Novel in vivo imaging techniques for the liver microvasculature

Abstract

Intravital microscopy is a valuable tool in studying the liver, a complex organ with a unique sinusoidal microcirculation and both metabolic and immune functions. The liver is also subject to a large variety of diseases including viral, bacterial and parasitic infections. We developed novel recording techniques to visualize dynamic events in the hepatic microvasculature without the need of fluorescent markers. In combination with cellular and molecular probes, reporter mice and Plasmodium as a hepatotropic model microorganism, we demonstrate the power of these techniques in monitoring the development of the malaria parasite and the response of the hepatic microenvironment to infection.

Keywords: :

• intravital imaging
• liver
• vasculature
• endothelium
• Kupffer cell
• T cell
• hepatocytes
• blood flow
• reflection
• autofluorescence

Introduction

The liver is unique in that it receives not only oxygenated arterial, but also nutrient-rich portal venous blood from the gastrointestinal tract. Upon entry into the liver lobule, the two blood sources merge and flow along the sinusoids toward the central venule. Because portal venous blood carries food antigens and microbial products from the intestines, the liver is exposed to a plethora of innocuous dietary and commensal antigens, to which it typically responds with tolerance, as well as potential pathogens or toxins against which it normally mounts an immune response.^1,2 Kupffer cells, resident professional phagocytes of the liver, are a key component of the unique hepatic innate immune system and crucially involved in the dichotomy between induction of tolerance or immunity.³ Based on these unique features, combined with easy accessibility and the continuously improving imaging technology, the liver has been a major focus of intravital microscopy (IVM) for many years. While early studies relied on video microscopy and transillumination, which limited analysis to the edge of the liver, the subsequent use of wide-field epifluorescence microscopy greatly extended the area of examination.^4-7 Technical advances in instrument design over the last years have to led to confocal, spinning disc and 2P microscopes with vastly improved IVM capabilities.^8-14

Designed as a barrier between the digestive system and the rest of the body, the liver is subject to a multitude of microbial infections. For example, Plasmodium undergoes a round of multiplication in hepatocytes and, by choosing this route of entry into the host, may be able to exploit the tolerogenic properties of the liver.^15,16 IVM has been instrumental in clarifying the dynamics of Plasmodium sporozoite entry into the liver and the release of thousands of merozoites into the blood,^13,17,18 which eventually leads to the typical malaria-associated symptoms and pathology. Despite decades of research, however, the elusive clinically silent Plasmodium liver stage (LS) is still a major focus of anti-malarial intervention and vaccine development,^19-22 highlighting our incomplete understanding of the basic immunobiology of Plasmodium within the hepatic microenvironment.

Here we present novel recording techniques that visualize dynamic events in the hepatic microvasculature without the need of fluorescent markers. Using Plasmodium yoelii as a hepatotropic model microorganism, we demonstrate, in real-time and under normal blood flow conditions, the development of the malaria parasite in the liver of live animals with an intact immune system and preserved innervation.

Results

To obtain a more detailed view of the hepatic architecture, in particular the sinusoidal microvasculature, we developed a novel approach based on simultaneous recording of reflected and autofluorescent signal for use in non-fluorescent mice. This technique, which we call intravital reflection recording (IRR), visualizes the sinusoidal walls and blood cells in the absence of a fluorescent tracer (Fig. 1A–C). When IRR is recorded in inverse color mode, blood cells appear as dark objects traveling through the microvasculature in bright (non-fluorescent) plasma (Fig. 1D–F). Combined with reporter mice, for example ECFP (Fig. 1B) or DsRed (Fig. 1E) mice that express fluorescence predominantly in hepatocytes, IRR dramatically improves visualization of the overall liver architecture by generating an image of the sinusoidal microvasculature complimentary to the fluorescent parenchymal signal. Direct (Vid. S1) and inverse IRR (Vid. S2) are ideally suited to monitor the blood flow or immune cells patrolling the sinusoidal microvasculature. Considering the multitude of fluorescent markers and reporter mice available, the ability to use any laser line for IRR in a complementary color presents a major advantage of this technique.

