the different environmental challenges and requirements during the prenatal and postnatal period, and throughout life. Not surprisingly, these age-specific differences are most profound either very early in life (generally described as the ontogeny and maturation of the immune system), or late in life (referred to as immunosenescence). Infants and young children rely heavily on cell-intrinsic and innate immune defenses to control infection. This is because their adaptive immune system has yet to fully mature, and antigen-specific effector B and T cells are absent during the early course of primary infection (the first exposure to a given pathogen). Moreover, we [1–3] and other groups [4–12] have shown substantial age- and gestational age-dependent differences in the innate immune function in early life, which may explain the vulnerability of some newborns, infants and young children to particular pathogens. Indeed, infants and young children are disproportionally burdened by severe infections with a variety of common viruses and bacteria that they become exposed to during birth (e.g. Group B streptococci), in the first few months of life (e.g. respiratory syncytial virus and other ‘common cold’ viruses), or even in utero (e.g. cytomegalovirus). It should be noted that most microorganisms associated with severe diseases in early childhood (or late in life) are common, can also be isolated from asymptomatic individuals and/or individuals with mild disease, and are in most cases unknowingly transmitted through the contact with infected family members or caregivers. It remains largely unclear why some children develop severe clinical illness following exposure to these microbes, while most others exposed to the same microbes develop only mild symptoms. This knowledge gap has been a major obstacle for the development of neonatal vaccines and therapeutic interventions for young pediatric patients. Various studies have indicated an important role of germline genetics (i.e. inborn errors), as well as developmental factors (i.e. epigenetics) in the clinical outcome of primary infection. Here we focus on developmental factors, specifically on age-dependent differences in the genome-wide expression profiles (transcriptome) of whole blood, blood mononuclear cells, or specific blood cell subsets that play an important role in immunity of neonates to infection. To gain further insight into the function and regulation of the immune system in early life, we applied a recently developed web-based systems immunology toolkit [13] to allow the visualization and analysis of a collection of datasets publically available in the Gene Expression Omnibus (GEO) database of the US National Center for Biotechnology Information (NCBI).

Material and Methods

Human neonatal immune defenses are most commonly studied by utilizing placenta-derived cord blood, because it is relatively easy to obtain in larger quantities, and the collection is non-invasive. Less frequently, peripheral blood samples are also obtained from healthy or sick neonates, or fetal blood samples are utilized, in order to study the ontogeny of the immune system. To harness available transcriptomics data and to allow a ‘cross-study’ comparison in order to reveal novel insights from publically available collective data, we queried NCBI's GEO database for search terms “neonate”, “neonatal”, “fetal” or “cord”, as well as “blood”, “PBMC”, “lymphocyte”, “B cell”, ”plasma cell”, ”T cell”, “Treg”, “monocyte”, “dendritic”, ”DC”, “natural killer”, “NK”, NKT”, “ neutrophil”, “erythroblast”, “erythroid”, or lineage markers CD19, CD20, CD3, CD4, CD8, or CD71. Data sets were restricted to those generated from human samples using expression profiling by array or high throughput sequencing. Data sets with keywords including “cancer”, “leukemia”, “lymphoma”, “cell line”, “myeloma”, “hepatocyte”, “mesenchymal” or “endothelial” were excluded. Further, the generated list of data sets using this query was manually curated to exclude studies that are unlikely to reveal insights into neonatal or fetal immune function and regulation. For refining and curating the query result, we used an alternative search engine to NCBI, namely GEOmetadb [14]. This search engine allows the use of SQL language to query a local copy of NCBI's GEO database, and to export any metadata field that is needed for the manual curation of the query result. Next, we adopted a novel approach to knowledge discovery and hypotheses generation for future mechanistic studies [15]. In brief, we first performed a screen aimed at identifying knowledge gaps in one of the above identified data sets that was retrieved from GEO (GSE25087), where the authors generated global gene expression data using sort-purified human fetal and adult T cell subsets, including CD4+CD25+ regulatory T cells (Tregs) and naïve CD4+ T cells [16]. The primary objective of the original study was to assess lineage differences in adult and fetal T cells. Here, we re-assessed data set GSE25087 by ranking the genes according to differential gene expression between fetal and adult Tregs, in order to reveal as yet unknown genes that are involved in immune regulation in early life. Of the genes that ranked among the top 20, we computed a knowledge gap score (KGS) as follows: Existing knowledge was numerated for each of the top-ranking genes by querying NCBI's National Library of Medicine's Pubmed search engine using the official gene symbol, name, as well as aliases. A knowledge gap score was then calculated by dividing the sum of publications with the sum of publications using the same query AND a keyword plus 1 [KGS =  number the Sum(publications)/ (Sum(publications) AND keyword)+1]. For this particular data set, we used “Treg” or “neonatal” as keywords for the denominator, because our objective was to reveal genes that have not yet been associated with Treg function, or with immune regulation during fetal or neonatal life. A high KGS indicated a knowledge gap, whereas a low KGS indicated that the gene in question had already been described to play a functional and/or regulatory role in immunity.

