SEARCH

SEARCH BY CITATION

Keywords:

  • APC;
  • DC;
  • Immune regulation;
  • Macrophages;
  • Mucosal immunity

Abstract

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

A specialized network of immune cells is positioned directly in the mucosa of the intestinal tract. In this Viewpoint, we discuss the nature and function of DC and macrophages occupying the intestinal lamina propria of mice. DC and macrophages share phenotypic traits and functional plasticity, properties that preclude simple classification of the two cell types. Nevertheless, the information available appears to have reached “critical mass” to allow for a clear demarcation between intestinal macrophages and DC.


The particle zoo

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

DC and macrophages control key aspects of innate and adaptive immune responses. In particular, in extra-lymphatic tissues, such as the intestinal lamina propria (LP) discussed in this Viewpoint, there are more and more cell types discovered or re-discovered that share properties of DC and macrophages 1–5. There is an accumulating wealth of exciting information on what these particular cells can do and how they contribute to the function of the intestinal immune system. These new insights, however, are confounded by a growing confusion regarding the identity of the DC/macrophages. Figure 1 and Table 1 summarize some of the more recent reports investigating the function of DC/macrophage subsets in the intestinal LP and highlights the proposed functions of these cells. It is obvious that a direct comparison of subset compositions reported in individual studies is not always straightforward. Even if this may be due in part to divergent phenotypic marker sets used to identify LP cells, it is conceivable that a lack of comparability is an inevitable result of pseudo-cell types.

thumbnail image

Figure 1. Immune relevant cells residing in the intestinal LP. This sketch shows a set of cells identified in the reports listed in the table. This limited selection of cells and the markers used to identify them raises an identity problem: which is a “look-alike” and which is a genuine cell type? Different shapes account for the classification as DC or macrophages used in various publications. In this Viewpoint, cell types depicted in blue and orange would be named as DC and macrophage, respectively.

Download figure to PowerPoint

Table 1. Phenotypes and ascribed functions of intestinal LP DC and macrophages
Cell phenotypeCell functionReference
CD11cCD11b+F4/80+MHCII+Constitutively produce interleukin-10, promote Foxp3+ Treg differentiation11
CD11chighCD11b+CX3CR1+Induce interleukin-17 in T cells11
CD11chighCD11bhighTLR5+Induce IgA class switch; promote interleukin-17 production and Th1 differentiation15
CD11c+CD11b+CCR3+SiglecF+Describes these cells as long-lived intestinal eosinophils9
CD11c+CD11b+CD103+CX3CR1Migrate in lymph; potent antigen presenting cells10
CD11c+CD11b+CD103CX3CR1+Not present in lymph; low antigen presenting capacity10
CD11c+CD103+CX3CR1Derive from pre-DC, require FLT3L, carry Salmonella into mLN12, 14
CD11c+CD103CX3CR1+Derive from MDP, require M-CSF12, 14
CD11c+CD11b+MHCII+CX3CR1+CCR6Extend transepithelial cell processes in vivo, sample intestinal bacteria8
CD11c+CD11b+MHCII+CD103+Induce gut homing T cells, reduced in mLN of CCR7 deficient mice32
CD11c+CD11b+MHCII+CD103+Induce Foxp3+ regulatory T cells in an retinoic acid-dependent manner35

Apart from potential pseudo-cell types, the classical markers currently used may be too ambiguous to allow for coherent definitions of DC and macrophages. CD11c, for example, has a long tradition as a DC marker and many studies used CD11c expression levels to distinguish different subtypes of DC/macrophages. Indeed, CD11c discriminates DC and macrophages reasonably well, with DC expressing higher CD11c levels. Yet, CD11c expression levels alone by far cannot be used to unambiguously identify DC, as illustrated, for example, by CD11c+ intestinal eosinophils and gut resident CX3CR1+ DC/macrophages (Fig. 1). Another commonly used marker is CD11b; however, CD11b is expressed by various DC/macrophages in the LP and does not seem to discern functionally distinct cell populations.

Is it worthwhile to distinguish DC from macrophage-like cells in the LP or is this rather “much ado about nothing”? In this Viewpoint, we outline a revised concept to classify DC and macrophages residing in the LP based on some key facts. This may help to further clear up the current puzzling heterogeneity of DC/macrophage cell types. Moreover, the fundamental subdivision into DC and macrophage-like cells may allow for better understanding of issues, such as the maintenance of intestinal homeostasis, which is critically dependent on a division of labor between migratory DC and local macrophages.

