Intravital microscopy: Imaging host–parasite interactions in lymphoid organs

Intravital microscopy allows imaging of biological phenomena within living animals, including host–parasite interactions. This has advanced our understanding of both, the function of lymphoid organs during parasitic infections, and the effect of parasites on such organs to allow their survival. In parasitic research, recent developments in this technique have been crucial for the direct study of host–parasite interactions within organs at depths, speeds and resolution previously difficult to achieve. Lymphoid organs have gained more attention as we start to understand their function during parasitic infections and the effect of parasites on them. In this review, we summarise technical and biological findings achieved by intravital microscopy with respect to the interaction of various parasites with host lymphoid organs, namely the bone marrow, thymus, lymph nodes, spleen and the mucosa‐associated lymphoid tissue, and present a view into possible future applications.


| INTRODUCTION
Intravital microscopy (IVM) allows visualization of organs in living animals, down to subcellular resolution, to study cellular interactions, cell dynamics, motility, adhesion, rheology and anatomical changes in different tissue compartments through time. Major advances in imaging have allowed more organs, and a wider range of physiological phenomena, to be visualised in vivo (reviewed by De Niz et al., 2019).
Among the most studied systems by IVM is the immune system (reviewed by Secklehner et al2017). Altogether, various methods have been developed to study lymphoid organs, and the voyage of immune cell populations across the body of living animals upon challenges including tumours, infection and inflammation Chtanova et al., 2009;Chtanova et al., 2014;Hampton et al 2015;Ladi et al2008;Torcellan et al 2017).
Immune cells are organised within primary and secondary lymphoid organs, which are vital for innate and adaptive immunity.
Primary lymphoid organs (PLOs) include two anatomically complex tissues: the bone marrow and the thymus, where lymphocyte differentiation and maturation occurs. Secondary lymphoid organs (SLOs) include the lymph nodes, spleen and mucosa associated lymphoid tissues (MALT). These organs host diverse populations of functionally mature, naïve lymphocytes, and are optimally localised across the body to enable efficient surveillance, detection of and response against foreign antigens (reviewed by Ruddle & Akirav, 2009). In parasitology, primary and secondary lymphoid organs have been the focus of interest in three main contexts: immune response generation upon parasitic infection; organ invasion and remodelling by parasites; and the advantages of tropism for parasite biology. In this review, we will focus on key findings on parasite and host-immune cell interactions, visualised within primary and secondary lymphoid organs using IVM. Equally, we will discuss the advances on surgical procedures, optical windows and imaging platforms to visualise these organs.

| PRIMARY LYMPHOID ORGANS IN PARASITIC INFECTIONS
The bone marrow and thymus are the PLOs where the largest part of lymphocyte development occurs. Hosting complex interactions between bone and immune compartments, the bone marrow is critical for haematopoiesis, immunological memory and bone regeneration.
Critical for the development of T cells, the thymus is vital for the generation of strong, yet self-tolerant, adaptive immune responses. Below, we discuss the relevance of performing IVM in these organs, available techniques and key biological findings obtained by IVM in the context of parasitology (summarised in Fig. 1).

| Biological relevance of the bone marrow
The bone marrow is among the organs where IVM has permeated the least, perhaps because of the challenges it represents for imaging.
In addition, the bone marrow possesses a dense vascular network, which occupies about 30% of this tissue's volume. This vascular network is largely heterogeneous, with sinusoidal blood vessels being the most prominent vessel type (Itkin et al., 2016;Spencer et al., 2014). This is relevant in the context of parasite colonization, as the sinusoids are characterized by specific haemodynamics, including slow flow rate and high permeability (Itkin et al., 2016;Jung et al., 2018), which affect whether and how parasites cross the vascular endothelium and establish in the bone marrow.
Bone marrow studies envisaging the use of IVM must consider three key aspects: a) the heterogeneity of the bone marrow niche across the body; b) the suitability and comparability of animal models with human bone marrow composition; and c) accessibility for IVM, which varies across bone locations, making some sites more accessible for imaging.

