Of men, mice and pigs: looking at their plasmacytoid dentritic cells

Protection from pathogen-borne disease is provided through complex interactions of the innate and the adaptive immune systems. To initiate this, pathogens must be recognized as a danger. Dendritic cells (DCs) possess key molecules for the recognition of danger signals as well as for the initiation of immune responses – an ideal configuration for bridging innate and adaptive immune responses.¹ In their immature state, DCs act as environmental sentinels. They continuously sample their environment for danger. For the recognition of pathogen-associated molecular patterns (PAMPs), DCs are equipped with a multiplicity of pattern-recognition receptors (PRRs). Among those are the Toll-like receptors (TLRs), an evolutionary highly conserved family of receptors, responsible for the detection of multiple PAMPs.² To date, 11 different TLRs have been identified in mammals. The large number of PAMPs that trigger signalling by TLRs include peptidoglycans (TLR2), double-stranded RNA (TLR3), lipopolysaccharide (TLR4), flagellin (TLR5), single-stranded RNA (TLR7 and TLR8) and bacterial DNA containing non-methylated CpG motifs (TLR9).³

The interaction of PAMPs and TLRs induces multiple changes in DCs. These include the reduction of the uptake of antigens through phagocytosis or endocytosis, the change of migratory patterns, and the up-regulation of major histocompatibility complex (MHC) and costimulatory molecules. As a result, professional antigen-presenting cells (APCs) evolve with optimal T-cell stimulatory properties.¹ Additionally, upon activation, DCs produce soluble immune molecules, such as interleukin-12 or interferon-α (IFN-α), which are able to modulate the quantity and quality of immune responses. Even though many cell types are able to produce type one interferon (IFN-I), one cell type stands out by virtue of its enormous capacity to produce IFN-α. For many decades, this rare cell population in human peripheral blood was known by virologists as the ‘natural interferon-producing cell’ (NIPC). Independently, a rare population of cells with plasmacytoid morphology and an unique phenotype of T-cell and myeloid-related antigens was noted and named as ‘plasmacytoid T cells’ or ‘plasmacytoid monocytes’. Based on the more recent finding that, upon stimulation, these cells have the capacity to convert into DCs, they were renamed as plasmacytoid DCs (pDCs).¹ In fact, it became clear that pDCs and NIPCs are indeed identical. After the characterization of human pDCs, murine pDCs were also identified. Murine and human pDCs show many phenotypic and functional similarities.⁴ Importantly, pDCs of both species express predominantly TLR7 and TLR9, and both are able to produce high levels of IFN-α to the single-stranded RNA or CpG-DNA, respectively, of the natural ligand.⁵ Moreover, pathogens such as herpes simplex virus or influenza virus induce high levels of IFN-I in both human and mouse pDCs.⁴ The discovery of the murine pDC has lent insight into various aspects of pDC biology, many features of which will probably translate to human pDCs. However, one major and important difference between mice and humans concerns the expression pattern of TLR9. In humans, only pDC and B cells express TLR9 and thus respond to TLR9 ligands. All other effects of TLR9 ligands on human immune cells seem to be indirect and depend on factors produced by pDCs and B cells. The situation in the mouse is different because not only pDCs and B cells, but also most other DC subsets, as well as macrophages, express TLR9 and thus respond directly to TLR9 ligation. Given this important species-specific difference in TLR9 expression, the mouse seems not to be an ideal animal model for using to establish TLR9-based therapeutic strategies that will probably translate into successful clinical treatments. Bacterial or viral DNA, the natural ligands for TLR9, can be mimicked by oligodeoxynucleotides (ODNs) with special CpG motifs (CpG-ODNs).⁶ The majority of in vivo studies have exploited the mouse as an animal model and have shown that CpG-ODNs are effective both as adjuvants and for therapeutic intervention in infectious and tumour model systems. However, the differences between CpG-ODN-responsive cells in mice and humans are a caveat in translating the murine data into the human system. Besides rodent models, a few studies have analysed the immune response of CpG-ODNs in other animals.⁶ Non-human primates have been used and shown to respond to CpG-ODNs, but for financial, social and ethical reasons, monkeys will remain a suboptimal animal model. Some studies have proven that farm animals, such as sheep, cattle or pigs, respond to CpG-ODNs, and, for pigs, pDCs have now been identified and characterized.^7,8 In the current issue of Immunology, Guzylack-Piriou⁹ and colleagues demonstrate that pig pDCs are the main producers of IFN-α in response to certain CpG-ODNs. Importantly, the authors additionally show that myeloid DCs and monocytes/macrophages are refractory to CpG-ODNs. Thus, the CpG-ODN responsiveness in pig seems to mimic the situation in primates, including humans, and therefore recommends, besides non-human primates, the pig as an animal model for preclinical studies with CpG-ODN.

However, some differences remain in the CpG-ODN responsiveness between pig and human immune cells. CpG-ODNs can be distinguished into at least two major groups which differ in chemical composition, species specificity, cytokine responses and in vivo immune responses. B-type CpG-ODNs are poor inducers of IFN-α in pDCs, but induce large amounts of other cytokines and B-cell proliferation.⁶ The recognition of B-type CpG-ODNs by TLR9 is species specific.¹⁰ The B-type CpG-1668, for example, predominantly stimulates murine cells, whereas CpG-2006 stimulates predominantly human immune cells. CpG-2006 was also used by Guzylack-Piriou et al., but did not show any stimulatory capacity for pig immune cells, suggesting that tests with B-type CpG-ODNs in pigs would have to be carried out with pig-specific CpG-ODN, some of which have been defined previously.^11,12 In contrast, A-type CpG-ODNs induce high production of IFN-α in pDCs, but are less potent in the induction of other cytokines and do not promote B-cell proliferation. A-type CpG-ODNs have poly(G) sequences at the 5′- and 3′ ends, which are partially phosphorothioate modified and contain a CpG-bearing palindromic sequence with a diester backbone. A-type specific CpG-ODNs are less species specific, and CpG-2216, originally defined for human pDCs, was shown to function in murine pDCs and (in the study by Guzylack-Piriou et al.) also in pig pDCs.^4,13 Another type of CpG-ODN, named C-type CpG-ODN, has been recently characterized and found to display mixed structural, as well as biological, properties.¹⁴ For example, it induced high levels of IFN-α in human pDCs but promoted, additionally, the proliferation of B cells. Little is yet know about the species specificity of these C-type ODNs sequences.

Taken together, the study of Guzylack-Piriou provides good news for those considering the pig as an alternative preclinical animal model system to evaluate the immunobiology of CpG-ODNs.