Quantity matters

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

A frequently neglected pitfall in the study of LP DC/macrophages relates to the procedure applied to isolate LP cells for flow cytometric characterizations or in vitro assays. Counting the number of villi of the murine small intestine, we observe about 5000 villi per square centimetre. Although this number represents only a rough estimate and also varies depending on the intestinal segment examined, it allows for the conclusion that the murine small intestine harbors a minimum of 1×105 villi. DC and macrophages are readily visible in conventional tissue sections stained for markers, such as CD11c, CD11b or F4/80. More recently, multi-photon confocal microscopy using transgenic mice expressing fluorescent proteins under control of the CD11c, MHC class II (MHC II) or CX3CR1 promoter 6–8 allowed visualization of the dense network of DC/macrophages in the LP. From such studies, we know that DC/macrophages are present in significant numbers in the small intestinal LP: at least ten DC/macrophages per villus. This represents a conservative estimate and it is likely that the “true” number is much higher. We postulate that a minimum of 1 million DC/macrophages is present in the small intestine. To our knowledge, currently, there are no methods available that allow isolating as many DC/macrophages from the LP. Therefore, it should be considered that depending on the preparation technique applied, a protocol-specific cell-cocktail is obtained, which may not reflect the composition originally present in the intestine. Such biases may lead to overestimating – or underestimating – the importance of certain cell types depending on how easily they are isolated.

Intestinal eosinophils, which in the intestine express CD11c 9, might serve as an instructive example. Microscopy detects only few eosinophils per villus and the number of eosinophils that can be isolated is similar to that predicted by microscopic analysis. Microscopic inspection also demonstrates that an interesting population of LP cells expressing the chemokine receptor CX3CR1+ is far more frequent in the intestine compared with eosinophils; yet, the frequency of these CX3CR1+ cells as judged by flow cytometric analysis of isolated cell populations is lower than that of eosinophils (our unpublished observation). These observations suggest that (i) given the particular difficulty in isolating LP cells, a detailed description of the isolation method and gating strategy is mandatory. The development of optimized isolation methods might substantially facilitate further studies; (ii) the frequency of isolated cells does not necessarily reflect their numbers in the tissue. Knowing the total number of cells is critical for judging the relevance of experimental observations.

Three's a crowd

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

On the premise that numbers matter, two particularly abundant cell types with properties of macrophages and/or DC can be identified in the small intestinal LP under steady-state conditions. A first subset is characterized by expression of CD11c, MHCII and a lack of CX3CR1. The majority but not all of these cells co-express CD11b and CD103, indicating the existence of further phenotypically distinct subpopulations (Fig. 1) 10–13. A second major population expresses CX3CR1 and CD11b but lacks expression of CD103 and has a comparably lower expression of CD11c and MHCII 8, 10–12, 14. Thus, expression of CX3CR1 might be considered to be exceptionally well suited to allocate subpopulations of LP mononuclear phagocytes (MP) (Fig. 1). Such condensed appraisal of only two major MP subsets in the LP does not deny that further phenotypically and functionally distinct populations exist. But first, it accounts for the apparently high numbers of these two subsets as observed in the tissue context, i.e. in the absence of any biased isolation technique, and second, it correlates well with shared functional and developmental properties as discussed in the following section.

There is a general consensus that the CX3CR1 MP subset represents bona fide DC. In contrast, the nature of CX3CR1+ MP is more controversial. Numerous recent reports classified both subsets as DC (see, for example, 8, 15), whereas others emphasized that CX3CR1+ MP also display properties typical of macrophages. Therefore, CX3CR1+ MP have also been termed macrophages 11, myeloid DC-like cells 16, or even simply myeloid or LP cells 17, 18. The classification of CX3CR1+ MP as DC seems to fit well to the observation that CX3CR1+ MP can extend cell processes through the epithelial layer into the gut lumen 6, 19, 20. These cell protrusions morphologically resemble dendrites and support the rationale of calling these cells DC. Yet, another morphological feature, the presence of extensive vacuoles in CX3CR1+ MP 12 would rather suggest that these cells are macrophages.