| Methods for IVM-based visualization of the bone marrow
A commonly preferred site for bone marrow IVM is the calvarium, between the sagittal suture bifurcation and the intersection of sagittal and coronal sutures (Fig. 1Ai). Historically, bone marrow imaging was first performed in the calvarium (Mazo et al., 1998). This location is preferred mostly because it provides easy surgical access, and the calvarial bone marrow is sufficiently thin and transparent to allow light penetration for high resolution image acquisition. To perform IVM in the calvarial bone marrow, one of the possible surgical procedures is to open the scalp skin to access the calvarial bone, and then suture after imaging. This method represents a challenge for longitudinal imaging because the scalp must be opened and sutured for each imaging session, causing significant scarring and tissue damage that often induce inflammation, degrade image quality, and limit the number of imaging sessions that can be performed at the same site (Lo Celso et al, 2011). Therefore, it is preferable to pursue methods that, consistent with the 3R perspective relative to work with living animals, diminish discomfort and damage to the animal. To refine the procedure for improved animal welfare, the mouse calvarial window model was developed. Here, a small section of the scalp is removed, a cover glass is attached to the frontoparietal region of the calvarium, and the exposed skull area is covered with a mixture of dental cement powder and cyanoacrylate glue to prevent re-growth of the membrane layer (Le et al., 2017). Alternatively, the cortical bone can be removed, and only a thin layer of bone left. This removal can be done mechanically, or by laser ablation (Lo Celso et al., 2011, Lo Celso, Flemming, et al 2009, Lo Celso, Wu, & Lin, 2009Turcotte et al 2014).
In 2009, Köhler et al. presented IVM methodology for imaging bone marrow in long bones (Köhler et al., 2009). In their work, the authors argued that although the calvarium had been consistently imaged by IVM, it was unclear whether this area was representative of events taking place in long bones (e.g. the tibia). They described a procedure for imaging the tibia, consisting on removing the skin and muscle on top of the bone with a scalpel and/or an electric drill to obtain a very thin (30-50 μm) layer of bone tissue covering the bone marrow (Fig. 1Aii). In this technique, it is important not to damage this thin bone layer, to prevent vascular collapse in the bone marrow cavity. In 2017, an advance for this procedure was published, whereby in addition to the surgical procedure, a medium-to-longterm imaging window could be incorporated . Conversely, an alternative non-destructive method has been explored to investigate the bone marrow niche, and allows visualization of the intact tibia by two-photon microscopy (Lawson, et al., 2015).
A more recent method involves the development of a microendoscopic multiphoton imaging approach to perform longitudinal imaging deep within the bone marrow (LIMB) at various anatomical locations, including the calvarium, the tibia and the femur (Reismann et al., 2017) (Fig. 1Aiii). This approach consists on surgically implanting into the mouse femur, a biocompatible fixation plate containing a gradient refractive index (GRIN) lens. This setup is based on a fixation plate originally developed to stabilise the femur after an osteotomy (Matthys & Perren, 2009