Figure 1. IRR visualizes blood flow in the non-fluorescent sinusoidal microvasculature. (A) IRR was obtained with a 633 nm HeNe laser at 25% power output and collected at 630–850 nm. In this 3D projection, sinusoids appear as red tubes due to the sinusoidal wall and dense packing of blood cells reflecting the excitation light. (B) ECFP mouse liver fluorescence was excited with a 405 nm Diode laser to visualize the hepatocyte plates. (C) Overlay shows the intricate network of liver parenchyma and highly anastomozed sinusoids. (D) Inverse IRR highlights the sinusoidal microvasculature, while DsRed mouse fluorescence, obtained with 543 nm HeNe laser, visualizes the hepatocyte plates (E). (F) Overlay allows for a 3D representation of the non-fluorescent structural details of the sinusoids (red) and the DsRed fluorescence of the liver tissue (green pseudocolor). Scale bars 20 μm.

Another novel approach for IVM of the hepatic microvasculature includes new membrane stains. CellMask Orange, for example, provides a detailed outline of the sinusoidal wall due to the high cell membrane density along the Disse space (Fig. 2A and C) and visualizes the sinusoidal blood flow (Vid. S3). In the avascular subcapsular region of the liver, CellMask Orange primarily stains hepatocyte membranes (Fig. 2B). Conventional MitoTracker dyes represent excellent parenchymal markers due to the high concentration of mitochondria in hepatocytes (Fig. 2D). In combination with MitoTracker, CellMask (Fig. 2E) provides a detailed view of the sinusoidal microvasculature and the radial hepatic plates, thus visualizing the overall architecture of the liver lobule (Fig. 2F). Even though representing a more traditional approach, endothelial labeling with CD31-specific antibodies (Fig. 3A) offers enhanced information on the microvasculature if used in conjunction with IRR.

Figure 2. Cell membrane and mitochondrial staining visualizes the sinusoidal liver architecture. (A–C) In the liver, the cell membrane stain CellMask Orange (shown in green) highlights the sinusoidal microvasculature due to the high density of cell membranes along the space of Disse. (A) 3D reconstruction of a z-stack. (B) A z-slice from the avascular subcapsular region of the liver demonstrates that CellMask Orange predominantly outlines hepatocytes. (C) Another z-slice, acquired 40 μm from the liver surface, visualizes that CellMask Orange outlines the microvasculature as well as individual blood cells within the sinusoidal lumen. (D–F) Liver from a mouse injected with MitoTracker Deep Red (D) and CellMask Orange (E). The overlay (F) visualizes the sinusoidal microvasculature outlined with CellMask Orange, while the hepatocyte cytoplasm is filled with red fluorescent mitochondria. Distortion of the RBCs in the sinusoidal lumen indicates preserved blood flow. Scale bar 20 μm.

Figure 3. Novel complementary approaches for visualization of the sinusoidal microvasculature. (A) In this 3D reconstruction of a lysM-EGFP mouse liver z-stack, the sinusoidal microvasculature is labeled with anti-CD31 Alexa Fluor 647 (red). Myelomonocytic cells in this reporter mouse fluoresce green and nuclei are blue (Hoechst). (B) Z-slice of a Tie2-GFP mouse liver with fluorescent endothelia (green). Hepatocytes are red (MitoTracker), CD8⁺ T cells were labeled with anti-mouse CD8a-PE (green outline) and nuclei are blue (Hoechst). See Vid. S8 for complete z-stack. (C) 3D projection of a liver z-stack showing anti-CD8a-PE labeled CD8⁺ T cells (red circles) patrolling the sinusoidal endothelium (green). Nuclei are blue (Hoechst).(D) Z-slice showing red hepatocyte plates (MitoTracker) lined with green fluorescent sinusoidal endothelia. Nuclei are blue (Hoechst). Scale bars 20 μm.

