Cytokine reporter mice in immunological research: perspectives and lessons learned

Abstract

Cytokines are soluble messenger molecules with important regulatory functions throughout the immune system. ‘Cytokine reporter’ strains express marker molecules under control of elements from cytokine genes allowing for easy identification of their cellular sources. Such systems are well-accepted tools for research of cytokine function. The value of these strains lies in the ability to perform experiments relying on identification and isolation of live cytokine-expressing cells, provided that the reporter faithfully reflects the proper cytokine mRNA and protein production. As more diverse cell subsets are defined by their cytokine expression, the field has adapted with the generation of more sophisticated strains. In this review we summarize the evolution of cytokine detection methods and give examples of knowledge gained using cytokine reporter mice for cell types expressing interferon-γ and interleukin-4, -10 and -17. We also discuss current options for generating such reporter strains and their potential pitfalls.

Cytokines and T helper cells

Communication between cells of the immune system and other cells in the body are essential for successful immune responses, maintenance of immune homeostasis and implementation of tolerance. This ‘cross-talk’ is facilitated by signals generated through interacting surface molecules as well as production and binding of soluble messenger proteins, glycoproteins or peptides. These soluble immune mediators, nowadays broadly termed ‘cytokines’, are very similar to hormones but act in most instances more locally. They attract immune cells to sites of inflammation and orchestrate the type of immune response by supporting differentiation of specialized T-cell subsets, inducing B-cell isotype switching, and attracting and activating phagocytes.

The first cytokine to be described was macrophage migration inhibitory factor.¹ In these early days of cytokine research, bio-assays were the method of choice to detect the presence of a certain cytokine. Activities that could be easily observed such as migration, proliferation, or cytotoxicity were exploited, as in the frequently used murine cell line CTLL-2 whose growth depends on the presence of interleukin-2 (IL-2).² The development of ELISA allowed a molecularly standardized analysis and led later to the development of the enzyme-linked immunosorbent spot (ELISPOT),³ in which secreted cytokines are visualized around cultured cells. In flow cytometry, the use of intracellular cytokine staining allows the easy analysis of cell surface differentiation markers in combination with various cytokines. However, this method requires fixation and permeabilization of the cells for intracellular cytokine detection. A combination of ELISPOT and flow cytometry, called cytokine secretion assay, allows the cell-surface capture and detection of secreted cytokine through heterodimeric antibody conjugates that bind to the cell surface and also to the cytokine of interest to retain it on the surface of the secreting cell. Use of a second cytokine-specific antibody coupled to a fluorochrome enables the researcher to identify the live producers and so sort viable cells defined by secreted cytokines for further functional studies.^4,5 Importantly, this technology allows the detection and purification of viable cytokine-secreting cells at the single-cell level independent of any transgenic manipulation in mouse and man, hence reflecting the proper cytokine gene expression and protein production. However, the cells have to be activated in this assay to detect transient cytokine secretion.

A different way to detect cytokine production is the use of transgenic expression of marker molecules under the control of a cytokine gene promoter. An early approach employed promoter sequences of the hormone prolactin to drive luciferase expression.⁶ Only later, the first cytokine (IL-2) was analysed in a transgenic reporter system using β-galactosidase.⁷ Finally, the discovery of naturally occurring fluorescent proteins made transgenic fluorescent marking of cytokine-producing cells possible as was again shown first for IL-2.⁸ It is this method of placing a gene encoding a protein with an easily measurable property, be that fluorescence, enzymatic activity or antigenic properties, under the same transcriptional control as the cytokine of interest, that has acquired the term ‘cytokine reporter’.

Up until now most of our knowledge about in vivo cytokine function is the result of experiments using transgenic cytokine-over-expressing mouse lines and mice with targeted gene disruptions.⁹ In fact, some of the first targeted genomic mutations were in genes encoding the cytokines IL-4,¹⁰ IL-2¹¹ and IL-10.¹² As our understanding of the immune system deepens, it becomes increasingly obvious that universal gene deficiencies are not able to answer more intricate questions. The dilemma becomes evident in mice deficient for cytokines shaping development of a specific lineage. For example, in animals lacking IL-4, development of the T helper type 2 (Th2) cell lineage, although not completely absent, is strongly reduced.¹³ Hence, it is difficult to analyse the function of IL-4 in Th2 cells beyond the requirement for their differentiation. Likewise, deletion of the gene encoding IL-17A will not entirely answer what role Th17 cells play in vivo, because of the numerous other cellular sources of this cytokine.^14,15 Given the pluripotent effects of cytokines and potentially widespread receptor expression, deletion of an entire cytokine frequently shows a multitude of effects not restricted to immune regulation. This was overcome by conditional gene targeting, a method to delete genes in cell and tissue types.¹⁶ A complementary approach is the use of ‘cytokine reporter’ mice that allow us to study cytokine expression without modifying protein expression itself (Table 1). Using such mice, questions regarding development and differentiation, proliferation, survival, and migration by giving the localization and phenotype of cytokine-secreting cells in an unaltered environment can be answered. They can also be combined with other genetic modifications to enhance the potential of previously existing models.

