Infection and RNA-seq analysis of a zebrafish tlr2 mutant shows a broad function of this toll-like receptor in transcriptional and metabolic control and defense to Mycobacterium marinum infection
The function of Toll-like receptor 2 (TLR2) in host defense against pathogens, especially Mycobacterium tuberculosis (Mtb) is poorly understood. To investigate the role of TLR2 during mycobacterial infection, we analyzed the response of tlr2 zebrafish mutant larvae to infection with Mycobacterium marinum (Mm), a close relative to Mtb, as a model for tuberculosis. We measured infection phenotypes and transcriptome responses using RNA deep sequencing in mutant and control larvae.
tlr2 mutant embryos at 2 dpf do not show differences in numbers of macrophages and neutrophils compared to control embryos. However, we found substantial changes in gene expression in these mutants, particularly in metabolic pathways, when compared with the heterozygote tlr2+/− control. At 4 days after Mm infection, the total bacterial burden and the presence of extracellular bacteria were higher in tlr2−/− larvae than in tlr2+/−, or tlr2+/+ larvae, whereas granuloma numbers were reduced, showing a function of Tlr2 in zebrafish host defense. RNAseq analysis of infected tlr2−/− versus tlr2+/− shows that the number of up-regulated and down-regulated genes in response to infection was greatly diminished in tlr2 mutants by at least 2 fold and 10 fold, respectively. Analysis of the transcriptome data and qPCR validation shows that Mm infection of tlr2 mutants leads to decreased mRNA levels of genes involved in inflammation and immune responses, including il1b, tnfb, cxcl11aa/ac, fosl1a, and cebpb. Furthermore, RNAseq analyses revealed that the expression of genes for Maf family transcription factors, vitamin D receptors, and Dicps proteins is altered in tlr2 mutants with or without infection. In addition, the data indicate a function of Tlr2 in the control of induction of cytokines and chemokines, such as the CXCR3-CXCL11 signaling axis.
The transcriptome and infection burden analyses show a function of Tlr2 as a protective factor against mycobacteria. Transcriptome analysis revealed tlr2-specific pathways involved in Mm infection, which are related to responses to Mtb infection in human macrophages. Considering its dominant function in control of transcriptional processes that govern defense responses and metabolism, the TLR2 protein can be expected to be also of importance for other infectious diseases and interactions with the microbiome.
KeywordsTuberculosis Mycobacterium TLR2 RNAseq RNA deep sequencing Transcriptomics Zebrafish
Confocal laser scanning microscopy
Differential expressed genes
Days post fertilization
Days post injection
Fluorescence-activated cell sorting
False discovery rate
Gene-Set Enrichment Analysis
Hours post fertilization
Hours post injection
Non-sense mediated mRNA decay
Quantitative polymerase chain reaction
Deep sequencing of cDNA derived from polyA RNA
Sub-Network Enrichment Analysis
Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis (TB), which infects nearly 23% of the world’s population, and kills about 1.6 million people annually (WHO Global Tuberculosis Report 2018). TB is characterized by the formation of granulomas, aggregates of infected macrophages and other immune cells, not only in the lung but also in other tissues and organs . The formation of granulomas is the result of a concerted action of host innate and adaptive immunity [2, 3, 4].
Innate immune responses play a critical role in defense against TB infection in the host, and for a major part these processes are mediated by Toll-like receptors (TLRs), a conserved family of pattern recognition receptors. TLR2 is one of the most widely reported members of the TLR family to be involved in defense against Mtb by virtue of its recognition of cell wall-associated components associated with this pathogen [5, 6, 7]. Following mycobacterial infection in human cell cultures, TLR2 dimerizes with TLR1 or TLR6, and recognizes mycobacterial components such as cell wall glycolipids LAM and LM , 38-kDa and 19-kDa mycobacterial glycoprotein (LpqH) [9, 10, 11], phosphatidylinositol mannoside (PIM) , and triacylated (TLR2/TLR1) [13, 14] or diacylated (TLR2/TLR6) lipoproteins [15, 16]. Then, these heterodimers recruit the MYD88 and TIRAP (MAL) proteins to activate the IRAK (1 and 4)/TRAF6/IKK (α or β) cascade, which subsequently leads to the ubiquitination of IκBα and the activation of transcription factor NF-κB or AP-1 to induce the expression of host defense genes and cytokine and chemokine responses [17, 18]. Once released after further processing, these cytokines and chemokines attract migration of macrophages and neutrophils to the infection site and activate the microbicidal functions of these cells.