A large variety of reporter mice are available to study interactions between liver antigen presenting cells (APC) and immune cells. As sinusoidal endothelia can function as APC,^3,23 Tie2-GFP reporter mice²⁴ are useful tools for monitoring CD8⁺ T cell adhesion to the sinusoidal microvasculature (Fig. 3B). When separate PMTs are used to collect signal from of anti-CD8a-phycoerythrin (PE) labeled CD8⁺ T cells and GFP expressing endothelia (Fig. 3C; Vid. S4), the blood flow remains invisible without IRR (Vid. S5). When PE and GFP signal is acquired with one detector and IRR with another (Vid. S6), CD8⁺ T cells patrolling the sinusoids can be monitored together with the blood flow (compare Vids. S5 and S6). Tie2-GFP mice may also be used in conjunction with MitoTracker to visualize the Disse space (Fig. 3D) or with anti-CD8a-PE to monitor CD8⁺ T cells (Vids. S7 and S8). Further, lysM-EGFP reporter mice with fluorescent myelomonocytic cells²⁵ are valuable to study innate immune cells or interactions between antigen-presenting Kupffer cells and CD8⁺ T cells. Anti-CD8a-PE surface labeling allows differentiation of CD8⁺ T cells from solid green GFP⁺ myelomonocytic cells, despite considerable overlap of the two emission spectra (Fig. 4A), but the blood flow remains invisible without IRR (Vid. S9). Capturing surface and cytosolic signal from different cell types in one PMT provides the advantage that other detectors remain available for IRR (Vid. S10), inverse IRR (Vid. S11), or additional fluorochromes such as nuclear stains (Vid. S12). CD11c-EYFP mice²⁶ injected with mitochondrial markers allow monitoring dendritic cell motility within the liver parenchyma (Fig. 4B) and together with IRR, observation of blood flow (Vid. S13). Finally, crossbreeding reporter mice that express complimentary fluorochromes in different cell types represents another useful scheme for IVM. For example, DsRed x lysM-EGFP hybrid mice enable study of the behavior of myelomonocytic cells in the uniformly red fluorescent liver parenchyma (Fig. 4C).

Figure 4. Complementary imaging techniques for reporter mice with fluorescent immune cells. (A) Single z-slice of a lysM-EGFP mouse liver with fluorescent myelomonocytic cells (solid green cytoplasm, in this case a Kupffer cell). MitoTracker was used to visualize hepatocytes (red), CD8⁺ T cells were labeled with anti-mouse CD8a-PE (green outline) and nuclei are blue (Hoechst). (B) 3D reconstruction of a z-stack from a CD11c-EYFP mouse liver with fluorescent dendritic cells (yellow). Parenchymal cells are red (MitoTracker), nuclei are blue (Hoechst). Scale bar 50 μm. (C) A DsRed x lysM-EGFP hybrid mouse has red fluorescent hepatocytes and green fluorescent Kupffer and myelomonocytic cells. All other scale bars 20 µm.

Imaging malaria parasites in the liver

Using the P. yoelii infected BALB/c mouse, currently the model of choice due to its superior LS development rate and negligible induction of unspecific inflammation,²⁷ we demonstrate the ability of IRR to visualize dynamic events in the microvasculature surrounding the parasites. Fluorescence emission by GFP expressing P. yoelii XNL (PyXNL-GFP) allows detection of both immature (Fig. 5A; Vid. S14) and mature LS (Fig. 5B). Direct (Fig. 5C) and inverse (Fig. 5D) IRR dramatically enhances visualization of the sinusoidal architecture. Supported by traditional organelle markers that visualize the non-fluorescent liver parenchyma, these techniques readily demonstrate vascular perfusion around the LS (Vid. S15).

Figure 5. Using IRR to visualize the sinusoidal blood flow surrounding Plasmodium LS in non-fluorescent mice. (A) 3D volume view of a mouse liver harboring a 22 h PyXNL-GFP LS (green). Hepatocytes are labeled with MitoTracker (red), nuclei are blue (Hoechst). (B) MitoTracker (red) labels the hepatocyte plates surrounding two 42 h PyXNL-GFP LS (green). (C) IRR highlights the sinusoidal microvasculature (red) surrounding a 42 h PyXNL-GFP LS (green). Optical xy-xz-yz slice view, obtained with 50% 594 nm and 633 nm HeNe laser output power and PMT set to 590–850 nm. Nuclei are blue (Hoechst). (D) 3D volume view shows a 42 h PyXNL-GFP LS (green) embedded in non-fluorescent liver tissue. The dense sinusoidal network (red) was visualized using 50% 594 nm and 633 nm HeNe laser output power and setting the PMT to 590–850 nm. Nuclei are blue (Hoechst). Scale bars 20 μm.

In situations where reporter mice and parasites share the same fluorescent protein, for example PyXNL-GFP infected Tie2-GFP mice, Plasmodium LS can nevertheless be distinguished from endothelia based on morphology (Fig. 6A). Parasite host combinations with complementary fluorochromes, for instance PyXNL-DsRed infected lysM-EGFP mice, are preferable as they allow analysis of close encounters between parasite and host immune cells (Fig. S1). Combined with IRR, this approach significantly improves visualization of immune cell adhesion to the endothelium (Vid. S16). Similarly, dendritic cell interactions with LS are readily traceable in PyXNL-DsRed infected CD11c-EYFP mice (Fig. 6B).