Table 1.

Cytokine reporter mice. Previously published cytokine reporter strains used in immunological research. Cytokine genes targeted, reporter protein used and nomenclature associated with these strains are listed. References to the publications are provided in the table. Although we tried to include all published strains this list may not be comprehensive

Cytokine	Reporter gene	Type	Name	References
IL-2	LacZ	Transgene	IL-2_LacZ	7
IL-2	GFP	Transgene	IL-2-GFP-CD2.27	8
IL-4	GFP	Targeted	IL-4/GFP	65
IL-4	IRES GFP	Targeted	C.129-Il4tm1Lky/J	27
IL-4	YFP	Transgene	IL-4YFP	66
IL-4	Human CD2	Targeted	KN-2	30
IL-7	Human CD25	BAC	IL-7.hCD25	67
IL-7	Cre	BAC	IL-7.Cre	67
IL-7	GFP, Cre, DTR, Luc	BAC	IL-7GCDL	62
IL-10	IRES GFP	Targeted	B6.129S6-Il10tm1Flv/J	38
IL-10	Thy1.1	Transgene	10BiT	39
IL-10	EGFP	BAC	Tiger	38
IL-10	EYFP	Targeted	IL-10eYFP	68
IL-12β (p40)	IRES EYFP	Targeted	B6.129-Il12btm1Lky/J	69
IL-17F	Cre	BAC	IL-17F-Cre	48
IL-17F	Thy1.1	Transgene	IL-17FThy1.1	49
IL-17F	RFP	Targeted	IL-17F-RFP	47
IFNβ1	IRES EYFP	Targeted	B6.129-Ifnb1tm1Lky/J	70
IFN-γ	EYFP	Transgene	YETI	33
CCL17/Tarc	EGFP	Targeted	CCL17-EGFP	71
TGFβ1	GFP	Targeted	TGFβ null	72

BAC, bacterial artificial chromosome; EGFP, enhanced green fluorescent protein; EYFP, enhanced yellow fluorescent protein; GFP, green fluorescent protein; IFN, interferon; IL, interleukin; IRES, internal ribosome entry site; RFP, red fluorescent protein; TGF, transforming growth factor; YFP, yellow fluorescent protein.

The dogma of Th1 and Th2 cells developed by Mosmann et al.¹⁷ introduced the concept of specialized T helper cell subsets of which not only the development but also the function depends critically on defined cytokine signalling. Although this initial concept is now perhaps too simplistic, it has been greatly expanded in recent years. Numerous T helper and regulatory cell subsets have now been identified and studied in depth and although the definition of these subsets as separate terminally differentiated lineages is somewhat controversial,^18–22 T helper cell subsets are still defined by their cytokine signatures: interferon-γ (IFN-γ) and IL-4 for Th1 and Th2 cells, respectively, and IL-17A for Th17 cells.^23,24 However, follicular T helper cells are defined by the surface markers CXCR5 and inducible co-stimulator rather than a subset-specific cytokine profile.^25,26

Major findings using cytokine reporter mice

IL-4 and IFN-γ, early markers of T helper cell function

The hallmark cytokine for Th2 cells is IL-4, the discovery of which is closely linked with its role in immunity against extracellular pathogens. Early efforts to detect IL-4-expressing cells in vivo were hampered by its quick release from activated T helper cells, a problem common to the majority of cytokines. Usually this is ‘overcome’ by ex vivo re-stimulation in the presence of the secretion inhibitor Brefeldin A to accumulate cytokine intracellularly. To measure IL-4 expression in situ, Mohrs et al.²⁷ generated a bicistronic IL-4 reporter mouse, the ‘4get’ mouse, by inserting an internal ribosome entry site–green fluorescent protein (IRES-GFP) cassette into the 3′ untranslated region of the Il4 locus. By comparing it with mRNA analyses this 4get reporter was shown to faithfully report IL-4 expression under Th2 polarizing conditions.^28–32 It also marked other constitutive IL-4 producers such as natural killer (NK) T cells, basophils, eosinophils and mast cells.^28,32,33