| Biological findings in parasitology by IVM: the bone marrow
Various parasites home to the bone marrow including Leishmania spp., Trypanosoma spp., Schistosoma spp. and Plasmodium spp. Histology studies on autopsies from humans, cattle and rodent models, FIGURE 1 Windows for IVM and biological findings on primary lymphoid organs (PLOs): bone marrow (A-B) and thymus (C-D). (Ai) The calvarial BM is located close to the intersection of the cortical and sagittal sutures, and the bifurcation of the sagittal suture. Access requires an incision of the scalp to expose the cranium, and either direct access with an objective, or the implantation of a glass window. Methods to access long bones include (Aii) ventral exposure of the tibia and femur, with careful separation of tendons and muscles to allow access to the BM close to the bone head. This can allow for IVM imaging with or without a chronic window; and (Aiii) surgical implantation of a device to allow entry of an endoscope tubing and a GRIN lens for long term imaging and access to deeper sites of the BM of various bones. (B) Key findings in parasitology by IVM include the observation of Plasmodium berghei gametocyte homing to and development within the extravascular space of the BM of long bones. Primary figure shows a still frame from an IVM visualizing mCherry-tagged P. berghei gametocytes in the BM of a UBC-GFP reporter mouse (Adapted from De Niz et al., 2018). The other lymphoid organ is the thymus (C), which is located in the thorax, above the heart. Direct access to the thymus is challenging. A technique used for IVM is the transplantation of the thymus to the kidney capsule, and visualization of the kidney capsule using a dorso-lateral window. Imaging can be done in the dorsal position. (D) Although no IVM has been reported for the thymus in the context of parasitology, various studies have shown that parasites including T. cruzi, T. gondii and Plasmodium alter the thymic architecture, inducing phenomena such as thymic atrophy, accelerated T cell release (T. cruzi and Plasmodium), and increased endothelial cell and CD4+ T cell destruction (T. gondii). (E) Ex vivo images of the thymus of Left panel: control (bottom) and T. cruzi-infected mice (top) showing increased deposition of CXCL12 (green) and fibronectin (red) in infected mice, consistent with altered thymocyte migration (Adapted from Mendes-da-Cruz, Silva, Cotta-de-Almeida, & Savino, 2006). Middle panel: control (bottom) and P. berghei-infected mice (top) showing increased laminin deposition (green) in the thymus of infected mice, consistent with changes in extracellular matrix components upon infection (Adapted from Gameiro et al., 2009)   and Toxoplasma have been detected directly on human bone marrow (Brouland et al., 1996;Jones & Leday, 2014). In rodent models, the presence of Schistosoma influences differentiation at the progenitor level, resulting in induced granulopoietic activity (Joshi et al 2008).
This is thought to be highly linked with maturation of Schistosoma worms, and oviposition (Azevedo et al., 2015;Elkhafif et al., 2010;Kamal et al., 1989). To our knowledge, neither Leishmania spp.,  Care should be taken to avoid damaging vasculature and intestinal epithelium, as well as dehydration. Requires control of peristaltic motion. . IVM showed that as infection progressed, vascular leakage contributes to the entry of sexual and asexual parasite stages to the extravascular space. Finally, mature gametocyte re-entry to the peripheral blood was observed, and was shown to require great deformability (De Niz et al., 2018) (findings are summarised in Fig. 1B).
Altogether, the bone marrow represents a puzzling environment so far greatly understudied in parasitology, for which IVM could contribute significantly in the context of immune responses, host-pathogen interactions, chronic and acute BM remodelling, parasite latency and drug delivery.

| Biological relevance of the thymus
The thymus is a primary lymphoid organ which supports the 'education' of T lymphocytes, resulting in aT cell pool that is self-restricted and selftolerant. Anatomically, the thymus is located in front of the heart, and behind the sternum (Fig. 1Ci). It consists of two merged lobes

| Methods for IVM-based visualization of the thymus
Although T cell responses during infection are a relevant topic for various fields of parasitology, the thymus has not directly been imaged by IVM in situ. Due to its anatomical location, the thymus presents important challenges for direct and continuous visualization. To circumvent such challenges, a frequently used alternative has been the transplantation of the thymus to the kidney capsule (Caetano et al2012). After acceptance of the thymus, the transplanted kidney is exposed, fixed in a stereotactic organ holder, and kept under physiological conditions for in vivo imaging (Fig. 1Cii). To prevent the kidney from returning to its original position, improved methods suggest closing the lateral sides of the skin incision with stitches (Caetano et al., 2012). Although the transplanted thymus approach has been successfully used in various studies, limitations of this approach include the complexity of the surgery required for transplantation, and the fact that different corporal conditions exist in the surrounding environment of a thymus transplanted in the kidney as opposed to a thymus located in the thoracic cavity (Aghaallaei & Bajoghli, 2018). Ex vivo alternatives to the complex transplantation procedure include extraction of individual thymic lobes perfused with oxygenated medium, followed by imaging by two photon microscopy , and imaging thymic slices (Bousso et Pérez et al., 2007;Savino et al, 1989), Plasmodium spp., Francelin et al, 2011;Gameiro et al., 2010) and T. gondii gondii infection (Kugler et al., 2016) (Fig. 1E).
None of the studies performed so far have studied parasite interactions with the thymic resident cells, nor the effects of parasite presence on events such as lymphocyte migration and development in vivo. IVM would enable a better understanding on T cell-mediated immunity, and alterations during infections due to thymic compromise.

