Inter-individual variability and genetic influences on cytokine responses to bacteria and fungi
REVIEW nature medicine
Inter-individual variability and genetic influences on cytokine responses to bacteria and fungi
Yang Li, Marije Oosting, Patrick Deelen, Isis Ricaño-Ponce, Sanne Smeekens, Martin Jaeger, Vasiliki Matzaraki, Morris A Swertz, Ramnik J Xavier, Lude Franke, Cisca Wijmenga, Leo A B Joosten, Vinod Kumar & Mihai G Netea
Key Results
#Stimulation increase inter individual variability in cytokine levels
First, they measured seven cytokines that were induced by ten different stimuli
After strickly quality control of cytokine distributions, they obtained a total of 62 (cytokine–stimulus pairs) different cytokine measurements.
Cytokine production followed a nonGaussian or bimodal distribution, with a few exceptions
In this figure, these graphs are cytokine-CA(Candida) pairs, Blue indicate normal(P is higher than 0.05), Yellow indicate non normal (P is lower than 0.05)and gray indicate distributions were not tested. there are normal distribution and non gaussian normal distribution. there are about 200 samples so I think there are nearly zero value ( not related) and others have peak at 200 because they have relation.
In figure (e,f )RPMI is basal unstimulated state and comparing that ,Individuals exhibited significantly increased inter-individual variability (P < 0.01) in cytokine secretion following stimulation,
- Cytokine responses are organized in a pathogen- specific manner
Correlations between the levels of various cytokines were found in response to stimulation with a certain pathogen, rather than in a cytokine pathway, and this conclusion was validated by additional analyses in the cohort bacterial-induced (lipopolysaccharide (LPS), Escherichia coli and Mycobacterium tuberculosis (MTB))
cytokines (TNF-α, IL-6 and IL-10) were strongly clustered together(red color) and were clearly separated from the fungus-induced (Candida albicans) cytokine cluster.(blue color)
Interferon (IFN)-γ and IL-17 production were exceptions to this rule; however, the magnitude of IFN-γ or IL-17 production by PBMCs from any individual correlated independently of the identity of the pathogen stimulus . This suggests an important evolutionary role for T helper 17 cell responses, which may be a general host defense pathway for both bacteria and fungi.
In addition, the differentiation of naive T cells into TH1 or TH17 effector lymphocytes was under the control of monocyte-derived cytokines.
Thus, a strong or weak monocyte-dependent cytokine production capacity may be associated with strong or weak helper T cell responses.
Following pair-wise correlation of cytokines, however, they observed two clusters, with cluster 1 consisting of cytokines mainly produced by monocytes, and cluster 2 consisting of cytokines known to be released mainly by T cells
However, there was a poor correlation between monocyte-derived cytokine production and T cell cytokine production
This indicates that one particular individual could be a high responder in terms of one set of cytokines, but a low responder for other cytokines associated with strong or weak helper T cell responses.
|
Others show strong correlation, whereas IL17 show poor correlation with others
This indicates that one particular individual could be a high responder in terms
of one set of cytokines, but a low responder for other cytokines
|
# Genome wide cQTL mapping identifies cell-count independednt cQTLs
The cytokine and genotype data available enabled us to study cQTLs for three stimuli: a Gram-negative stimulus (LPS), a mycobacterium (MTB) and a fungus (C. albicans), which provided 18 measurements (three stimulations × six cytokines; IL-6, IL-8, IL-10, IL1-RA, IL-1β and TNF-α).
Raw cytokine levels were first log-transformed and then mapped to genotype data using a linear regression model with age and gender as covariates.
This analysis revealed six significant cQTLs that have Pvlaue lower than 5 × 10−8)
They identified two independent cQTLs for C. albicans–induced IL-6 levels (Fig. 3a–c), two independent cQTLs for MTB-induced IL-8 levels (Fig. 3d–f), one for C. albicans–induced TNF-α and one for LPS-induced IL-10 levels,
(a,d) Manhattan plots showing the genome-wide QTL mapping results for C. albicans–induced IL-6 levels (a) and MTB-induced IL-8 levels (d). Horizontal dashed line corresponds to P < 5 × 10−8.