TLR2 has been shown to be important for granuloma formation in Mtb infection in mouse infection model . Other studies have reported that Tlr2−/− mice lose control to high dose infection of Mtb delivered by aerosol administration and show higher susceptibility to Mtb infection compared to the wild type [20, 21]. However, using lower doses of infection of 100 bacteria, there are controversial results as to the function of TLR2 in defense to Mtb in rodent models [20, 21, 22]. Furthermore, the function of TLR2 in susceptibility to Mtb is still unclear because several independent studies have reported polymorphisms of TLR2 (TLR1 and TLR6) in humans that have been linked with TB susceptibility, whereas others have been unable to find such links [23, 24].
Several studies suggest that the interaction between TLR2 and Mtb or other pathogens does not always promote the killing of bacteria, but can in fact be part of the pathogens’s strategy to evade the immune system [25, 26, 27]. TLR2 has been reported to inhibit MHC-II expression on the surface of murine macrophages, thereby preventing presentation of Mtb antigens, which may allow intracellular Mtb to evade immune surveillance and maintain chronic infection [11, 28]. Furthermore, LprG from Mtb was reported to inhibit human macrophage class II MHC antigen processing through TLR signaling . In other infection systems, Tlr2 mutant mice have an increased resistance against infection of Candida albicans  and Yersinia pestis . It has been proposed that this phenotype is the result of Tlr2-dependent induction of the anti-inflammatory cytokine IL-10 . Furthermore, TLR2 activation inhibits the release of IL-12 via activation of the cFos transcription factor. It also has been shown that IFN-γ or IFN-γ-induced signals  are inhibited after infection of murine macrophages in a Tlr2-dependent fashion . These effects show that TLR2 activation can yield a bias to T helper Type 2 (Th2) cells , and by breaking the Th1/Th2 balance can lead to less Th1 type responses and reducing the killing of intracellular pathogens. However, most of the molecular mechanisms underlying TLR2 functions remain unknown and a better understanding of the TLR2-mediated immune response and immune evasion can help in planning prevention and therapy strategies against Mtb infection.
Animal models have shown their power in studies of the mechanisms of interaction of host and TB pathogens, and discovering new anti-TB drugs. Zebrafish adult and larvae models have become a useful complement for rodent studies, for three important reasons. First, zebrafish have a 3–4 weeks separation stage between development of innate and adaptive immunity after fertilization [34, 35], which gives the possibility to study the host innate immune response to infection in the absence of adaptive immune responses. Second, zebrafish can be infected by Mycobacterium marinum (Mm), a natural pathogen of cold-blooded vertebrates and a close relative of Mtb, which can induce granuloma formation in zebrafish, similar to human TB [36, 37]. Third, the transparent larvae are ideal for imaging the early steps of the infection process in real time. Hence, zebrafish has earned its place of being a versatile tuberculosis model [1, 38, 39]. Although zebrafish has no lungs and therefore the route of infection is through other tissues, it has been shown that the parasitism of immune cells and the aggregation of infected cells into granulomatous aggregates are highly similar to that in mammals . Nevertheless, there are some notable difference with the progression of the tuberculosis infection processes in mice .
The TLR signaling pathway in zebrafish appears to be highly conserved with that in mammals . The zebrafish tlr2 gene and the genes encoding its adaptor proteins are highly similar in sequence to those of mammals and mutation of the down-stream signaling gene myd88 leads to phenotypes that correlate with those seen in rodent and human cell culture studies [43, 44]. In our previous study, we demonstrated that the mammalian TLR2 ligand Pam3CSK4, a synthetic triacylated lipopeptide that mimics the triacylated lipoprotein of mycobacteria, could also specifically activate the zebrafish Tlr2 pathway, inducing fosl1a and cebpb gene upregulation . In the current study, to further explore the involvement of Tlr2 in Mm infection, we conducted infection studies using tlr2 mutant zebrafish. We found that a tlr2 mutation led to an overall increased Mm bacterial burden, corresponding with reduced presence of macrophages in the granulomatous aggregates and more extracellular growth. These results indicate that Tlr2 plays an important role in protecting the host during the early stage of mycobacterial infection. In addition, we performed RNA deep sequencing (RNAseq) and determined a Tlr2-specific gene list for the response to Mm infection. This analysis revealed that most of transcriptional downregulation caused by Mm infection in control animals was abrogated by tlr2 mutation, in addition to a dampening effect on the upregulation of transcription factors and inflammatory genes.
