A contrast agent recognizing activated platelets reveals murine cerebral malaria pathology undetectable by conventional MRI

Human and murine cerebral malaria are associated with elevated levels of cytokines in the brain and adherence of platelets to the microvasculature. Here we demonstrated that the accumulation of platelets in the brain microvasculature can be detected with MRI, using what we believe to be a novel contrast agent, at a time when the pathology is undetectable by conventional MRI. Ligand-induced binding sites (LIBS) on activated platelet glycoprotein IIb/IIIa receptors were detected in the brains of malaria-infected mice 6 days after inoculation with Plasmodium berghei using microparticles of iron oxide (MPIOs) conjugated to a single-chain antibody specific for the LIBS (LIBS-MPIO). No binding of the LIBS-MPIO contrast agent was detected in uninfected animals. A combination of LIBS-MPIO MRI, confocal microscopy, and transmission electron microscopy revealed that the proinflammatory cytokine TNF-α, but not IL-1β or lymphotoxin-α (LT-α), induced adherence of platelets to cerebrovascular endothelium. Peak platelet adhesion was found 12 h after TNF-α injection and was readily detected with LIBS-MPIO contrast-enhanced MRI. Temporal studies revealed that the level of MPIO-induced contrast was proportional to the number of platelets bound. Thus, the LIBS-MPIO contrast agent enabled noninvasive detection of otherwise undetectable cerebral pathology by in vivo MRI before the appearance of clinical disease, highlighting the potential of targeted contrast agents for diagnostic, mechanistic, and therapeutic studies.

Cerebral malaria (CM) is a major cause of death in tropical and subtropical regions, killing over 3 million individuals worldwide per year (1). Platelet sequestration in the cerebral microvasculature plays a pivotal role in the pathogenesis of CM (2), and antiplatelet therapy has been shown to improve outcome in CM patients (3, 4). The histopathology of mice infected with Plasmodium berghei ANKA reveals extensive damage to vascular endothelial cells and plugging of vessels caused by platelet thrombi (5). Similarly, immunohistochemistry for the platelet-specific glycoprotein IIb/IIIa (GPIIb/IIIa) receptor, the activated conformation of which is responsible for platelet linkage via fibrinogen, revealed that platelet accumulation occurs in the microvasculature of patients with CM (6). The mechanisms of platelet aggregation and adhesion are not completely understood. However, the local production of the proinflammatory cytokine TNF may be a contributing factor (7, 8). TNF is elevated in postmortem brain tissue of CM patients and in murine models of CM, where it is known to upregulate the expression of the selectins on the cerebral vasculature (9). Other proinflammatory cytokines, such as IL-1β and lymphotoxin-α (LT-α), are also expressed within the brain parenchyma of mice and humans with CM and may contribute to the sequestration of platelets by activating the microvasculature (10). However, each of these cytokines gives rise to the recruitment of distinct populations of leukocytes across the intact brain endothelium despite the induction of the same pattern of adhesion molecule expression. The differential induction of chemokines is likely to determine which populations are recruited, but it is not known whether platelet adhesion to the brain microvasculature is dependent on the expression of specific cytokines.

A noninvasive approach for the specific detection of activated platelets or platelet thrombi in the cerebral microvasculature under different conditions of cytokine expression could help in determining the influence of cytokines on vascular platelet adhesion in CM and other neuropathologies. Recent progress in MRI has enabled the detection of such molecular targets by designing contrast agents that bind to cellular receptors or surface antigens (11, 12). By delivering high payloads of contrast agent such as iron oxide particles to molecular epitopes, imaging of even sparsely distributed molecules is possible (13, 14). We have previously demonstrated that vascular inflammation, such as is found in multiple sclerosis, can be detected in vivo using VCAM-1–targeted microparticles of iron oxide (MPIOs) in conjunction with MRI (15). In addition, we have described the unique possibility of selectively targeting activated platelets using single-chain antibodies that recognize the ligand binding pocket or ligand-induced binding sites (LIBS) of GPIIb/IIIa (16–18). The latter epitope becomes exposed only upon activation through receptor-ligand binding and therefore offers the opportunity to target activated platelets, such as are found on damaged endothelium caused by inflammation or atherosclerotic plaque rupture. By conjugating the LIBS antibody to MPIOs (referred to herein as LIBS-MPIO), we recently demonstrated the possibility of detecting endovascular platelet adhesion by ex vivo MRI (19).