(b,c,e,f) Box plots showing the association of genotypes at chromosome 9 SNP rs11141235 (b), chromosome 15 SNP rs77181278 with C. albicans–induced IL-6 levels (c), chromosome 1 SNP rs75839717 (e), and chromosome 7 SNP rs74513903 with MTB-induced IL-8 levels
They identified two independent cQTLs for C. albicans–induced IL-6 levels (Fig. a–c),
two independent cQTLs for MTB-induced IL-8 levels (Fig. d–f),
one for C. albicans–induced TNF-α
one for LPS-induced IL-10 levels
First, they analyzed the correlation structure between cell counts and cytokine measurements and observed weak correlations,
For example, candida induced IL6 levels in PBMC showed a weak correlation with monocyte counts, but not with other cell type.
they tested the association of Candida-induced cQTLs with cytokine levels in the cohort after correcting for age, gender and cell counts.
Of those three cQTLs tested, SNP rs11141235, which associated with Candida-induced IL-6 levels, showed a clear replication of association (P = 0.017; Supplementary Table 4) even after correction for monocyte cell counts (P = 0.030).
In contrast, none of the six cQTLs were directly associated with cell counts , suggesting an independent role for genetic variation in regulating cytokine production
To assess whether this observation may denote a genetic component, they tested whether strong cQTLs of one pathogen-induced cytokine could also be associated with cytokine levels induced by other pathogen,
Figure A, QTLs of IL-10 were more likely to be pathogen specific ,
whereas In B, QTLs of other cytokines (IL-6, IL-8 and TNF-α) were more likely to be shared genetic loci that respond to all pathogens
# Correlations between cytokine responses are partially genetically determined
Figure A, SNPs affecting fungus-induced IL-6 and IL-8 were also strongly associated with bacteria-induced IL-6 and IL-8 levels, but not with IL-10 and TNF-α
This indicates that the SNPs associated with IL-6 and IL-8 levels are pathogen independent, and are therefore shared between pathogens
they maybe because these SNPs act on genes or proteins that are downstream of pathogen recognition receptors
The top association for Candida induced IL-6 at the chr9q21 locus provides an illustrative example for a strong shared cQTL (Fig. 4b,c), where the minor allele C at SNP rs11141235 was not only associated with lower C. albicans–induced IL-6 production that show before, but also with LPS-induced (Fig. 4b) and MTB-induced IL-6 production (Fig. 4c).
|
- A cQTL gene GOLM1 on chr9q21 modulates cytokine production
(P values for differential expression of genes (±250 kb around the SNP) selected from genome-wide significant
cQTL loci following different stimulations in human PBMCs (n = 8). P values were from differential expression
analysis using threshold of false discovery rate (FDR) = 0.05 and fold change > 2. Genes were selected on the
basis of their physical positions, which were in a ±250-kb window around the SNP. PBMC stimulations were
carried out for either 4 h or 24 h. Results are shown for both 4-h and 24-h stimulations. Red indicates upregulation,
and blue indicates downregulation; Asterisks indicate genes with suggestive cQTLs in RNA-seq data to identify
the putative causal genes at six significant cQTLs, they tested the expression levels of all genes located within a
500-kb cis-window of the six cQTLs in PBMCs stimulated with different microbial antigens (Fig. 4d). Genes
identified by this differential expression analysis were not cytokine genes, suggesting that the cQTLs that they
identified are mainly trans-QTLs of regulatory genes modulating cytokine production.))
The top associated cQTL, rs11141235, which is on the chromosome 9q21 region, was associated with Candida
induced IL-6 levels (Fig. 3a).
To identify the causal mechanism at this locus, they generated gene expression data by RNA sequencing in
PBMCs from 70 individuals with and without Candida stimulation.
they reconfirmed the significant differential expression of GOLM1 in response to Candida stimulation
|
Next mapped Candida-response cQTL at rs11141235 and at another SNP, rs11141242, in the locus (D′ = 0.95), which is a more frequent polymorphism. The cQTL results revealed that rs11141242 was significantly associated with the expression levels of Golgi membrane protein 1 (GOLM1), and the minor allele was associated with lower levels of GOLM1 (Fig. 5a), with rs11141235 showing a similar trend (Fig. 5b), suggesting the regulatory role of haplotypes in regulating GOLM1 expression.