Characterisation of a tlr2 mutant zebrafish line
To confirm whether tlr2 mutation blocks its downstream pathway, we analyzed the gene expression profiles of zebrafish treated with the TLR2 agonist, Pam3CSK4 similarly as in our previous work , now also including the heterozygote mutant (tlr2+/−). Pam3CSK4 was injected into the blood island of zebrafish embryos at 27 h post fertilization (hpf). One hour after injection (hpi), we collected samples and performed qPCR to analyze the expression levels of CCAAT/enhancer-binding protein beta (cebpb) and FOS Like Antigen 1a (fosl1a), previously shown to be specific targets of Tlr2 signaling . In wild-type siblings and tlr2+/ sa19423 heterozygotes (tlr2+/−), the expression levels of cebpb and fosl1a, as well as the inflammatory gene il1b were significantly induced upon Pam3CSK4 injection, whereas tlr2−/− showed no significant response (Fig. 1b-d). tlr2+/− animals showed a lower induction of these marker genes than the wild-types, indicating that there is an effect of the tlr2 mutation even in the heterozygote. To confirm that these results are specific to the Tlr2 pathway, we injected flagellin, a Tlr5 agonist, into 27 hpf embryos. Flagellin induced il1b expression, but not cebpb and fosl1a expression in wild-type siblings, tlr2+/− and tlr2−/− larvae (Fig. 1e-g). Overall, these data show that tlr2 mutation specifically blocks the Tlr2 downstream pathways.
To determine if immune cell development was affected by tlr2 mutation, we used the double-transgenic line tlr2+/+ Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) and tlr2−/− Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) to count the number of macrophages and neutrophils at 2 dpf. The results show that there is no significant difference in the number of macrophages and neutrophils at 2 dpf between the wild type and mutant (Fig. 1j, k).
Comparison of gene expression profiles of tlr2 homozygote and heterozygote mutants in the absence of infection
To investigate whether the basal expression level differences of genes might be relevant to the Tlr2 pathway, we performed GO analysis of a set of genes that were expressed differently with the very low FDR adjust P value of 10− 10 (Additional file 13: Table S2). Gene ontology analyses of the gene sets that differ with an FDR of 10− 10 in both analyses show that there is a large enrichment of genes belonging to the GO terms related to neural development (Additional file 13: Table S2). In addition, we have also analyzed differences with a fold change criterion of two and FDR of 0.05. These analyses indicated that under the GO category “Transcription factor genes” only two categories of genes including the c-Maf transcription factor were present (Additional file 14: Table S3). Pathway analysis shows that genes that function in glucose metabolism are differentially expressed in the tlr2 homozygote versus heterozygote mutant (Additional file 3: Figure S3).
Tlr2-specific gene expression profiles after Mm infection
To show the relationship between DEGs, we constructed networks based on common expression targets in the 97 up- and 92 down-regulated genes . The involved networks with the up-regulated genes contain lipc, nr0b2, hmgcr, itln1, fgf23, vdr, irak3 and tlr8 (Additional file 8: Figure S8). In tlr2+/− controls we observed positively regulated hmgcr, fgf23, vdr and tlr8, and negatively regulated nr0b2, whereas tlr2−/− showed an opposite regulation for most of these genes. However, lipc and irak3, were positively regulated in both tlr2+/− and tlr2−/−. For the 92 downregulated genes, a network containing nr0b2 and mafb was constructed (Additional file 9: Figure S9). nr0b2 and mafb were positively regulated in tlr2−/− zebrafish and down regulated in tlr2+/−.
Enrichment analysis of Tlr2 specific genes after Mm infection
To link our data to gene sets defined based on prior biological knowledge (including GO), we conducted Gene-Set Enrichment Analysis (GSEA) of the differently expressed genes. This method derives its power by focusing on gene sets, that is, groups of genes that share common biological function, chromosomal location, or regulation . Because the number of predicted gene sets were too large for our analysis method (more than 1000 with P < 0.05), we focused on the gene sets related to metabolic, immunological and inflammation pathways. GSEA predicted 61 pathways in tlr2+/− and 67 pathways in tlr2−/− zebrafish responsive to Mm infection (Additional file 15: Table S4A and 4B). Interestingly, most of these pathways are common in tlr2+/− and tlr2−/−, however some of them are detected as being anti-correlated in regulation, including some pathways underlying natural killer cell functions and omega-6-fatty acid metabolism (Additional file 15: Table S4C). We also performed Sub-Network Enrichment Analysis (SNEA)  to identify possible key genes that are responsible for the difference in response of the tlr2+/− and tlr2−/− group to Mm infection (P < 0.05). SNEA predicted 565 and 503 pathways for tlr2+/− and tlr2−/− zebrafish that are linked to the response to infection, respectively (Additional file 15: Table S4D and 4E), and 264 and 202 of them are specific for the response in tlr2+/− and tlr2−/− fish, respectively (Additional file 15: Table S4F). Since the RNAseq was conducted with the total RNA from whole body of zebrafish, it is not possible to define which pathways are involved in macrophage functions related to tlr2 expression. Therefore, we analyzed previously published DNA microarray data (GDS4781) of human macrophages transfected with Mtb from Gene Expression Omnibus . SNEA analysis of human macrophages shows that 659 pathways are linked to Mtb infection (Additional file 15: Table S4G). By comparing the human SNEA result to tlr2+/− specific pathways in zebrafish, 56 pathways were defined as tlr2+/−-specific (Additional file 15: Table S4H). Of these, the pathway of TLR8, which has the lowest P-value in both zebrafish and human enrichment (Additional file 15: Table S4H), and its network is depicted as example in Additional file 10: Figure S10. Overall, these analyses show that tlr2−/− mutants have a strongly altered immune response of which many pathways that are linked to human tuberculosis are differently responding.