| SECONDARY LYMPHOID ORGANS IN PARASITIC INFECTIONS
Following their development in the primary lymphoid organs, competent lymphocytes populate secondary lymphoid organs such as the lymph nodes, the spleen and the MALT. These organs have a key distribution across the body to allow surveillance and efficient responses upon challenges including infections. The LN then lies within the popliteal fossa, and must be extremely carefully separated from surrounding muscles and adipose tissues.

| Biological relevance of the lymph nodes
Once exposed, it is crucial to maintain proper body temperature and prevent dehydration. For the latter point, the LN can be submerged in saline prior to coverage with the glass coverslip (Liou et al., 2012;Mempel et al., 2004). Once the mouse is secure in the customised stage, it is advised that the tail is carefully secured, as it will provide stability to the leg while imaging (Fig. 2Bi). The mouse can then be imaged using an upright microscope directly above the exposed popliteal LN. Various protocols have been suggested to generate the customised stage that best supports mice for IVM (Liou et al., 2012).
Although not yet used in a context of parasitology, a recent development which could be useful to the field is the chronic lymph node window (CLNW) model, which allows longitudinal imaging of the inguinal LNs (Jeong et al., 2015;Meijer et al., 2017). In their work, Meijer et al. designed a stage specifically to allow access to this LN, while  Table 1. where they partially differentiate into exoerythrocytic forms (EEF) (Amino et al., 2006). The skin was also shown to support the development of sporozoites till complete maturation into red blood cells infective forms, a process that was thought to occur exclusively in the liver (Coppi et al., 2011;Gueirard et al., 2010). This implies that the LN draining the inoculation site will receive parasite antigens not just from sporozoites (sporozoites actively reaching the LN or dead sporozoites left in the skin) but also from differentiating parasites (skin EEF aborting at various stages of their development).

| Biological findings in parasitology: the lymph nodes
Regarding the immunological implications of these findings, so far it is known that after an infectious mosquito bite the first cohort of protective CD8+ T cells is primed by DCs in cutaneous LNs (Chakravarty et al., 2007;Obeid et al., 2013). Dynamic in vivo and static imaging has shown the uptake of parasites by LN DC followed by the DC cluster with CD8 + T cells. Indeed, CD8 + T cells are primed by resident CD8α + DCs with no apparent role for skin-derived DCs.
This study established a critical role for LN resident CD8α + DCs in CD8 + T cell priming to sporozoite antigens while emphasizing a requirement for motile sporozoites in the induction of CD8 + T cell-  (Hurrell et al., 2015). IVM also elucidated the dynamics of NK cells following Leishmania major infection. NK cells, which reside in both the LN medulla and paracortex in steady state conditions, were shown to be recruited from the blood to the paracortex where they produce IFN-y that activates parasite-specific CD4 + T cells (Bajénoff et al., 2006) (Fig. 2Ciii and Diii).
IVM imaging of T. gondii-infected mice has shown that neutrophils are able to migrate in a coordinated manner within the LNs . Cooperative action of neutrophils and parasites egressing from host cells triggers neutrophil swarm formation leading to the removal of macrophages that line the subscapular sinus of the lymph node. These results provide insight into the cellular mechanisms that lead to neutrophil swarms and suggest new potential functions for neutrophils in LNs Coombes & Robey, 2010) ( Fig. 2Civ and Div).

| Biological relevance of the spleen
The spleen is a specialised organ that combines the innate and adaptive immune system within a complex architecture. The structure of the spleen enables it to remove older erythrocytes from circulation, thus constantly maintaining a pool of red blood cells best adapted to transport oxygen and iron across tissues. Equally, the spleen's architecture allows efficient removal of blood-borne microorganisms, including parasites. The spleen is organised into three main compartments: the red pulp, where pathogens and senescent red blood cells are removed from the blood by specialised macrophages; the white pulp, a highly organised lymphoid region composed of B and T cell zones; and the marginal zone, which constitutes a bridge between innate and adaptive immune responses due to its specialised macrophage and B cell subsets. The specialised architecture of the spleen is coordinated by the expression of lipid mediators, adhesion molecules and chemokines, which direct the migration and retention of specific lymphoid subsets across the splenic compartments (Reviewed by (Mebius & Kraal, 2005)).