- GOLM1 cQTL is associated with susceptibility to candidemia
Pathway enrichment analysis on strongly co-expressed genes with GOLM1 during Candida stimulation revealed with cytokine signaling in figure D. The enrichment of genes for cytokine signaling after stimulation could also be a consequence of the stimulation and not necessarily specific to GOLM1 co-expression.
Thus, they tested whether the extent of gene enrichment for IL-6 signaling was specifically linked to GOLM1
|
with Candida stimulation when compared with randomly chosen differentially expressed genes in response to Candida stimulation.
They found a much stronger enrichment of genes co-expressed with GOLM1 for IL-6 signaling than randomly chosen genes
- Correlation between secreted IL-6 levels and genotypes at rs7036187 of 117 candidemia patients
In addition, in patients with candidemia, they also assessed the effects of the rs7036187 polymorphism, a SNP that is associated with susceptibility to disease, on serum IL-6 concentrations.
This SNP was associated with circulating IL-6 concentration, and heterozygous AG genotypes were associated with lower levels of IL-6 (P = 0.015), suggesting that the polymorphisms in the GOLM1 locus may influence Candida-induced cytokines and susceptibility to candidemia.
|
rs7036187 |
- Cytokine QTLs overlap with human-disease-associated SNPs
they tested whether SNPs that were previously associated with human diseases, particularly with infectious diseases, are enriched with cQTLs. they extracted GWAS SNPs from the National Human Genome Research Institute GWAS catalog and binned them into eight categories on the basis of their association with different human phenotypes (Online Methods).
Next they identified all cQTLs that were associated with cytokine levels with P < 0.05 (Supplementary Table 6) and tested whether these cQTLs are linked to GWAS SNPs or their proxies.
In figure A, they found that 61% of infectious-disease-associated SNPs were also cQTLs, and 43% of immune-mediated disease– associated SNPs were also cQTLs.
They used height-associated SNPs as null set of SNPs to determine whether cQTLs are more often associated with a particular human disease. (the reason why height is standard)
Given that a large number of genetic loci were associated with height (for example, 19,798 SNPs from 955 loci), they expected to find a considerable proportion of these SNPs to be co-localized with cytokine QTLs just by chance. Indeed, they found that 38.5% of height-associated SNPs were also cQTLs.
Thus, they used these height-associated SNPs as a background set to test whether other disease-associated SNPs are overrepresented in cQTLs than in height-associated SNPs.
They observed a significant enrichment (P < 9.99 × 10−8) of cQTLs among infectious-disease-associated SNPs. Nearly 60% of heart disease-associated SNPs were also cQTLs, suggesting a role for cytokine pathways in the pathogenesis of cardiovascular diseases.
Conclusions
First, They not only confirmed the non-normal nature of cytokine production distribution of monocyte-derived pro-inflammatory cytokines, but our findings also extend this to T lymphocyte– derived cytokines.
Second, the production capacity of various cytokines strongly correlated with cells being stimulated with a specific pathogen, whereas the correlation was poor when comparing bacterial versus fungal stimulation.