Results and Conclusion

The query of NCBI's GEO database using the keywords and restrictive criteria listed above generated a list of >400 data sets, a selection of our curated data set collection is available at http://developmentalimmunology.gxbsidra.org/dm3/geneBrowser/list (including data set GSE25087). This website is running an instance of an open source web tool called gene expression browser (GXB) [13], which hosts the curated data set. Interestingly, our knowledge gap assessment on data set GSE25087 revealed a high KGS for PMCH, a gene encoding a precursor of melanin-concentrating hormone (MCH), a cyclic neuropeptide. PMCH expression profiles and results from previous studies primarily suggested a neurotransmitter or neuromodulator role for MCH in a broad array of neuronal functions directed toward the regulation of behavior. We further assessed the expression of PMCH in a selection of other GEO data sets that we had identified with the query described above. This revealed PMCH to also be selectively expressed in immature erythroid cells, which were sort-purified from cord blood samples of healthy neonates along with a broad panel of other hematopoietic cell subsets (GSE24759). These immature (nucleated) red blood cells are found in high frequencies in human cord blood samples but decline rapidly in frequency in the first few days of life [17]. Recent studies have shown an important immunosuppressive role of CD71+ erythroid cells, which was associated with increased susceptibility to Listeria monocytogenes and Escherichia coli infection in a mouse model [7, 18]. Our findings indicate that PMCH is expressed by two important immune regulatory cell types of hematopoietic origin and may play a critical, as yet unrecognized role in immune suppression and tolerance during fetal development, which may in turn render newborns vulnerable to infection. To our knowledge, a role of PMCH in immune regulation has not been described, apart from two studies. One of these studies reported PMCH to be induced by TH2 cytokines in human vascular endothelial cells from normal lung blood vessels, suggesting a role in asthma development [19]. The other study found PMCH to be expressed in human Th2 cells which was activation responsive and paralleled with the expression of Th2 cytokine genes [20]. In conclusion, we employed a web-based systems immunology application that allows the visualization and comparative analysis of a collection of public human functional genomics data sets. This proved to be a powerful approach for new discoveries and hypothesis generation, which can then be further validated with in-depth mechanistic ‘wet laboratory’ studies.

References

[1] Marr, N., et al., J Immunol, 2014. 192(3): p. 948–57.

[2] Sharma, A.A., et al., Neonatology, 2014. 106(1): p. 1–9.

[3] You, D., et al., J Leukoc Biol, 2013. 93(6): p. 933–42.

[4] Dowling, D.J. and O. Levy, Trends Immunol, 2014. 35(7): p. 299–310.

[5] Zhang, X., et al., PLoS One, 2013. 8(1): p. e52003.

[6] Corbett, N.P., et al., PLoS One, 2010. 5(11): p. e15041.

[7] Elahi, S., et al., Nature, 2013. 504(7478): p. 158–62.

[8] Ghazal, P., et al., Curr Opin Infect Dis, 2013. 26(3): p. 213–8.

[9] Kollmann, T.R., et al., J Immunol, 2009. 183(11): p. 7150–60.

[10] Kollmann, T.R., et al., Immunity, 2012. 37(5): p. 771–83.

[11] Lavoie, P.M., et al., J Infect Dis, 2010. 202(11): p. 1754–63.

[12] Lissner, M.M., et al., PLoS One, 2015. 10(7): p. e0132061.

[13] Speake, C., et al., J Transl Med, 2015. 13: p. 196.

[14] Zhu, Y., et al., Bioinformatics, 2008. 24(23): p. 2798–2800.

[15] Rinchai, D., et al. [version 1; referees: 3 approved with reservations], 2015. 4:89 (doi: 10.12688/f1000research.6241.1)

[16] Mold, J.E., et al., Science, 2010. 330(6011): p. 1695–9.

[17] Hermansen, M.C., Arch Dis Child Fetal Neonatal Ed, 2001. 84(3): p. F211–5.

[18] Elahi, S., Front Immunol, 2014. 5: p. 376.

[19] Orihara, K., et al., J Allergy Clin Immunol, 2009. 124(3): p. 612–4, 614 e1–2.

[20] Sandig, H., et al., Proc Natl Acad Sci U S A, 2007. 104(30): p. 12440–4.

Loading

Article metrics loading...

/content/papers/10.5339/qfarc.2016.HBPP2361
2016-03-21
2020-02-28
Loading full text...

Full text loading...

http://instance.metastore.ingenta.com/content/papers/10.5339/qfarc.2016.HBPP2361
Loading
This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error