Using a dual reporter system in which two IL-4 cytokine reporter mice, namely the 4get and the hCD2-expressing KN2 reporter mice, Mohrs et al.³⁰ described widespread cytokine expression during Heligmosigmoidespolygyrus infection. They could distinguish between IL-4-competent cells, which are GFP⁺, and IL-4-producing cells, which are GFP⁺ hCD2⁺. In combination with both IL-4 receptor -deficient and signal transducer and activator of transcription 6-deficient backgrounds, these dual reporter mice were used recently to show that the majority of naive T cells are responsive to IL-4.³⁴ During H. polygyrus infection, sustained IL-4 conditioning of the Th2-reactive lymph node altered the response of naive T cells to antigenic encounter. This is important as it shows that naive T cells can potentially be conditioned by a cytokine before T-cell receptor engagement and that polarization can in fact begin before T-cell receptor engagement. Hence, these reporter strains have been instrumental in re-evaluating the existing dogma that cytokine secretors have to be in close proximity to their target cells.^35–37

In a similar fashion to the 4get mice, Locksley et al. generated an IFN-γ reporter mouse, in which enhanced yellow fluorescent protein (EYFP) fluorescence reported IFN-γ gene expression. This ‘YETI’ strain was used in conjunction with the 4get strain as an IFN-γ/IL-4 double-reporter. They demonstrated that naive NK-T cells and NK cells, but not naive T cells, are continually expressing these two genes (Fig. 1).³³ Naive T cells were shown to up-regulate expression of these cytokines only after activation under respective polarizing conditions. These findings led to the important conclusion that NK-T cells and NK cells initiate cytokine expression in response to developmental cues in the thymus and bone marrow, respectively, and not only in response to stimulation by a pathogen. This finding has important implications with respect to induction of a subsequent Th1 and Th2 response by these innate cell types, given that this rapid cytokine secretion from NK or NK-T cells favours a polarization to Th1 or Th2 differentiation.

Figure 1.

Conceptual insights from cytokine reporter mice. Summary of findings obtained using reporter strains for interferon-γ (IFN-γ; YETI), interleukin-4 (IL-4; 4get), IL-10 (Tiger and 10BiT) and IL-17F (IL-17F-RFP and IL-17F-Cre^EYFP). (a) Both natural killer (NK) and NK-T cells can express IL-4 and IFN-γ before activation. (b) Activation of T cells induces IL-10 expression by T cells from mesenteric lymph nodes (MLN) and intra-epithelial lymphocytes (IEL) in Tiger mice. 10BiT mice were used to show that IL-10 expression can be induced from both Foxp3⁻ and Foxp3⁺ T cells in Peyer’s patches (PP), small intestine intraepithelial lymphocytes (SI IEL) and lamina propria lymphocytes of both colon and small intestine (cLPL and SI LPL, respectively). This was dependent on transforming growth factor-β (TGF-β) signalling. (c) IL-17F-RFP mice were used to show that natural Foxp3⁺ (nTreg) cells can up-regulate IL-17F in response to IL-6, IL-1 and IL-23 signalling. IL-17F-Cre^EYFP mice were used to show that fully differentiated Th17 cells are resistant to Foxp3 expression in response to TGF-β signalling.

IL-10, the generic immune suppressor

Numerous cytokine reporters were also generated for the immune-suppressive cytokine IL-10. Flavell et al. published an IL-10 reporter strain (tiger), in which IRES-GFP was introduced into the 3′ untranslated region of the IL-10 gene.³⁸ In these tiger mice, IL-10 expression was readily induced in intraepithelial lymphocytes and mesenteric lymph node CD4⁺ T cells after multiple rounds of anti-CD3 injection (Fig. 1). Although the GFP signal of this transcriptional reporter was clearly associated with IL-10 protein, the authors also show a small GFP⁺ IL-10⁻ population, possibly because such a reporter does not take into account the regulation of mRNA stability and translation as well as protein accumulation and degradation²⁷ (Fig. 2). Also, GFP may be inherently more stable than IL-10, remaining in cells after IL-10 has been secreted or degraded (Fig. 2). While the authors, to their credit, point out these complications, this observation highlights the caution required when interpreting data from cytokine reporters.³⁸ In fact, transcriptional reporters such as the tiger mouse may be of great value for the identification of post-transcriptional regulatory mechanisms.