Third, they identified six genome-wide significant cQTLs that influence cytokine responses
Notably, regulation of the pathogen-specific cytokine responses is most likely only partially genetic, as some genetic polymorphisms regulate multiple cytokine responses to a certain pathogen (especially for monocyte-derived production), whereas others regulate the monocyte-derived production of cytokine responses resulting from multiple pathogens (see below). It is therefore likely that nongenetic external factors encountered during one’s lifetime also are important for long-term modulation of cytokine responses, and epigenetic regulation may represent one of the molecular substrates for this process32,33. One notable exception to the rule of pathogen-centric responses is represented by specific lymphocyte responses, such as IL-17 production, which represents a separate, strongly correlated cluster independent of the type of pathogen and may be an important aspect of IL-17 biology
REFERENCES
:
< >Fumagalli, M. et al. Signatures of environmental genetic adaptation pinpoint pathogens as the main selective pressure through human evolution. PLoS Genet. 7, e1002355 (2011).Netea, M.G., Wijmenga, C. & O'Neill, L.A. Genetic variation in toll-like receptors and disease susceptibility. Nat. Immunol. 13, 535–542 (2012). Karlsson, E.K., Kwiatkowski, D.P. & Sabeti, P.C. Natural selection and infectious disease in human populations. Nat. Rev. Genet. 15, 379–393 (2014).Hill, A.V. Evolution, revolution and heresy in the genetics of infectious disease susceptibility.Phil. Trans. R. Soc. Lond. B 367, 840–849 (2012).Jostins, L. et al. Host–microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature 491, 119–124 (2012).Kumar, V., Wijmenga, C. & Xavier, R.J. Genetics of immune-mediated disorders: from genome-wide association to molecular mechanism. Curr. Opin. Immunol. 31, 51–57 (2014).Zhernakova, A. et al. Evolutionary and functional analysis of celiac risk loci reveals SH2B3as a protective factor against bacterial infection. Am. J. Hum. Genet. 86, 970–977 (2010).Fairfax, B.P. et al. Innate immune activity conditions the effect of regulatory variants upon monocyte gene expression. Science 343, 1246949 (2014).Lee, M.N. et al. Common genetic variants modulate pathogen-sensing responses in human dendritic cells. Science 343, 1246980 (2014).Berry, M.P. et al. An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature 466, 973–977 (2010).Ye, C.J. et al. Intersection of population variation and autoimmunity genetics in human T cell activation. Science 345, 1254665 (2014).Raj, T. et al. Polarization of the effects of autoimmune and neurodegenerative risk alleles in leukocytes. Science 344, 519–523 (2014).Vogel, C. & Marcotte, E.M. Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat. Rev. Genet. 13, 227–232 (2012).Naitza, S. et al. A genome-wide association scan on the levels of markers of inflammation in Sardinians reveals associations that underpin its complex regulation. PLoS Genet. 8,e1002480 (2012).Fehrmann, R.S. et al. Gene expression analysis identifies global gene dosage sensitivity in cancer. Nat. Genet. 47, 115–125 (2015).Zambrano-Zaragoza, J.F., Romo-Martínez, E.J., Durán-Avelar, Mde.J., García-Magallanes, N. & Vibanco-Pérez, N. TH17 cells in autoimmune and infectious diseases. Int. J. Inflamm.2014, 651503 (2014).Mills, K.H., Dungan, L.S., Jones, S.A. & Harris, J. The role of inflammasome-derived IL-1 in driving IL-17 responses. J. Leukoc. Biol. 93, 489–497 (2013).Genome of the Netherlands Consortium. Whole-genome sequence variation, population structure and demographic history of the Dutch population. Nat. Genet. 46, 818–825 (2014).Kladney, R.D. et al. GP73, a novel Golgi-localized protein upregulated by viral infection.Gene 249, 53–65 (2000).Kumar, V. et al. Immunochip SNP array identifies novel genetic variants conferring susceptibility to candidemia. Nat. Commun. 5, 4675 (2014).Welter, D. et al. The NHGRI GWAS Catalog, a curated resource of SNP–trait associations.Nucleic Acids Res. 42, D1001–D1006 (2014).Orrù, V. et al. Genetic variants regulating immune cell levels in health and disease. Cell 155,242–256 (2013).Roederer, M. et al. The genetic architecture of the human immune system: a bioresource for autoimmunity and disease pathogenesis. Cell 161, 387–403 (2015).Brodin, P. et al. Variation in the human immune system is largely driven by nonheritable influences. Cell 160, 37–47 (2015).Rubicz, R. et al. Genome-wide genetic investigation of serological measures of common infections. Eur. J. Hum. Genet. 