TLR2 has been shown to play a role in host defense against Mtb in several rodent studies but its role in host innate immunity during infection is still not clear. Moreover, little is known about the systemic regulation of down-stream signaling of TLR receptors in animal models. As part of this study we generated a tlr2 zebrafish mutant to study Tlr2 function in the innate immune system during mycobacterial infection. Our results show a function of Tlr2 in defense against mycobacterial infection. This function is apparent by a larger number of extracellular bacteria and a lower number of granulomas in the tlr2 mutant compared to the controls. This result is in agreement with in vivo studies in mice infected with tuberculosis bacteria that show a defect in granuloma formation in a Tlr2 mutant . To obtain explanations and genetic markers for further studies of the effect of the tlr2 mutation on infection we performed deep RNA sequencing to study the whole transcriptome profile in our mycobacterial infection model at the systems level. There are only a few RNA sequencing results of tuberculosis studies in rodents [57, 58], human serum , human pulmonary epithelial cells , and bovine systems.
To characterize the effect of tlr2 mutation in the absence of infection, we compared the transcriptome of homozygous mutant larvae with that of heterozygote larvae, thereby excluding the effect of non-dominant background polymorphisms that might have resulted from ENU mutagenesis. The results show large differences between these genotypes, for instance in genes involved in glycolysis (Additional file 3: Figure S3). In accordance, a previous study in human peripheral blood mononuclear cells and an in vivo mouse model for tuberculosis showed a switch in host cellular metabolism toward aerobic glycolysis after mycobacterial infection that is dependent on TLR2 . In agreement, our previous results in zebrafish larvae suggested a role in metabolism also for Myd88, the adaptor in TLR signaling . The largest category of genes that was significantly affected in tlr2 homozygous mutants was “neurological system process” (Additional file 13: Table S2). Many recent studies show that a mutation in Tlr2 in mice resulted in effects of neuronal development and responses to injury [62, 63]. Some of these studies show a connection of Tlr2 deficiency to neuronal defects that could be related to Il10 function and autophagy [64, 65]. When focusing on the signaling pathways that could be involved, we observed that there was a significant effect in the GO category of transcription factors, namely the c-Maf factors that totals up to 546 representatives that were affected (Additional file 14: Table S3). The members of the Maf family of transcription factors, c-Maf and Mafb are specifically expressed in monocyte and macrophage lineages [66, 67], and in addition, c-Maf is also expressed in T helper cells . c-Maf was also reported to directly regulate IL-10 expression induced by LPS in macrophages . Double deficiency of Mafb/c-Maf promotes self-renewal of differentiated macrophages . We did not detect differences in macrophage numbers between tlr2 mutants and controls, but it remains to be further investigated if macrophage differentiation is altered as a result of Tlr2 deficiency.
Recently, we have shown that Tlr2 and its adaptor MyD88 are important for the response of the zebrafish host to the microbiome . Therefore, it is very well possible that the transcriptional differences we find in the tlr2 mutant versus the controls are due to an aberrant response to the microbiome. Similarly, we have found that mutation in myd88 leads to a large difference in the zebrafish gene expression profile, even in the absence of pathogenic infection . This difference appeared to be dependent on the presence of a microbiome. In future work we therefore aim to study the transcriptome response of the tlr2 mutant under germfree conditions and in the presence of a microbiome under gnotobiotic conditions and investigate whether the dysregulation of neural development pathways in tlr2 mutants might be linked to control of the gut-brain axis [70, 71, 72, 73, 74]. Such studies could also show whether the composition of the microbiome affects the progression of mycobacterial infection. This could be clinically relevant, considering that recent studies show a correlation of the occurrence of tuberculosis with aberrations of the microbiome in human patients .