| Methods for IVM-based visualization of the spleen
The spleen presents important challenges for IVM. Two significant microscopy-based limitations imposed by the spleen are a) that the optically dense capsule contributes to significant light scattering, making some anatomical sections of the spleen difficult to access for imaging; and b) that the spleen is highly vascularised and highly complex in its cellular composition, with multiple regions of very fast circulation (Discussed in (Grayson et al, 2001)). Many confocal systems lack the capability of scanning or recording data at a speed rapid enough to allow distinguishing cellular interactions or subcellular events at high speed. In other organs, poor penetration depth has been largely addressed by the use of two-photon microscopes, while speed limitations have been overcome to a certain extent, by spinning disc methods, or opto-acoustic deflectors achieving a minimum of 30 frames per second (Discussed in (Grayson et al., 2001)). IVM techniques adapted for spleen imaging not based on fluorescence, but which overcome the above challenges, include optical frequency domain imaging (OFDI) (Kubo et al., 2017;Otake et al., 2018;Yun et al, 2003), and coherent anti-Stokes Raman spectroscopy (CARS) (Vogler et al, 2015). The former allows detection of scattering properties of tissues at multiple depths to detect angiogenesis and tissue viability, while the latter uses multiple photons to detect intrinsic molecular vibrations, allowing imaging of chemical structures. These techniques have been used in other areas of research, but remain to be introduced in parasitology. An alternative to these methods, is the use of time-and polarization-resolved fluorescence detection, such as fluorescence lifetime imaging (FLIM), whereby the combination of high-speed acquisition and novel methods of image processing, allow for visualization almost in real time (Niesner et al 2008).
In terms of imaging windows, various types of setups have been designed to image the spleen. One includes the abdominal imaging window (AIW) for imaging with either an upright or an inverted microscope (Fig. 3Ai). In 2012, Ritsma et al. developed the abdominal imaging window (AIW), which consists of a titanium ring with a 1mm groove on the side, and a coverslip which can be fixed on the top with glue, and exchanged as required (Ritsma et al., 2012;Ritsma et al., 2013). Following this, the AIW is implanted in the skin and abdominal wall, and held in position by a purse-string suture, which prevents the mice from biting or removing the sutures. The AIW was reported to be used over a maximum of 28 days, without considerable changes in the anatomical position of the window and without disturbing physiological processes. Although the AIW offers major advantages for long term imaging, less complex, temporary windows which simply consist on surgically exposing the organ of interest, hydrating it, and attaching it to a coverslip using glue, can also be used for short-term (4-8 h) imaging. Such setup has been used for Plasmodium imaging, and involves the exposure of the spleen through a small incision on the mouse flank, followed by attachment to a glass coverslip for imaging on an inverted microscope (De Niz et al., 2016;Ferrer et al, 2012) ( Fig. 3Aii). While using glue for attaching the abdominal organs is possible for short-term imaging (i.e. no more than 12 h, continuously), it is not suitable for imaging over a period of days. Alternatively, a vacuumcoupled window can be implemented for imaging using an upright microscope (Fig. 3Aiii); or a glass-bottom cell culture dish, whereby the spleen is exposed and immobilised over sterile saline, allowing use of an inverted microscope (Grayson et al., 2001).
Specific challenges during splenic surgery and exposure include a) that care should be taken to avoid the vasculature while performing the incisions on cutaneous and muscle layers to expose the organ, to prevent bleeding over the spleen during imaging; b) that extreme care should be taken upon pulling the spleen through the muscle, peritoneal and skin incisions in order to neither compress the spleen vasculature (which would hinder circulation and all chances of visualizing cells in motion), nor to cause bleeding or incisions in the spleen (which could be lethal due to rapid exsanguination in extreme cases). Moreover, the intercostal incision should be relatively small, because with larger incisions, the risk that the spleen retracts into the abdominal cavity and away from the window is higher. IVM of the spleen upon even minor splenomegaly (often observed in parasitic infections) should be performed with extreme caution, as the tissue is very fragile and could easily rupture.
In addition to the surgical considerations for imaging, an equally important issue is the biological relevance of the mouse spleen relative to the human spleen. Three key differences between the mouse and human spleens include a) that the marginal zone of the human spleen lacks a delimited marginal sinus and is surrounded by an additional perifollicular zone; b) that the human spleen is sinusoidal, while the mouse spleen is less so; and c) that while the mouse spleen is a key organ for erythropoiesis, the human spleen is less so. This has been relevant in the study of malaria, and might equally be so in the study of other parasitic diseases for which animal models exist. Although pathogens including Plasmodium, T. brucei, T. cruzi, Leishmania, Schistosoma, Babesia, Echinococcus and Paragonimus cause spleen involvement and pathology, to our knowledge only Plasmodium has been imaged by IVM. Window types, and their advantages and limitations, are summarised in Table 1.