Figure 2.

Challenges associated with the use of cytokine reporter mice. The potential pitfalls in the use of cytokine reporter mice are grouped into those associated with the actual process of transgenesis (in green) and those as a result of the reporter itself in relation to its biological significance (in red). Green: During the process of transgenesis short promoter and other control sequences are used to drive expression. This fact as well as integration site-dependent variegation may lead to an expression pattern different from that of the actual cytokine. In the knock-in approach, however, genetic deficiency can be the end result depending on the strategy used. Either in the homozygous state or as haploinsufficiency this can lead to unexpected cytokine-related phenotypes. In bacterial artificial chromosome (BAC) -transgenesis additional genes that frequently lie on the same BAC are co-integrated and may result in phenotypes through dosage effects. Red: mRNA levels may be regulated through stabilizing or destabilizing elements in untranslated regions. If these are used to integrate an internal ribosome entry site or 2A-connected marker gene, stability of the mRNA may be changed, which results in divergence from the actual cytokine mRNA levels. When the reporter mRNA or protein is less/more stable than that of the cytokine, the reported outcome is different from the levels of the actual investigated cytokine. Additional levels of consideration are the translational and stability control of the investigated cytokine. Also, some cytokines have to be modified and activated before being biologically active whereas others may be stored for some time in vesicles and are released when necessary. These processes, important for the understanding of cytokine biology, are all not reflected in reporter mouse systems.

Another IL-10 reporter strain (10BiT), generated by Weaver et al., expresses the surface marker CD90-1 transgenically under control of the IL-10 promoter³⁹ and was used to describe the diversity, distribution and developmental origins of regulatory T (Treg) cell subsets. In the steady-state, IL-10 expression was limited to T cells, and in particular to gut-dwelling Treg cells.³⁹ In addition, it was shown that dependent on transforming growth factor-β (TGF-β) signalling both Foxp3⁺ and Foxp3⁻ T cells are able to produce IL-10 (Fig. 1). A surprising finding, however, was that within Treg cells defined by the transcription factor Foxp3, the expression status of IL-10 was variable depending on localization. Colonic lamina propria-dwelling Treg cells expressed high levels of CD90-1 (IL-10), whereas mesenteric lymph node-derived Treg cells expressed low levels of CD90-1. One may therefore be tempted to predict that such Foxp3⁺ Treg-derived IL-10 is responsible for protection against the severe colitis observed in IL-10-deficient mice.⁴⁰ These findings demonstrate clearly that a combination of cytokine and transcription factor reporters can be used to further define ‘subsets within subsets’.

IL-17A and IL-17F, Beelzebub and Lucifer

A new research field that developed around IL-17A and IL-17F-expressing CD4⁺ T cells (Th17) has been another arena in which cytokine reporter mice have begun to make a difference. The Th17 cells have attracted an overwhelming interest because of their obvious association with autoimmune diseases such as rheumatoid arthritis, multiple sclerosis and psoriasis.^14,41,42 Research performed using the multiple sclerosis model, experimental autoimmune encephalomyelitis, has led to the unravelling of the transcription factors and cytokine requirements in development and maintenance of these Th17 cells.^43–46 The arrival of Th17 reporter was greatly anticipated, and results derived from experiments using these mice have revealed many facets of Th17 regulation and phenotypic plasticity. Interestingly, even though the hallmark cytokine of Th17 cells, IL-17A, was used to make all of the defining research uncovering the existence of Th17 cells, the first three reporter mice to be published all placed either red fluorescent protein (RFP), Cre recombinase or CD90-1 under the control of the IL-17F promoter.^47–49 Furthermore, expression of IL-17A and IL-17F, although largely overlapping, does indeed vary between reports. However, IL-17A reporter mice will no doubt soon become available given the interest surrounding this cytokine and implications in autoimmune disorders and host defence against extracellular microorganisms.