The analysis of differential gene expression during infection showed that there is a very pronounced effect of mutation of the tlr2 gene. However, the expression of many genes is independent of tlr2 as expected, since also other TLR receptors play a role in recognition of mycobacterial infections [76, 77, 78, 79]. With respect to the number of genes affected, the strongest effect was observed in genes that are down regulated during infection, since this category was strongly diminished in the tlr2 mutant (Fig. 6a, b and Additional file 5: Figure S5). This observation suggests that Tlr2 has an important function in anti-inflammatory responses, in line with previous reports of studies in mice that showed a strongly decreased anti-inflammatory response in Tlr2 knockouts [77, 78]. In a further analysis of the quantitative effects on the differences in expression levels of genes in both the up regulated and down regulated groups, we selected a number of genes that were most significantly affected in GO analysis (Fig. 6d, e). This GO analysis showed many groups to be affected with the immune response as the biggest group. We also performed GSEA and SNEA analyses (Additional file 8: Figure S8, Additional file 9: Figure S9, Additional file 10: Figure S10 and Additional file 15: Table S4) showing that the tlr2 mutant has a very different immune response than heterozygote controls. Furthermore, we show that many signaling pathways that have been reported to be linked to tuberculosis infection in humans are differentially regulated in our data set. Most significantly, activation of the Tlr8 pathway was strongly affected (Additional file 10: Figure S10). This evidence suggests that the Tlr2 signaling is strongly connected with Tlr8 function. TLR8 mutations (polymorphisms) increase susceptibility to mycobacteria in the human population [79, 80]. Recently it was shown that TLR8 is a sensor of various bacteria in human primary monocytes . This indication that TLR8 is a more important sensor in the antibacterial defense system than previously known might explain our SNEA results. In addition, our analysis revealed differential expression of three other interesting categories of immune genes, discussed below.
The vitamin D receptor pathway genes that are normally up-regulated during infection in zebrafish larvae were down regulated in the tlr2 mutant. Furthermore, pathway analysis (Additional file 8: Figure S8 and Additional file 10: Figure S10) also implicated the expression of the Tlr8 pathway connected to vitamin D signaling as being strongly affected in the tlr2 mutant. Vitamin D has been shown to be an important regulatory factor during tuberculosis  and has been linked previously to TLR2 function in studies in cell cultures . Therefore, aberrant vitamin D signaling could be a major contributing factor to the hyper-susceptibility phenotype of tlr2 mutants in Mm infection.
Tlr2 has been shown to be essential for the up-regulation of a group of genes that encode the Diverse Immunoglobulin Domain-Containing Proteins (DICPS) (Fig. 6d). This group is as novel multigene family encoding diversified immune receptors. Haire et al.  reported that recombinant DICP Ig domains bind lipids and lipid extracts of different bacteria, including Mtb and Mm, a property shared by mammalian CD300 and TREM family members. In the down-regulated set also several DICP members appear to be dependent on Tlr2, such as dicp1.17, dicp3.3, that are linked to the GO term insulin-like growth factor binding. These correlations might relate to functions of Tlr2 in other processes such as the control of diabetes type II by gut microbiota. However, the DICP gene family lacks easily recognizable genetic homologs in mammals, making a translation to a function in mammalian tuberculosis and other diseases currently not yet possible .
Another highly relevant category of genes of which the induction or repression during infection is dependent on Tlr2 includes the chemokines. In a previous study of our laboratory, Torraca et al. demonstrated the function of the Cxcr3-Cxcl11 axis in macrophage recruitment to infection foci and showed that disruption of this axis by cxcr3.2 mutation increases the resistance to mycobacterial infection . Furthermore, by Fluorescence-activated cell sorting (FACS) of Mm-infected cells, we recently showed that cxcl11a is a robust marker of infected macrophages . The infection induced expression of this chemokine is dependent on Myd88, the common adaptor of the majority of Tlrs, including tlr2 . In agreement, the tlr2 mutant shows a significant lower expression of cxcl11aa and also of an related chemokine, cxcl11ac, during Mm infection (Fig. 5). Considering the large number of other chemokines that are controlled by Tlr2 during infection, it is clear that the integrative network of connections cannot yet be understood from these expression studies and need more detailed functional analyses, e.g. by combinations of different mutations or directed studies on responses to chemokines as shown by Torraca et al . However, we can state that the phenotype of the tlr2 mutant, at the infection level, and the level of transcriptional control such as the mentioned effects on regulation of MafB/c-Maf and chemokines shows a clear connection with macrophage chemotaxis. These evidences do not exclude that Tlr2 has many other functions during infection such as phagocytosis. For instance, Blander et al.  and Rahman et al.  showed that phagocytosis of bacteria and phagosome maturation are impaired in the absence of TLR signaling. Therefore, the large number of unannotated genes of which the expression during infection is dependent on Tlr2 is also worth studying in more detail in future studies.
Our study shows that Tlr2, as a part of innate immunity, plays an important role in controlling mycobacterial infection as observed on the transcriptome and infection level. This function may be mediated by several mechanisms, including a general attenuation of the inflammatory response, reduced mycobacterial dissemination by dampening of CXCR3-CXCL11 signaling, and anti-mycobacterial effects like vitamin D signaling. Our results show that Tlr2 is a major Tlr family member upstream of Myd88 that activates the CXCR3-CXCL11 signaling axis. The tlr2 mutant is therefore a valuable model for further studies using published infection models for other pathogens and the study of the interactions with gut microbiota in zebrafish larvae.