23, 1544–1548 (2015).Ghazalpour, A. et al. Comparative analysis of proteome and transcriptome variation in mouse. PLoS Genet. 7, e1001393 (2011).Jüschke, C. et al. Transcriptome and proteome quantification of a tumor model provides novel insights into post-transcriptional gene regulation. Genome Biol. 14, r133 (2013).Battle, A. et al. Genomic variation. Impact of regulatory variation from RNA to protein.Science 347, 664–667 (2015).Wu, L. et al. Variation and genetic control of protein abundance in humans. Nature 499,79–82 (2013).Endres, S., Ghorbani, R., Lonnemann, G., van der Meer, J.W. & Dinarello, C.A.Measurement of immunoreactive interleukin-1β from human mononuclear cells: optimization of recovery, intra-subject consistency and comparison with interleukin-1α and tumor necrosis factor. Clin. Immunol. Immunopathol. 49, 424–438 (1988).Endres, S. et al. In vitro production of IL-1β, IL-1α, TNF and IL-2 in healthy subjects: distribution, effect of cyclooxygenase inhibition and evidence of independent gene regulation.Eur. J. Immunol. 19, 2327–2333 (1989).Saeed, S. et al. Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345, 1251086 (2014).Cheng, S.C. et al. mTOR- and HIF-1α-mediated aerobic glycolysis as metabolic basis for trained immunity. Science 345, 1250684 (2014).\Kayama, H., Nishimura, J. & Takeda, K. Regulation of intestinal homeostasis by innate immune cells. Immune Netw. 13, 227–234 (2013).van de Veerdonk, F.L. et al. STAT1 mutations in autosomal dominant chronic mucocutaneous candidiasis. N. Engl. J. Med. 365, 54–61 (2011).Liu, L. et al. Gain-of-function human STAT1 mutations impair IL-17 immunity and underlie chronic mucocutaneous candidiasis. J. Exp. Med. 208, 1635–1648 (2011).Milner, J.D. et al. Impaired TH17 cell differentiation in subjects with autosomal dominant hyper-IgE syndrome. Nature 452, 773–776 (2008).Farh, K.K. et al. Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518, 337–343 (2015).Kladney, R.D., Cui, X., Bulla, G.A., Brunt, E.M. & Fimmel, C.J. Expression of GP73, a resident Golgi membrane protein, in viral and nonviral liver disease. Hepatology 35,1431–1440 (2002).Liu, X. et al. Golgi protein 73(GP73), a useful serum marker in liver diseases. Clin. Chem. Lab. Med. 49, 1311–1316 (2011).Liang, H. et al. Interleukin-6 and oncostatin M are elevated in liver disease in conjunction with candidate hepatocellular carcinoma biomarker GP73. Cancer Biomark. 11, 161–171(2012).Netea, M.G. et al. A semi-quantitative reverse transcriptase polymerase chain reaction method for measurement of mRNA for TNF-α and IL-1β in whole-blood cultures: its application in typhoid fever and exentric exercise. Cytokine 8, 739–744 (1996).van Crevel, R. et al. Disease-specific ex vivo stimulation of whole blood for cytokine production: applications in the study of tuberculosis. J. Immunol. Methods 222, 145–153(1999).Hirschfeld, M., Ma, Y., Weis, J.H., Vogel, S.N. & Weis, J.J. Cutting edge: repurification of lipopolysaccharide eliminates signaling through both human and murine toll-like receptor 2.J. Immunol. 165, 618–622 (2000) Sutmuller, R.P. et al. Toll-like receptor 2 controls expansion and function of regulatory T cells.J. Clin. Invest. 116, 485–494 (2006) Shah, T.S. et al. optiCall: a robust genotype-calling algorithm for rare, low-frequency and common variants. Bioinformatics 28, 1598–1603 (2012).Deelen, P. et al. Genotype harmonizer: automatic strand alignment and format conversion for genotype data integration. BMC Res. Notes 7, 901 (2014).Delaneau, O., Zagury, J.F. & Marchini, J. Improved whole-chromosome phasing for disease and population genetic studies. Nat. Methods 10, 5–6 (2013).Howie, B., Marchini, J. & Stephens, M. Genotype imputation with thousands of genomes. G3 (Bethesda) 1, 457–470 (2011).Deelen, P. et al. Genome of Netherlands Consortium. Improved imputation quality of low-frequency and rare variants in European samples using the 'Genome of the Netherlands'.Eur. J. Hum. Genet. 22, 1321–1326 (2014).Smeekens, S.P. et al. Functional genomics identifies type I interferon pathway as central for host defense against Candida albicans. Nat. Commun. 4, 1342 (2013).Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).Love, M.I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). Shabalin, A.A. Matrix eQTL: ultrafast eQTL analysis via large matrix operations.Bioinformatics 28, 1353–1358 (2012).