In the present study, we used the above-described activation-specific platelet contrast agent LIBS-MPIO in an animal model to detect vascular platelet aggregation associated with CM before pathology was visible by conventional in vivo MRI. We further demonstrated that platelet accumulation was induced in the brain microvasculature by the proinflammatory cytokine TNF, but not by either IL-1β or LT-α. Following platelet accumulation, TNF, but not IL-1β, also induced the adherence of mononuclear cells to the cerebral vasculature in a manner very reminiscent of CM pathology.

Imaging of CM with conventional MRI. Conventional MRI was performed throughout the development of disease in mice infected with the CM parasite. No abnormalities were detected until day 7, when breakdown of the blood-brain barrier was evident as hyperintense areas on T₁-weighted images obtained after injection of gadopentetic acid (Gd-DTPA) that were not present prior to Gd-DTPA injection (Figure 1, A and B). Discrete hyperintensities were also present on T₂-weighted images (Figure 1C), which coincided with the Gd-DTPA–enhancing lesions. The evaluation of T₂ maps revealed a significant increase in T₂ within the hippocampus (Figure 1D). However, by day 7 the mice were moribund, and MRI at this time provided little additional information on the pathogenic process. Before day 7, no overt clinical signs were evident. Thus our studies using conventional MRI techniques failed to reveal the presence of CNS pathology before the appearance of overt clinical signs.

Figure 1

Conventional imaging of murine CM. (A) Coronal T₁-weighted image acquired 7 days after the injection of 10⁶P. berghei ANKA–parasitized rbcs. (B) T₁-weighted image from the same animal following injection of Gd-DTPA. Note the regions of increased signal intensity compared with A, indicating blood-brain barrier breakdown (arrow). (C) T₂-weighted image of the same slice showing increased signal intensity in the same regions as the Gd-DTPA enhancement. (D) Significant elevation in T₂ in the hippocampi of day-7 terminal-stage CM mice (P < 0.05).

Imaging of CM using a platelet-specific contrast agent. Murine and human CM is associated with the adherence of platelets to an intact brain endothelium (20, 21). The aim of this study was to determine whether we could distinguish between CM and control mice at a time when no overt disease was present using a contrast agent that recognizes activated platelets. Our in vitro experiments revealed that the functionalized MPIOs, which recognize the LIBS of GPIIb/IIIa receptors (LIBS-MPIO), bound to ADP-activated platelets alone. No significant binding was observed with MPIOs conjugated to control antibodies (control-MPIO) or nonfunctionalized MPIOs (Figure 2).

Figure 2

In vitro MPIO platelet binding. (A–C) Nonactivated or ADP-activated platelets were incubated with LIBS-MPIO (A), control-MPIO (B), or nonfunctionalized MPIOs (C). Note the presence of LIBS-MPIO binding on the surface of the activated platelets on the thrombus, which was not observed on nonactivated platelets or with the control MPIO or nonfunctionalized MPIO.

When MRI was performed at day 5 after P. berghei ANKA infection, some contrast enhancement, in the form of isolated focal hypointensities, was observed on T₂-weighted images after LIBS-MPIO injection (Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI33314DS1). However, after injection of LIBS-MPIO in CM-infected mice at day 6, more extensive MPIO-associated hypointensity was evident in and around cortical vessels, delineating areas of LIBS-MPIO binding as demonstrated in 2 representative slices at different levels of the same brain (Figure 3A, arrows). Conversely, CM-infected animals injected with control-MPIO exhibited no negative contrast in corresponding areas (Figure 3B). Using a 3D reconstruction of the original MRI data stack, we observed enhanced binding of LIBS-MPIO in cortical regions of the brain (Figure 3C and Supplemental Movie 1), while injection of control-MPIO did not give rise to specific binding (Figure 3D and Supplemental Movie 2). Using volumetric quantification, a significant increase in the extent of signal voids per volume was confirmed for LIBS-MPIO– compared with control-MPIO–injected animals (2,261 ± 623 vs. 282 ± 101 μm³ × 10^–6; P = 0.035; Figure 3E). Furthermore, histological evaluation of LIBS-injected CM animals revealed the presence of MPIO bound to aggregated endovascular platelets (see below). These data confirm that there is significant binding of LIBS-MPIO to areas of CM pathology and demonstrate that the use of this targeted contrast agent enables detection of pathology before the onset of overt clinical signs.