All zebrafish were handled in compliance with the local animal welfare regulations and maintained according to standard protocols (zfin.org). Larvae were raised in egg water (60 g/ml Instant Ocean sea salts) at 28.5 °C. For the duration of bacterial injections, larvae were kept under anesthesia in egg water containing 0.02% buffered 3-aminobenzoic acid ethyl ester (Tricaine, Sigma-Aldrich, the Netherlands). The culture of zebrafish with mutations in immune genes was approved by the local animal welfare committee (DEC) of the University of Leiden (protocol 14,198). All protocols adhered to the international guidelines specified by the EU Animal Protection Directive 2010/63/EU.
The tlr2sa19423 mutant line (ENU-mutagenized) was obtained from the Sanger Institute Zebrafish Mutation Resource (Hinxton, Cambridge, UK) and shipped by the Zebrafish Resource Center of the Karlsruhe Institute of Technology. The mutant allele was identified by sequencing. Heterozygous carriers of the mutation were outcrossed three times against wild type (AB strain), and were subsequently incrossed three times. Heterozygous fish of the resulting family were used to produce embryos. Homozygous mutants were outcrossed to the Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) double transgenic line [88, 89], and the offspring with GFP and mCherry fluorescence were subsequently incrossed to produce the Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) line.
Bacterial strain preparation
The bacterial strain, Mycobacterium marinum m20 (Mma20) expressing mCherry fluorescent protein , was used in this study. For the infection to zebrafish larvae, the bacteria were prepared as previously described . The infection inoculum was prepared in 2% polyvinylpyrrolidone40 solution (Calbiochem, the Netherlands), and 150 colony-forming units (CFU) of bacteria were injected into the blood stream at 28 h post fertilization (hpf) as previously described .
Purified Pam3CSK4 (InvivoGen, France) and flagellin from S. typhimurium (Flagellin FliC VacciGrade™, Invitrogen, France) were diluted in 1 mg/ml and 100 μg/ml in sterile water, respectively. For injection, 1 nl of the ligand solutions were injected into the blood stream at 28hpf. Sterile water was injected as a control experiment. Injections were performed using a FemtoJet microinjector (Eppendorf, the Netherlands) equipped with a capillary glass needle.
Bacterial burden imaging and quantification
Pools of 20 larvae were collected at 3- and 4-day post infection (dpi) and imaged by using the Leica MZ16FA Fluorescence Stereo Microscope (Leica Microsystems, Wetzlar Germany) equipped with the DFC420C color camera (Leica Microsystems). Bacterial loads were analyzed using dedicated pixel counting software as previously described .
Confocal laser scanning microscopy imaging and image quantification
Larvae (2 dpf) were embedded in 1% low melting point agarose (Sigma Aldrich), and image acquisition was performed by using a Leica TCS SP8 confocal microscope (Leica Microsystems) with a 10 times magnification objective (N.A. 0.40). Acquisition settings and area of imaging (in the caudal vein region) were kept the same across the groups for macrophages and neutrophils number counting (Fig. 1 j, k) and pixel counting (Additional file 1: Figure S1). Experiments were performed in two independent series. Double fluorescent lines tlr2+/+ Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) and tlr2−/− Tg (mpeg1:mCherry-F);TgBAC (mpx: EGFP) were used for macrophages and neutrophils number counting. Macrophage and neutrophil cell counting was either performed manually or by using the plugin Find Maxima in Fiji (http://imagej.nih.gov/ij/docs/menus/ process.html#find-maxima) (Fig. 1 j, k) using projections of the z-series. 25 individual larvae for each group were used for counting and representative images are given in Fig. 1h, i. With manual counting the z-series data was examined in cases when it was unclear whether the fluorescent pixels presented one or multiple cells. The result of manual cell counting (Additional file 11: Figure S11) is comparable with the result of automated cell counting using Fiji. Pixel counting of the double transgenic lines (Additional file 1: Figure S1 a, b) was performed using dedicated pixel counting software as well as previously described . For CLSM analysis of progression of infection of Fig. 4 we used Fiji software. To quantify the number of bacteria outside macrophages, total bacteria pixel count was quantified by Fiji software and followed by subtracting this number with the number of co-localised pixels. Co-localisation of macrophage and bacterial pixels was performed using a custom made script written in Java. For this experiment we used three objectives: 10x (N.A. 0.40), 20x (N.A. 0.75), 63x (oil immersion, N.A. 1.40) magnification.