Figure 6. Visualization of the sinusoidal architecture aids in analysis of the Plasmodium infected liver. (A) Optical xy-xz-yz slice view of a 42 h PyXNL-GFP LS in a Tie2-GFP mouse liver. Despite expression of the same green fluorescent protein, endothelia can readily be distinguished from the large LS (arrows). Kupffer cells were identified by phagocytic uptake of dextran-rhodamine (red). (B) 3D reconstruction showing a 42 h PyXNL-DsRed LS (red) surrounded by dendritic cells (green) in a CD11c-EYFP mouse liver. Nuclei are blue (Hoechst). (C) Optical xy-xz-yz slice view of a non-fluorescent 42 h PyXNL-wt LS in a DsRed mouse liver. The parasite is identifiable by the large number of small blue merozoite nuclei (Hoechst). The sinusoidal architecture was visualized by IRR (green) using 40% output power of a 633 nm HeNe laser and PMT set to 630–750 nm. DsRed fluorescence is most pronounced in hepatocytes. See Figure S2 for individual color channels. (D) A non-fluorescent 42 h PyXNL-wt LS in a lysM-EGFP mouse with green Kupffer and myelomonocytic cells. Hepatocytes are red (MitoTracker). The small punctate merozoite nuclei of the mature LS can readily be distinguished from the much larger hepatocyte nuclei (both stained blue with Hoechst). Scale bars 20 µm.

In contrast to fluorescent parasites, wt parasites are inherently difficult to locate in the liver. Because wt parasites frequently produce considerably higher sporozoite loads in the mosquitoes’ salivary glands compared with fluorescent mutants, and consequently less time needs to be spent on purification of the several million sporozoites required for IVM, we deemed it essential to develop a method for identification of wt LS in the liver when fluorescent parasites are scarce or not available.

Distinct punctate DNA staining of the hundreds of merozoite nuclei that eventually fill the entire volume of mature parasites readily identifies the location of non-fluorescent PyXNL-wt LS in the liver parenchyma of DsRed (Fig. 6C; Fig. S2) or Tie2-GFP mice (Fig. S3). In lysM-EGFP mice, combined mitochondrial and nuclear staining facilitates determining the position of PyXNL-wt LS relative to recruited myelomonocytic cells (Fig. 6D; Fig. S4). IRR dramatically improves visualization of blood flow and the behavior of myelomonocytic cells patrolling the sinusoids surrounding PyXNL-wt LS (Vid. S17). By providing structural details of the liver parenchyma, mitochondrial staining also reveals the space occupied by non-fluorescent LS (Fig. S4). We anticipate that these methods will be useful to visualize, in the absence of fluorescent reporters, P. falciparum parasites in humanized mouse models.²⁸

Discussion

Collection of IRR signal from non-fluorescent structures not only greatly improves visualization of the sinusoidal microvasculature and the flow of blood cells, but also provides advantages over more traditional approaches. IRR generates a completely uniform image of the microvasculature, while vascular endothelia of Tie2-GFP mice, for example, fluoresce in the arterial rather than the venous portion of the microvasculature and thus highlight the periphery rather than the center of the liver lobule¹⁸ (Frevert, unpublished data). Similarly, the widely used endothelial marker PECAM-1 is not universally expressed throughout the entire microvascular tree.²⁹ Also, cross-linking of cell surface molecules by antibody-mediated labeling may elicit unwanted cellular responses, an important consideration for IVM studies.

Without a fluorescent tracer, the hepatic microvascular plasma appears typically dark when imaged by epifluorescence/video, confocal or 2P microscopy.^13,18,30-32 IRR is particularly valuable for organs such as liver, spleen or bone marrow, whose sinusoidal capillaries are lined with fenestrated endothelia. Soluble vascular markers, while widely used for IVM of organs with a continuous endothelium, are less suitable for long-term microcirculatory studies of the liver because the sinusoidal hepatic microvasculature allows tracers such as fluorochrome-conjugated albumin or low molecular weight dextran to rapidly diffuse out of the vascular lumen into the space of Disse where they are internalized by hepatocytes, either for degradation or for secretion into the bile.^4,14 In fact, small dyes such as rhodamine 123 have been used to measure the kinetics of hepatocytic uptake.³¹ Further, Evans blue, a tracer that is commonly used as a fluorescent plasma marker based on its reaction with albumin,^33,34 conjugates to other proteins as well and also stains cells thus diminishing its value for liver IVM. Large fluorescent molecules, beads, or liposomes, while ideally suited for clearance studies, are equally inappropriate as tracers for the hepatic microcirculation, because any corpuscular material is rapidly cleared from the blood by Kupffer cells, resident phagocytes that line the liver sinusoids in large numbers.^18,35-41 IRR is also well suited for microcirculatory studies of other organs, for example the brain, where fluorescent tracers remain strictly confined to the vascular lumen. Monitoring blood brain barrier impairment and the ensuing uptake of plasma into the parenchyma will still require the use of vascular tracers, however (Movila et al., unpublished data).²⁹