Such discussion may seem semantic. Yet, names go along with associations and assumptions. Although changing a name does not change a cell's function, a correct naming provides the basis for communication and it might be worthy of careful consideration. In the following, we discuss recent observations on the nature and function of CX3CR1+ and CX3CR1 MP in the small intestine under steady-state conditions. We address CX3CR1+ MP as macrophages and CX3CR1 MP as DC. Such terminology is in disagreement with some (but by far not all) relevant original publications. Still, it may serve at least as a tool to more adequately discuss the function of LP DC and macrophages and to reconcile contradictory observations. With these designations in mind, the reader may ask: what about classical macrophages? We will comment on this at the end of this Viewpoint.

To migrate or not to migrate?

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

Macrophages are tissue resident cells with high phagocytic activity, scavenging dead and apoptotic cells as well as bacteria. Macrophages can be recruited as precursors from the blood to assist in clearance of an immunological insult but macrophages would not travel beyond their original tissue destination 1, 2. Early studies using F4/80 as a marker identified the largest population of F4/80+ macrophages in the intestine when compared with other organs 21, 22. F4/80+ macrophages are in close contact with the epithelium, express CD11b and intermediate levels of MHCII 10, 22. A more recent study used the differential expression of CD11b and CD11c to distinguish CD11b+CD11hi DC and CD11b+CD11cint macrophages 11. Taking into account the latest information available, it seems likely that the F4/80+ macrophages described more than 25 years ago as well as the CD11b+CD11cint macrophages 11 are identical or at least largely overlap with the population of CX3CR1+ cells addressed as DC in other studies. From this perspective, the re-discovery of intestinal macrophages to some extent is a result of a switch in terminology.

In contrast to macrophages, DC are characterized by a dynamic life cycle. Residing as sessile cells in the tissue, DC display an immature phenotype. DC are distinguished by a low expression of MHCII and costimulatory molecules and display a high endocytic activity. Upon stimulation, DC mature and upregulate MHCII and costimulatory molecule expression. Concomitant with their maturation, DC upregulate the chemokine receptor CCR7, which allows for DC entry into lymphatics and subsequently into the draining LN, where DC act as potent APC 23. CX3CR1 LP DC have been shown to migrate via lymph to the mesenteric LN (mLN). CX3CR1 LP DC are reduced in number in the mLN of CCR7-deficient mice 13 and can be observed in lymph vessels connecting the LP with the mLN 10. Observing CX3CR1 LP DC en route is a powerful piece of evidence illustrating a key feature of all extra-lymphatic DC, i.e. to transport antigen to the draining LN and consequently, it was also shown that CX3CR1 LP DC carry Salmonella into mLN after oral infection 12.

In contrast, CX3CR1+ LP macrophages are non-migratory cells. Intra-vital microscopy failed to detect CD11c+(GFP+) cells associated with intestinal epithelia egressing from the LP 7. Correspondingly, cells phenotypically resembling CX3CR1+ macrophages were not detected in mLN 10, 12 and CX3CR1+ MP do not contribute to the cell-associated spread of Salmonella from the intestine to the mLN 12. In support of these findings, a direct examination of cells present in intestinal lymph failed to detect CX3CR1-expressing cells passing through the lymphatic vessels 10. Last but not least, migration of LP DC to the draining LN is thought to require expression of CCR7 24, 25 and consistently CCR7 is present on maturing CX3CR1 DC but not CX3CR1+ macrophages in the LP 10, 12. Thus, the divergent migratory behavior of CX3CR1 and CX3CR1+ MP supports their designation as DC and macrophages, respectively, as proposed here.