| Biological findings in parasitology by IVM: the spleen
As part of its functions, the spleen destroys senescent red blood cells in health, and aids in the detection and immune response formation against blood-borne pathogens. The spleen's involvement in malaria is significant. Splenomegaly is a hallmark of the disease in endemic areas; furthermore, splenectomy both in rodents and humans has been consistently associated with more frequent fever, and higher Moreover, because of its role as a blood surveillance organ, it is believed that the presence of the spleen has driven the evolution of a plethora of evasion mechanisms in Plasmodium to avoid destruction.
Perhaps the most significant is the parasite's ability to sequester in the peripheral vasculature of multiple organsa phenomenon responsible for pathology and complications associated with malaria (reviewed by Engwerda et al 2016)). Previous work to study Plasmodium in the spleen, ex vivo, has included an ex vivo perfusion system whereby the spleen retained its clearing and processing functions and allowed visualization of P. falciparum (Buffet et al., 2006), and a spleen-on-a-chip approach which reproduced splenic basic units To our knowledge, the malaria-infected spleen has been imaged by IVM in four separate studies over the last 6 years. The first spleen IVM study in rodent malaria models focused on the differential remodelling of this organ induced by lethal and non-lethal GFP-expressing P. yoelii strains (Py17XL and Py17X respectively). A significantly higher number of parasites of the non-lethal strain were detected in the spleen, than those of the lethal strain. This was proven to neither be the result of different blood flow, not differential macrophage activity. The nonlethal strain, however, displayed an adhesive rolling-circle behaviour, while the lethal strain did not. Conversely, FITC-labelled uninfected RBCs and fluorescent beads displayed equal flow patterns in mice infected with either strain (Martin-Jaular et al., 2011) (Fig. 3Bi). This differential motion specific to the parasites, together with further MRI, EM and other ex vivo data, led to the conclusion that the nonlethal strain induces a spleen blood barrier of fibroblastic origin, to which Py17X-infected reticulocytes can adhere to, to escape from macrophage clearance. The image-processing pipeline developed to track parasite directionality, residence time, mean velocity and volumetric blood flow allowing for normalization of erythrocyte and lumen vessel diameters, was made available in parallel to the original study . In line with the aim of investigating Plasmodium interactions with splenic populations, a later study went on to explore the role of DCs in the control of blood stage infections with P. yoelii and the chronic malaria model P. chabaudi, at different phases of acute and chronic malaria infections. IVM in this study for the first time showed interactions between DCs and CD4+T cells at different phases of acute malaria, and it was the first work suggesting that aside of their role in antigen presentation, DCs also directly participate in the elimination of iRBCs during acute infection (da Silva et al., 2013). Finally, given that sequestration in the vascular endothelium is believed to be a mechanism developed by Plasmodium to avoid passage through the spleen, a recent study used IVM to demonstrate that nonsequestering P. berghei lines lacking export-mediating Maurer's cleft resident proteins MAHRP1a and SBP1 do not sequester, and extensively accumulate in the spleens, inducing exacerbated splenomegaly (De Niz et al., 2016) (Fig. 3Bii).
The spleen has recently gained great interest in the context of malaria transmission. Studies in humanised mice infected with P.
falciparum , and rodent models of malaria infected with P. berghei , have consistently shown increased numbers of immature gametocytes in the spleen. The relevance of this organ for host-to-vector transmission, as well as its study in the context of transmission-blocking drugs, remains to be fully explored, and IVM could be an extremely relevant tool for this purpose.
Although various technological improvements now allow for greater penetration depths and faster imaging to record cellular interactions, the greatest improvements likely to impact spleen IVM are imageprocessing algorithms for events occurring at high speed. In terms of biology, spleen involvement in the context of immunomodulation and immunopathology have received increased interest, and IVM is a key tool to potentially shed light on parasite-mediated responses. Given its relatively easy use, the implementation of this technique to study other parasites in the spleen, might provide interesting insights into other host-pathogen interactions.