RNA isolation, cDNA synthesis and qPCR
Total RNAs were extracted using TRIzol Reagent (Life Technologies) and purified using RNeasy MinElute Cleanup Kit (Qiagen, the Netherlands). The concentration and quality of RNAs were evaluated by NanoDrop 2000 (Thermo Scientific, the Netherlands). cDNAs were synthesized from 1 μg total RNAs and qPCR were performed by using the iScript™ cDNA Synthesis Kit (BioRad, the Netherlands) and iQ™ SYBR Green Supermix (BioRad) and normalized against the expression of ppial as a housekeeping gene . Results were analyzed using the ΔΔCt method . Primer sequences are described in Additional file 16: Table S5.
Deep sequencing and data analysis
Triplicates of 10 larvae of tlr2+/− and tlr2−/− with PBS (as control) or Mma20 injection, were homogenized in 300ul of TRIzol reagent, and total RNAs were purified as described above. RNAseq was performed using Illumina Hi-Seq 2500 as previously described . The raw data is available in the NCBI GEO database under accession number GSE102766. The RNAseq data were mapped on the zebrafish genome (version GRCz10) and tag counts were performed by Bowtie 2 using GeneTiles software (http://www.genetiles.com) . Then, we performed normalization and gene expression analysis using the R package and DESeq2 . After statistical tests, we performed further bioinformatics analyses Gene-Set Enrichment Analysis , Sub-Network Enrichment Analysis  and Pathway Enrichment Analysis . For creating gene networks based on common regulatory targets, we used Pathway Studio 9.0 (Elsevier, Amsterdam, the Netherlands) as previously described .
Comparison of edgeR and DEseq2: EdgeR and DEseq2 differ mainly in the aspects of normalization and estimation of the dispersion parameters. Normalization in edgeR is done via the trimmed mean of M values, while in DEseq2 this is done by comparing each library with a virtual library based on the relative log expressions. The dispersion parameters in edgeR are estimated by empirical Bayes and are therefore shrunken towards the overall mean of the estimates. Dispersion in DEseq2 is estimated by taking the maximum of the individual dispersions and the mean trend of the dispersions. As a consequence, edgeR tends to be more sensitive to outliers, while DEseq2 is less powerful .
Graphpad Prism software (Version 8.1.1; GraphPad Software, San Diego, CA, USA) was used for statistical analysis. All experiment data are shown as mean ± SEM. In the gene expression profiles of zebrafish treated with the TLR2 agonist (Fig. 1 b-g) and in immune gene expression in tlr2+/− and tlr2−/− fish lines infected with Mm (Fig. 5), statistical significance of differences was determined by two-way ANOVA with Tukey’s Multiple Comparison test as a post-hoc test. The other experiments were analyzed by using unpaired, two-tailed t-tests for comparisons between two groups and one-way ANOVA with Tukey’s multiple comparison methods as a post-hoc test for comparisons between more than two groups. For percent survival analysis (Fig. 3 f), statistical significance of difference was determined by a log-rank (Mantel-Cox) test. (ns, no significant difference; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).
We thank all members of the fish facility team for fish caretaking. We would like to thank our colleagues from Leiden University: Lanpeng Chen and Gerda E.M. Lamers for assistance with confocal laser scanning imaging and Joost J. Willemse for providing microscopic analysis programs. We also thank Vincenzo Torraca (London School of Hygiene & Tropical Medicine) for helpful discussions and the qPCR primers of cxcl11-like genes. The zebrafish tlr2 mutant was obtained from the Sanger Institute Zebrafish Mutation Resource (ZF-MODELS Integrated Project funded by the European Commission; contract number LSHG-CT-2003-503496), also sponsored by the Welcome Trust [grant number WT 077047/Z/05/Z].
SY and WH carried out all the experimental operations. SY, YS, and HPS performed RNAseq data analysis. SY, WH and HPS wrote the first draft of the manuscript. SY, WH and HPS participated in the design of the study and coordination. AHM and RMJ helped to revise the final manuscript. HPS conceived the study. MM assisted in statistical analysis of RNAseq data. All authors read and approved the final manuscript.
S. Y. and W. H were supported by grants from the China Scholarship Council (CSC). This funding body had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Ethics approval and consent to participate
No animals were used for experimentation. Larvae for experiments were obtained from zebrafish lines that were handled in compliance with the local animal welfare regulations and maintained according to standard protocols (zfin.org). The breeding of adult fish was approved by the local animal welfare committee (DEC) of the University of Leiden. All protocols adhered to the international guidelines specified by the EU Animal Protection Directive 2010/63/EU for which larvae under the age of 5 days post fertilization are not considered test animals.
Consent for publication
The authors declare that they have no competing interests.