Origin and development of intestinal macrophages and DC

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

Recently, two studies published by the groups of Miriam Merad and Steffen Jung analyzed the developmental origin of LP DC/macrophages 12, 14. Notably, in both studies the cells we categorize here as CX3CR1 DC and CX3CR1+ macrophages were named DC (Fig. 1). DC as well as monocytes/macrophages derive from a common precursor referred to as macrophage and DC precursor 26. Macrophage and DC precursors give rise to progenitors of the macrophage/monocyte lineage and a common DC precursor. Common DC precursors in turn represent precursors of plasmacytoid DC and of lymphoid organ pre-DC (pre-cDC) 27, which are selectively committed to produce classical DC 28. The differentiation of plasmacytoid DC as well as DC in lymphoid organs strongly depends on Fms-like thyrosine kinase 3 ligand (Flt3L). Remarkably, Flt3L also induces the expansion of LP DC 29. Conversely, a lack of Flt3L selectively reduced the CX3CR1 DC but not the CX3CR1+ macrophage subset in the LP 12. DC-committed pre-DC give rise to CX3CR1 DC but not CX3CR1+ macrophages. In contrast, CX3CR1+ macrophages differentiate from Ly6chi monocytes in a macrophage colony-stimulating factor-dependent process 12, 14. Therefore, both MP subsets derive from distinct progenitors and their development requires divergent growth factors. Notably, CX3CR1 DC somewhat differ in their late ontogeny from classical lymphoid organs DC, as suggested by the selective reduction in number affecting the latter but not CX3CR1 LP DC in interferon regulatory factor 8-deficient mice 12.

Function of LP DC and macrophages

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

The primary function of LP DC is to transport antigens into mLN to initiate immune responses. LP DC carry intestinal bacteria into the mLN 30, 31 and are critically required to induce oral tolerance to food proteins 25. A common phenotypic fingerprint of LP DC emerges as CD11chiMHCII+CX3CR1F4/80. Expression of CD11b,CD103, CD8α and other markers can be used to further subdivide LP DC. A particular interesting marker of LP DC is CD103. CD103 is expressed by the vast majority of LP DC and CD103's presence correlates with the ability to induce gut-homing molecules on T cells, which are primed by CD103 DC in the mLN 32–34. Moreover, CD103+ LP DC possess an extraordinary ability to direct the differentiation of naïve T cell into Foxp3-expressing Treg 35. Both these attributes of CD103+ LP DC correlate with their high activity of retinoic acid-producing enzymes 36. Interestingly, other DC subset incapable to confer gut-homing capacity to T cells in vitro readily do so in vivo in mLN following intra lymphatic injection 37. This indicates that retinoic acid-producing activity is not reserved to LP DC and might act in concert with other cells and/or mechanisms to induce gut-homing and Foxp3-induction.

Much less is known regarding the function of CD103 LP DC. A direct examination of cells travelling in the lymph following stimulation with the TLR7/8 ligand R848 showed that the majority of DC expressed CD103. Interestingly, the number of CD103 LP DC is reduced in the LP of Id2-deficient mice 12, and immuno-fluorescent microscopy revealed more CD103 DC residing in lymphoid aggregates compared with the surrounding LP 33. This indicates that the majority of CD103 DC might localize in lymphoid aggregates and not constitute bona fide LP DC.

We are left with a dilemma when it comes to the CX3CR1+ macrophages. As already mentioned, numerous studies addressed these cells as DC. Yet, CX3CR1+ macrophages do not migrate in lymph, develop from monocytes and share morphological and phenotypical traits of both DC and macrophages. Both these cell types express MHCII, present antigen and can activate T cells. Consistently, several reports showed in vitro activation of naïve T cells with purified LP macrophages in mice and human 11, 38, 39. But in a direct comparison of CX3CR1+ LP macrophages with LP DC, macrophages displayed a significantly inferior capacity to prime T cells in vitro and in vivo after direct injection into afferent lymph 10. Thus, the criterion antigen uptake/processing/presentation with subsequent capacity to stimulate T cells is only suitable to distinguish DC from macrophages when efficiency is included as an additional parameter. This more quantitative perspective supports the notion that CX3CR1+ LP cells resemble macrophages.

Although in most tissues recognition of bacteria by macrophages is associated with the production of pro-inflammatory cytokines, intestinal macrophages display a high bactericidal activity without overt induction of inflammation 40. This inert behavior of LP macrophages at least in part can be explained by the absence/low expression of TLR 41 and other receptors involved in bacteria-induced macrophage activation (reviewed in 3). Beyond that intestinal macrophages may actively counter-act an onset of inflammation. One of the key signals to induce the anti-inflammatory phenotype might be contributed by CX3CR1. The CX3CR1 ligand CX3CL1 is expressed in a membrane-bound form by gut epithelial cells. In vitro, CX3CL1 can reduce the production of pro-inflammatory cytokine by macrophages 42. Moreover, we observed that compared with wild-type intestinal macrophages those lacking CX3CR1 show a reduced production of IL-10 (our unpublished observations). Thus, the direct contact of CX3CR1-expressing macrophages with gut epithelial cells might contribute to the unique phenotype of intestinal macrophages.