Using an IL-17F-RFP reporter mouse crossed to a Foxp3-GFP strain, it was demonstrated that cells co-expressing CD4, GFP and RFP exist in cultures designed to polarize Th17 cells. This suggests that both Th17 and Treg cell developmental programmes are induced simultaneously, and coexist for a period of time before complete differentiation of Th17 cells.⁴⁷ In addition to the fact that IL-6 inhibits the differentiation of induced Treg cells, it was further shown that naturally occurring GFP⁺ Treg cells are able to ‘redifferentiate’ into RFP⁺ GFP⁻ Th17 cells in the presence of IL-6. We then showed that fully differentiated Th17 cells are resistant to TGF-β-mediated up-regulation of Foxp3, implying that terminal differentiation of Th17 cells inhibits subsequent reversion to a Treg phenotype (Fig. 1).⁴⁸ Also, we and others have found that upon adoptive transfer of live Th17 cells the Th17 phenotype is unstable and down-regulation of IL-17A and IL-17F occurs in the absence of pro-inflammatory cytokines.^48,50,51

A common feature of ‘Th17-mediated’ inflammatory states is the presence of IL-17A⁺ IFN-γ⁺ T cells in the inflamed organ. Using intracellular staining alone it was not clear if these double-positive cells are derived from differentiated Th17 cells or Th1 cells. Lee et al.⁴⁹ showed in a CD90-1-expressing IL-17F reporter strain that purified Th17 cells exposed to different cytokine milieus respond to IL-12 with up-regulation of IFN-γ. Therefore, it seems that IL-17⁺ IFN-γ⁺ cells are, at least in part, reprogrammed Th17 cells, but this point remains to be definitively proven using a fate mapping approach in vivo. Furthermore, TGF-β signalling was also shown to be necessary for the maintenance of IL-17A and IL-17F expression, which further supports the significant overlap in signalling requirements for both Th17 and Treg cells.⁴⁹

Cre under cytokine promoter control

A special type of reporter strain combines expression of the Cre recombinase with a reporter transgene in which a marker is transcriptionally activated by Cre-mediated recombination. To the best of our knowledge the first model using Cre recombinase expressed under the control of a cytokine promoter was published last year by Waisman and colleagues.⁴⁸ This strain carries a bacterial artificial chromosome (BAC) transgene in which the open reading frame of Cre recombinase replaced the first two exons of the IL-17F gene. The IL-17F-Cre allele was crossed to a conditional EYFP reporter strain,⁵² yielding the IL-17F-Cre^EYFP system. After adoptive transfer of live Th17 cells into wild-type recipients, we were able to use EYFP expression to track these cells in vivo. We used this strain to isolate and transfer live Th17 cells and showed that Th17 cells down-regulate expression of hallmark cytokines after a period of homeostatic expansion in RAG1-deficient hosts. Also, we found that pure Th17 cells are resistant to up-regulation of Foxp3 in response to TGF-β signalling (Fig. 1 and ref. 48). After crossing the IL-17F-Cre mice to another, red-fluorescent, Cre reporter allele,⁵³ we were able to identify permanently marked living Th17 cells. We found in the bone marrow of these mice RFP⁺ Ly6C⁺‘memory Th17’ cells that no longer expressed IL-17A or IL-17F⁵⁴ (A.L. Croxford, unpublished observations), a population that has not been described before. The questions that can be answered by a Cre-mediated ‘fate mapping’ approach are therefore separate from those of standard expression reporters.