- 9.Jung SB, Yang CS, Lee JS, Shin AR, Jung SS, Son JW, Harding CV, Kim HJ, Park JK, Paik TH, et al. The mycobacterial 38-kilodalton glycolipoprotein antigen activates the mitogen-activated protein kinase pathway and release of proinflammatory cytokines through Toll-like receptors 2 and 4 in human monocytes. Infect Immun. 2006;74(5):2686–96.CrossRefPubMedPubMedCentralGoogle Scholar
- 31.Gehring AJ, Rojas RE, Canaday DH, Lakey DL, Harding CV, Boom WH. The Mycobacterium tuberculosis 19-kilodalton lipoprotein inhibits gamma interferon-regulated HLA-DR and Fc gamma R1 on human macrophages through Toll-like receptor 2. Infect Immun. 2003;71(8):4487–97.CrossRefPubMedPubMedCentralGoogle Scholar
- 33.Agrawal S, Agrawal A, Doughty B, Gerwitz A, Blenis J, Van Dyke T, Pulendran B. Cutting edge: different Toll-like receptor agonists instruct dendritic cells to induce distinct Th responses via differential modulation of extracellular signal-regulated kinase-mitogen-activated protein kinase and c-Fos. J Immunol. 2003;171(10):4984–9.CrossRefPubMedPubMedCentralGoogle Scholar
- 52.Rougeot J, Torraca V, Zakrzewska A, Kanwal Z, Jansen HJ, Sommer F, Spaink HP, Meijer AH. RNAseq Profiling of Leukocyte Populations in Zebrafish Larvae Reveals a cxcl11 Chemokine Gene as a Marker of Macrophage Polarization During Mycobacterial Infection. Front Immunol. 2019;10:832.Google Scholar
- 54.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50.CrossRefPubMedPubMedCentralGoogle Scholar
- 56.Verway M, Bouttier M, Wang TT, Carrier M, Calderon M, An BS, Devemy E, Mcintosh F, Divangahi M, Behr MA, et al. Vitamin D induces interleukin-1beta expression: paracrine macrophage epithelial signaling controls M. tuberculosis infection. PLoS Pathog. 2013;9(6):e1003407.CrossRefPubMedPubMedCentralGoogle Scholar
- 61.Lachmandas E, Beigier-Bompadre M, Cheng SC, Kumar V, van Laarhoven A, Wang X, Ammerdorffer A, Boutens L, de Jong D, Kanneganti TD, et al. Rewiring cellular metabolism via the AKT/mTOR pathway contributes to host defence against Mycobacterium tuberculosis in human and murine cells. Eur J Immunol. 2016;46(11):2574–86.CrossRefPubMedPubMedCentralGoogle Scholar
- 77.Richardson ET, Shukla S, Sweet DR, Wearsch PA, Tsichlis PN, Boom WH, Harding CV. Toll-like receptor 2-dependent extracellular signal-regulated kinase signaling in Mycobacterium tuberculosis-infected macrophages drives anti-inflammatory responses and inhibits Th1 polarization of responding T cells. Infect Immun. 2015;83(6):2242–54.CrossRefPubMedPubMedCentralGoogle Scholar
- 78.Piermattei A, Migliara G, Di Sante G, Foti M, Hayrabedyan SB, Papagna A, Geloso MC, Corbi M, Valentini M, Sgambato A, et al. Toll-Like Receptor 2 Mediates In Vivo Pro- and Anti-inflammatory Effects of Mycobacterium tuberculosis and Modulates Autoimmune Encephalomyelitis. Front Immunol. 2016;7:191.CrossRefPubMedPubMedCentralGoogle Scholar
- 79.Davila S, Hibberd ML, Hari Dass R, Wong HE, Sahiratmadja E, Bonnard C, Alisjahbana B, Szeszko JS, Balabanova Y, Drobniewski F, et al. Genetic association and expression studies indicate a role of toll-like receptor 8 in pulmonary tuberculosis. PLoS Genet. 2008;4(10):e1000218.CrossRefPubMedPubMedCentralGoogle Scholar
- 91.Benard EL, van der Sar AM, Ellett F, Lieschke GJ, Spaink HP, Meijer AH. Infection of zebrafish embryos with intracellular bacterial pathogens. J Vis Exp. 2012;61:2.Google Scholar
- 93.Stoop EJM, Schipper T, Rosendahl Huber SK, Nezhinsky AE, Verbeek FJ, Gurcha SS, Besra GS, Vandenbroucke-Grauls CMJE, Bitter W, van der Sar AM. Zebrafish embryo screen for mycobacterial genes involved in the initiation of granuloma formation reveals a newly identified ESX-1 component. Dis Models Mech. 2011;4(4):526–36.CrossRefGoogle Scholar
- 96.Marin-Juez R, Jong-Raadsen S, Yang S, Spaink HP. Hyperinsulinemia induces insulin resistance and immune suppression via Ptpn6/Shp1 in zebrafish. J Endocrinol. 2014;222:229-41.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.