| Other immune organs: the Mucosa-associated lymphoid tissue (MALT)
The mucosa-associated lymphoid tissue (MALT) is situated along the surface of all mucosal tissues, and is responsible for initiating immune responses to antigens encountered at these sites. About half of the lymphocytes of the immune system are in the MALT, with the most studied of such locations being the gut-associated lymphoid tissue (GALT) (reviewed by (Jung et al 2010)), the nasopharynx-associated lymphoid tissue (NALT) and the bronchus-associated lymphoid tissue (BALT) (reviewed by (Bienenstock & McDermott, 2005)). These anatomical locations can be divided into two key sites: effector sites and inductive sites. Inductive sites contain secondary lymphoid tissues which support IgA class switching and clonal expansion of B cells in response to T cell activation. Following activation and IgA class switching, T and B cells migrate from inductive sites to effector sites.
Effector sites are present as diffuse lymphoid tissue along mucosal surfaces, and allow IgA transport across the mucosal epithelium. The functional compartments of the GALT, NALT and BALT are lymphoid follicles, the interfollicular region, the subepithelial dome region, and the follicle-associated epithelium (reviewed by (Cesta, 2006)).

| Methods for IVM-based visualization of the GALT
The intestines are a crucial site in the life cycle of most orallytransmitted parasites, as well as a main site of pathology for the host.
Despite the huge relevance of this site, intestine IVM has not been extensively reported. Aside from the abdominal imaging window ( Fig. 3Ai), recent work by (Kolesnikov et al 2015;Xu et al, 2012) described two alternative methods for imaging the intestines and Peyer's Patches. The surgical procedure involves performing a small incision through the skin along the abdominal midline to expose the peritoneal wall, after which an incision to the peritoneal wall itself follows. Next, a section of the intestine was externalised, and an incision made, further exposing the luminal surface. Care should be taken in particular to a) avoid disrupting the mucosal layer and b) avoid damaging the mesentery or mesenteric vasculature. For this purpose, the use of an electrocautery is recommended (Kolesnikov et al., 2015), rather than a scalpel, to seal off vessels damaged during the surgical procedure. Aside of preventing bleeding, controlling intestinal peristaltic movements and the associated motion artefacts is key. Imaging methods to overcome this issue have been discussed (Kolesnikov et al., 2015;Xu et al., 2012). If using an inverted microscope, upon completion of the surgery, the externalised intestine can be sandwiched between a coverslip and sterile gauze at the animal's abdomen. In this setup, the animal's weight stabilises the tissues and prevents dehydration (Xu et al., 2012). Conversely, an imaging chamber suitable for imaging using an upright microscope can be used, which is arguably better suited for physiological preparations (Kolesnikov et al., 2015). Two key considerations for intestine IVM which differ from other tissues are on one hand, that autofluorescence areas in the mucosal layer exist which may represent challenges for imaging (although such autofluorescence is absent from the epithelia, and the lamina propria). On the other hand, live intestinal tissue is more sensitive to imaging than the skin, spleen, or other tissues which allow long imaging periods. The intestinal tissue reportedly begins to deteriorate within a 3h window following surgery and exposure to air. Internal controls to monitor membrane integrity, vascular permeability and cellular motility are therefore highly recommended during intestinal IVM (Kolesnikov et al., 2015). Similar to ex vivo alternatives for other organs, imaging can be performed in intestinal sections cultured in aerated media (Chieppa et al, 2006;Coombes et al., 2013).
The downside of this method is that the vascular, lymphatic and nervous connections of the tissue are severed, thus preventing a full reproduction of physiological conditions of the intestines. Window types, and their advantages and limitations, are summarised in Table 1.
To our knowledge, only T. gondii has been imaged in the intestines by IVM (Coombes et al., 2013;Watanabe et al., 2018). A more recent study used IVM to investigate the leucocyte's behaviour in the mesenteric circulation, as well as leucocyteendothelium interactions following infection with T. gondii (Watanabe et al., 2018). The study's key findings were that haemodynamic changes already happen at the beginning of infection, in which molecules such as ICAM1, PECAM1 and P-selectin are upregulated on the vascular endothelium. These not only contribute to the parasite's adhesion and transmigration, but also to leucocyte mobilization within the mesenteric vasculature, resulting in a mild inflammatory response.