Approaches and considerations for the generation of cytokine reporter strains

Several different approaches for the generation of cytokine reporter strains are currently in use: conventional transgenic mouse strains, targeted insertions (‘knock-ins’), and BAC transgenes (Fig. 3). Conventional transgenes are the result of random genomic integration of the transgene (tg) after pronuclear injection (Fig. 3a) and their generation requires detailed knowledge about transcriptional control elements of the cytokine gene of interest. To circumvent this and rely solely on all endogenous elements, researchers have also used gene targeting to introduce marker genes into a locus of interest,⁵⁵ resulting in so-called knock-ins (Fig. 3b). When integrating the marker 5′ or 3′ of the open reading frame and use of either IRES⁵⁶ or viral 2A peptides⁵⁷ the generation of null alleles in this method can be avoided (Fig. 3d–f). An alternative that has been given more attention in recent years combines both approaches. The BAC-mediated marker expression alleviates the need for promoter characterization while maintaining the speed of conventional transgenesis⁵⁸ (Fig. 3c). It has the advantage of obtaining more transgenic founder lines resembling natural expression, yet care has to be taken when additional copies of genes also present in large BAC are co-introduced, leading to a phenotype of their own. Also the number of available marker genes has expanded beyond LacZ and GFP, enabling the researcher to choose the appropriate colour, suitable for analysis by histology, flow cytometry or in vivo imaging techniques.⁵⁹ Commonly used fluorescent proteins now include EYFP, RFP, mCherry and tdTomato. For most colours, variant proteins are now available which are more stable and fold more quickly, like Venus instead of YFP. However, this increased stability also has an inherent drawback in that a more stable protein will yield a signal even in cells that have ceased transcription of the respective locus (Fig. 2). To overcome this problem, a PEST degradation cassette rich in proline, glutamic acid, serine and threonine, can be attached to the fluorescent protein⁶⁰ reducing the half-life dramatically. We have used this approach for the generation of a BAC targeting cassette using either tdTomato alone or in tandem with p2A-cre (S. Haak, personal communication, ref. 58). Nevertheless, this method cannot overcome the problem of distinct translation and folding kinetics of non-secreted reporter proteins versus secreted cytokine proteins. In spite of these numerous advances, LacZ remains a frequent marker gene because it is, in combination with neomycin resistance (β-Geo), now part of many gene-trap cassettes and targeted/trapped embryonic stem cells can be ordered from various sources. Also ease of staining and high sensitivity make it a reliable marker gene.

Figure 3.

Generation of cytokine reporter mice. Three basic techniques can be used to generate cytokine reporters. In the knock-in approach (a) a marker gene is introduced into the genomic locus of the cytokine of interest by homologous recombination in embryonic stem cells. Co-introduced selection markers such as neomycin resistance are removed by use of DNA-recombinases such as Flp or Cre recognizing FRT or loxP sites, respectively (filled ovals). In conventional transgenesis (b) the reporter gene is placed on a construct containing defined promoter and other transcriptional control elements. This construct is injected into pronuclei to generate transgenes through random insertion into the genome. Usually the transgene is integrated in concatameric assemblies, and mice showing the expected expression pattern have to be identified from the different founder lines. In bacterial artificial chromosome (BAC) transgenesis (c) a larger transgenic vector is generated by modification of a BAC containing the whole genomic locus (and more) of the cytokine of interest. The modified BAC is again injected into pronuclei to generate transgenic mice by random insertion. Yet, because of the size of the construct transcriptional control elements do not have to be defined in advance and such transgenes are ‘well-insulated’ against variegation effects. For the knock-in approach three different variants can be performed. Either the cytokine gene is disrupted by the introduction of the marker (d) or the marker is introduced into the 3′ (e) or 5′ (f) untranslated regions (in violet). In the latter cases, translation of both open reading frames is facilitated through use of either internal ribosome entry site (IRES) or 2A elements.

Expression of a surface protein foreign to the host, such as CD90-1 or human CD4, represents another option allowing identification and sorting by flow cytometry. Because antibodies labelled with different fluorochromes can be used to identify marker-expressing cells, this approach is more versatile in multicolour staining protocols. Also, it is possible to deplete mice of cytokine-expressing cells after administration of an antibody directed against the foreign surface protein.

A significant challenge for the future will be the detection of cytokines by in vivo imaging using ultrasensitive cameras. It can be expected that despite formidable obstacles such an approach will yield numerous new insights into cytokine biology. Luciferase-based systems can be used but they require the injection of large quantities of expensive substrates. Also, cells from such mouse lines cannot be used for flow cytometric analyses. Because tissue is most permeable at long wavelengths, transgenesis with red fluorescent proteins offers an alternative.⁶¹ Another potential strategy for the generation of cytokine reporter may lie in the use of multi-cistronic transgenes, combining a fluorescent protein, Luciferase, Cre and possibly diphtheria toxin receptor [Shalapour (in press; DOI: 10.1002/eji.201040441)]. A further challenge in cytokine biology lies in monitoring not only expression, but also protein stability, storage and release of a cytokine. Attempts on achieving this have been made by fusion of IL-15 to EGFP (Fig. 2).^63,64

Concluding remarks

Taken together, numerous questions regarding the regulation of cytokine expression and the plasticity of T-cell subsets have been clarified using cytokine reporter mice. They have eased significantly the analysis of living cytokine-expressing cells by direct imaging. Caution has to be taken, however, because the reporter data relates only to transcriptional activity and stability of the reporting protein may skew the results.

Disclosures

Andrew Croxford and Thorsten Buch declared no conflicts of interest, including all relevant financial interests in any company or institution that might benefit from the publication.