Prerequisite for monitoring microbial intruders in the hepatic microenvironment and the ensuing innate or acquired response is visualization of immune cells and their interaction with the various non-parenchymal antigen-presenting cells of the liver.^3,23 Many years of liver IVM in the McCuskey lab have provided detailed knowledge on the biology and function of Kupffer cells and the regulation of the hepatic microcirculation in health and disease.^30,36-41 More recently, extensive liver IVM studies in the Kubes lab have revealed that recruitment of neutrophils and platelets in sepsis and sterile inflammation is mediated by CD44/hyaluronan and Mac-1/ICAM-1, respectively.^42-47 Liver IVM was used to monitor CD1d-reactive CXCR6-positive NKT cells patrolling the sinusoids based on recognition of endothelial CXCL16,¹⁰ to study macrophage and T cell dynamics in the course of mycobacterial granuloma formation,^11,12 and to characterize the behavior of CD8⁺ effector T cells in the livers of Plasmodium infected and immunized mice (Cabrera et al., manuscript in preparation). We envision that future studies aiming at visualizing immune cells traveling in the hepatic microvasculature, adhesion to the endothelium and invasion of the parenchyma will greatly benefit from direct or inverse IRR. Further, IRR will help associate the upregulation of molecules such as ICAM-1, CXCR6, or CD14 during vascular inflammation with specific sections of the microvasculature.^10,29,48 Last not least, because inverse IRR depicts (non-fluorescent) blood cells as bright objects on a dark background, this approach may eliminate the need for re-inoculation of ex vivo labeled erythrocytes for velocity measurements.⁵ Thus, the potential applications of this novel imaging technique are manifold.

Overall, facilitated by the increasing availability of imaging instrumentation with spectral capability and a vast fluorochrome library, the presented approaches are readily adaptable to two-photon microscopy for greater depth of data acquisition or spinning disk confocal microscopy for enhanced speed, for study of the broad spectrum of microbial infections of the liver, analysis of other hepatic functions and IVM of other organs.

Materials and Methods

Ethics statement

This study was conducted in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee, NYU School of Medicine (Protocol number 120213-01). All surgery was performed under ketamine/xylazine/acepromazine anesthesia, and all efforts were made to minimize suffering.

Parasites

Wild-type P. yoelii strain 17XNL (PyXNL-wt), P. yoelii 17XNL expressing GFP (PyXNL-GFP)¹³ and P. yoelii 17XNL expressing DsRed (PyXNL-DsRed)⁴⁹ were purified from the salivary glands of female Anopheles stephensi mosquitoes. PyXNL-GFP and PyXNL-DsRed parasites emit fluorescence throughout the entire life cycle.

Mice and crossbreeding

Swiss Webster mice (Taconic Farms, Inc.) were used for propagation of the parasite cycles. Because of the limited liver surface area available for IVM analysis, high parasite densities are essential for acquisition of sufficient data sets for statistical analysis. BALB/c mice (Taconic Farms, Inc.), in particular the BALB/cAnNHsd strain (Harlan Laboratories), were chosen based on their superior susceptibility to infection with P. yoelii. To generate maximum numbers of Plasmodium LS in fluorescent reporter mice, the following mouse strains were backcrossed into BALB/c background: DsRed mice that express red fluorescent protein in all cells (B6.Cg-Tg(CAG-DsRed*MST)1Nagy/J, Stock 006051, Jackson Laboratory); ECFP mice that express cyan fluorescent protein in all cells (B6.129(ICR)-Tg(CAG-ECFP)CK6Nagy/J, Stock 004218, Jackson Laboratory); Tie2-GFP mice that express GFP in vascular endothelia (TgN(TIE2GFP)287Sato/J, Stock 003658, Jackson Laboratory)²⁴; lysM-EGFP mice that express EGFP in the myelomonocytic lineage (a kind gift from Dr Thomas Graf)²⁵; and CD11c-YFP mice that express EYFP in dendritic cells [B6.Cg-Tg(Itgax-EYFP)1Mnz/J, Stock 008829, Jackson Laboratory].²⁶