|
Given the involvement of the gastrointestinal tract and the lungs by a large number of parasites, the relevance of the MALT for pathology and immunity, and the little understanding of the dynamics we have of parasites in these organs, the implementation of IVM has vast potential in the years to come. shows the abdominal intravital window, which gives access to the abdominal organs including the gut and the spleen (on the dorso-ventral side). The chronic window consists on the implantation of titanium rings to enable imaging. (Aii) shows a temporary window consisting on the exposure of the spleen, and visualization through a coverslip using an inverted microscope. (Aiii) is an alternative whereby the mouse can be imaged using an upright microscope. For stabilization, the coverslip and organ are immobilised using a vacuum. Main biological findings in the spleen using IVM involve Plasmodium infections, namely (Bi) that lethal and non-lethal strains of P. yoelii differentially remodel the spleen. The nonlethal strain induces the formation of a barrier of fibroblastic origin which protects the parasites from phagocytosis, causing vast splenomegaly. Primary image shows the spleen of a mouse infected with non-lethal P. yoelii 17X (green) at day 3 post-infection, showing FGF8 as a marker of a barrier of fibroblastic origin (red) and F4/80 macrophages (blue). (Adapted from Martin-Jaular et al., 2011). (Bii) shows that at early synchronised infections, P. berghei ANKA schizonts sequester in the vascular periphery, avoiding the spleen, while parasites unable to sequester (PbMAHRP1a and PbSBP1) are absent from the vasculature in the periphery, but are present in vast numbers in the spleen, where they are eliminated. Primary images (middle and bottom panels) show spleens of reporter mice (UBC-GFP, green) infected with P. berghei WT and PbSBP1 (red), 16-18 h following intravenous injection of synchronised schizonts. (Adapted from De Niz et al., 2016). GALT. The GALT can be observed using various windows including the AIW (Ai). Biological findings in parasitology by IVM are restricted to observation of T. gondii, whereby (C) neutrophil recruitment was observed to the villi and mucosal epithelium. Within neutrophils, T. gondii survives and disseminates to other organs. Primary image shows T. gondii-containing LysM-GFP cells (yellow) migrating across the intestinal epithelium (red). Images were obtained by two-photon microscopy. (Adapted from Coombes et al., 2013). All original images were published under a Creative Commons Attribution (CCA) license, and/or reproduced with permission

| OUTLOOK
In this review we have discussed the relevance of primary and secondary lymphoid organs in parasitic infections, and the knowledge we have gained as a community through the use of IVM to investigate dynamic interactions in organs including the bone marrow, spleen, lymph nodes and GALT.
Interestingly, an alternative technology that has been explored for two parasites (T. gondii and Plasmodium), in the context of lymphoid organs, is the use of organs-on-chip (OOCs), and of organoids.
Organoids consist of organ-specific adult mammalian stem cells grown in a three-dimensional structure while OOCs are based on a similar principal as organoids, but with the addition of microfluidic systems to allow for the distribution of nutrients and soluble mediators throughout the constructs. The cells in both systems self-organise into morphologically distinct layers that are capable of successfully mimicking organs. One of the major benefits of OOCs is that they can mimic blood flow, allowing for the assessment of shear stress and deformability of circulating cells.
The use of a spleen on a chip model demonstrated the poor deformability of P. falciparum pRBCs and their ability to occlude narrow vessels (Picot et al., 2015;Rigat-Brugarolas et al., 2014). Another important advance has been the generation of an organoid for the intestinal epithelium to study T. gondii interactions with the epithelium and immune cell populations (Derricott et al., 2019). While both organoids and OOCs are still in their infancy, they have great potential for assessing parasite behaviour in a consistent and reproducible manner.