Another well-recognized function of CX3CR1 is to enable the formation of transepithelial cell processes. The CX3CR1-based formation of these cell processes 8 can be triggered by TLR stimulation 7. In vitro, such cell processes facilitate the transport of bacteria across barriers. These observations stirred up much discussion on the relevance of this particular antigen-uptake mechanism in vivo19. The gut epithelium is covered by a thick layer of mucus, and few bacteria come into direct contact with the epithelial cell surface (except for a prominent group of segmented filamentous bacteria). Still, impaired formation of trans-epithelial protrusions in CX3CR1-deficient mice, correlated with reduced dissemination of intestinal bacteria from the intestine to the mLN 8 and reduced accumulation of Salmonella in the LP 43. Thus, it is likely that indeed trans-epithelial sampling occurs in vivo. Given their non-migratory phenotype, CX3CR1+ LP macrophages may not directly carry intestinal bacteria into mLN but instead sampled antigens might be passed on to other cells.

When used as APC in vitro, LP macrophages have been shown to induce differentiation of Foxp3+ Treg 11, as well as Th17 T cells 38; however, in vivo LP macrophages will only very infrequently encounter naïve T cells, which do not efficiently home to the LP. Thus, it appears unlikely that LP macrophages might contribute functionally to a noticeable activation of naive T cells in vivo. Instead, LP macrophages might modulate immune responses locally in the gut. This kind of locally restricted antigen presentation executes its modulatory power exclusively on those lymphocytes that were previously activated in other compartments where they acquired gut-homing capacity. Alternatively, LP macrophages might indirectly influence adaptive immune responses locally in the LP by the secretion of cytokines but independent of an APC-like function and last but not least might do what macrophages do best: phagocytose and eliminate bacteria.

The quest for classical macrophages

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

Finally, the LP contains classical macrophages. These cells seem to constitute a minor population in the healthy small intestine but are more abundant in the colon and their numbers dramatically increase during inflammation. It is difficult to distinguish classical macrophages from CX3CR1+ macrophages discussed before. Some authors used CD11c to discriminate classical CD11c macrophages and CD11c+CX3CR1+ macrophages. Yet, currently, it is unclear whether these are indeed distinct cell populations. In our hands, the vast majority of F4/80+ cells, which would comprise both subtypes, express CD11c when analyzed in tissue sections. Thus, there seems little room for an abundant population of CX3CR1 macrophages present in the healthy small intestine.

Conclusion and further perspective

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

The intestinal LP harbors a variety of hematopoietic and non-hematopoietic cells with antigen-presenting capacity. Apart from LP macrophages and DC discussed here, basophils promote Th2 responses in response to helminth infection 44. Gut epithelial cell express MHCII and might perpetuate the generation of Foxp3+ Treg. Thus, a picture is emerging that several cell types with APC function cooperate in a coordinated fashion to warrant appropriate intestinal immune responses. A classification like the one favored in this Viewpoint offers the potential to gain a better understanding of the immunological control circuits performing this function: classical DC migrate into mLN to activate naïve T cells; the antigen specificity is primarily determined during this LN-based step. Subsequently, these selected and antigen-experienced T cells home to the intestinal LP where local APC, including gut resident macrophages and epithelial cells might modulate the immune response. A particular attractive implication of such model is that the balance between immunity and tolerance, as well as the fine tuning of the response, would be controlled at several levels, i.e. not only at the time of T-cell activation but also locally in the LP.

Acknowledgements

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

This work was supported by Deutsche Forschungsgemeinschaft Grant SFB621-A11 to O.P. The authors apologize to all authors whose work could not be cited due to space constraints.

Conflict of interest: The authors declare no financial or commercial conflict of interest.

References

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. The particle zoo
  4. Quantity matters
  5. Three's a crowd
  6. To migrate or not to migrate?
  7. Origin and development of intestinal macrophages and DC
  8. Function of LP DC and macrophages
  9. The quest for classical macrophages
  10. Conclusion and further perspective
  11. Acknowledgements
  12. References
  13. Supporting Information

See accompanying article: http://dx.doi.org/10.1002/eji.201040562

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.