Anesthesia, surgery and intravital microscopy

Mice were infected by intravenous inoculation of at least 1–2 million sporozoites. At various time points after infection, mice were anesthetized with a cocktail of 50 mg/kg Ketamine HCl (Ketaset, Fort Dodge Animal Health), 10 mg/kg Xylazine (Lloyd Laboratories, 4821) and 1.7 mg/kg Acepromazine Maleate (Butler Animal Health Supply, 003845) (KXA mix) as described.^13,18 The mouse peritoneal cavity was opened along the rib cage and the liver exposed for IVM. The mouse was then positioned on its left side on a custom-made microscope stage adaptor with a 3 cm opening covered with a 45 × 50 mm coverslip (Fisher Scientific, 12-545H). The left liver lobe was gently teased out onto the coverslip with a cotton-tipped applicator and immobilized with gauze moistened with PBS pH 7.0. The adaptor was then inserted into the stage of an inverted Leica DMIRE2 microscope equipped with a temperature controlled Ludin chamber and the liver was analyzed with a Leica TCS SP2 AOBS confocal system (40× HCX PL APO 1.25–0.75 oil lens) with the following excitation light sources and laser lines: 405 nm (Diode laser), 488 nm (Argon/ Krypton laser) and 543, 594 and 633 nm (HeNe lasers). Periodic reinjection of KXA mix at 60–90 min intervals allowed imaging for at least 4 h.

Direct and inverse IRR

To visualize the architecture of the liver tissue and the sinusoidal blood flow in non-fluorescent mice, a combination of reflected light and autofluorescence was recorded. Liver autofluorescence was collected at a distance of least 5 nm from the excitation line using any one of the HeNe lasers. Reflected light was collected in a narrow wavelength setting of 3 nm on either side of the excitation line. IRR was done by collecting reflection plus autofluorescence in one of the AOBS photomultiplier tubes (PMT) by setting the spectra cutoffs to include a reflection excitation line and its corresponding autofluorescence emission, or by collecting autofluorescence from a shorter HeNe line plus reflection from the next longer laser excitation source. Inverse IRR was done correspondingly except that one of the inverse color Leica look-up tables (LUT) was used.

Cellular markers

Prior to imaging, mice were intravenously or intraperitoneally inoculated with 648 nmol of the nuclear stain Hoechst 33342 (λ_ex = 405 nm, Invitrogen, H3570), 18.4 nmol MitoTracker Deep Red (λ_ex = 633 nm, Invitrogen, M22426), 7.5 μg CellMask Orange plasma membrane stain (λ_ex = 543 nm, Invitrogen, C10045), 125 pmol rhodamine-dextran (MW 2,000 kDa, λ_ex = 543 nm) to visualize Kupffer cell phagocytosis (all from Invitrogen), 2 μg of PE-conjugated anti-mouse CD8a (Clone 53–6.7, λ_ex = 488 nm; eBiosciences, 12-0081), or 4.5 μg of Alexa Fluor 647 conjugated anti-mouse CD31 (Clone MEC13.3, λ_ex = 633nm; BioLegend, 102516) in matching color combinations.

Image processing

After acquisition with Leica LCS software, 2D, 3D or 4D data sets were reconstructed and processed in Imaris 7.4.2 (Bitplane). Composite figures were generated in Adobe Photoshop CS4 (Adobe Systems Inc.).

Abbreviations:
APC	=	antigen-presenting cell
CFP	=	cyan fluorescent protein
LS	=	liver stage
GFP	=	green fluorescent protein
IRR	=	intravital reflection recording
IVM	=	intravital microscopy
LUT	=	lookup table
PE	=	phycoerythrin
PMT	=	photomultiplier tube
PyXNL	=	Plasmodium yoelii strain XNL
YFP	=	yellow fluorescent protein
2P	=	two-photon

Acknowledgments

The authors wish to thank Drs Elizabeth Nardin, P’ng Loke, Jerome Vanderberg, and Stefan Kappe for helpful comments on the manuscript. The work was supported by grants RO1 AI070894 and S10 RR019288 from the National Institutes of Health (U.F.).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplemental Materials

Supplemental materials may be found here: www.landesbioscience.com/journals/intravital/article/23423