Figure 3

Data from animals with CM and LIBS-MPIO contrast agent injection. (A and B) T₂-weighted 3D gradient-echo images from CM mice following intravenous injection of LIBS-MPIO (A) and control-MPIO (B) are presented in 2 representative slices at 2 different levels within the same brain. Areas of MPIO-induced signal appeared as dark signal voids (arrows) in cortical regions of the LIBS-MPIO–injected mouse, but not in the control-MPIO–injected mouse. (C and D) 3D reconstruction confirmed the cortical binding pattern in the LIBS-MPIO–injected mouse (C), whereas only modest background binding was evident in the control-MPIO animal (D). (E) Quantification of signal voids demonstrated a significant difference between the LIBS-MPIO– and control-MPIO–injected animals.

In vivo MRI for platelet detection after TNF and IL-1β injection. To further investigate the mechanisms underlying cerebrovascular platelet aggregation in CM, we used MRI in conjunction with the LIBS-MPIO contrast agent to examine the spatial distribution of platelet aggregation in vivo following stereotactic injection of either TNF or IL-1β into the brains of normal mice.

In the TNF-injected animals using the LIBS-MPIO contrast agent, negative MRI contrast was observed bilaterally throughout the anterior portion of the forebrain (Figure 4A). Conversely, animals injected intracerebrally with IL-1β (Figure 4B) or saline (Figure 4C) showed no areas of MPIO binding (negative contrast) following injection of LIBS-MPIO. Nonspecific binding of LIBS-MPIO was excluded using control-MPIO in TNF-injected mice (Figure 4D). As with mice injected with IL-1β or saline, these animals showed no alteration in signal intensity.

Figure 4

In vivo T₂-weighted coronal images (4 images per brain, beginning at Bregma and moving backward in 700-μm increments) from 3D gradient-echo data sets each with approximately 90-μm isotropic resolution. (A) Animal injected intrastriatally with 1 μg TNF in 0.5 μl saline 11.5 h prior to intravenous injection of LIBS-MPIO (~4.5 mg/kg Fe). Intense low-signal areas (black) reflect the specific retention of MPIO on activated platelets adhering to the cerebrovascular endothelium. (B and C) In contrast, no effects of the LIBS-MPIO agent were detected in animals injected 11.5 h previously with 1 μg IL-1β in 0.5 μl saline (B) or with 0.5 μl saline alone (C). (D) Similarly, no nonspecific effects of the control-MPIO contrast agent were detected in animals injected intrastriatally with 1 μg TNF in 0.5 μl saline.

Using a 3D reconstruction of the original MRI data stack, we found binding of LIBS-MPIO to be enhanced in both cortical and striatal regions of the brain following TNF injection into the brain parenchyma (Figure 5A and Supplemental Movie 3), while no specific binding was evident following control-MPIO injection (Figure 5B and Supplemental Movie 4). As expected, at 12 h after TNF injection a similar binding pattern was found in both hemispheres, indicating bilateral aggregation, with slightly enhanced signal intensity changes on the side of injection. Volumetric quantification confirmed significantly greater LIBS-MPIO binding in TNF-injected animals compared with both IL-1β– and saline-injected animals, as well as TNF-injected animals receiving the control-MPIO agent (Figure 5C). These in vivo data indicate a role for TNF in platelet aggregation in cerebral vessels.

Figure 5

3D data and MRI signal quantification of animals with intracerebral TNF injection. (A) LIBS-MPIO binding pattern after intracerebral TNF-injection in a 3D reconstruction, showing the enhancement of cortical and central LIBS-MPIO binding. (B) Minimal background binding was observed in animals with control-MPIO injection. (C) Quantitative analysis showed significantly higher binding of LIBS-MPIO to TNF-injected brain areas compared with all controls.

By immunohistochemistry, MPIO binding was apparent in TNF-injected animals receiving the LIBS-MPIO (Figure 6). On cresyl-violet–stained paraffin-embedded sections, binding of MPIOs to areas near the vascular wall was confirmed (Figure 6A). Furthermore, costaining with CD41 (GPIIb subunit) confirmed that binding of LIBS-MPIO was specific to areas of wall-adherent platelets (Figure 6B). Attached to the red-stained thrombus area, 2 MPIOs in different focus (owing to their location in different focal planes through the section) were evident.

Figure 6

Histology of TNF-injected brains with LIBS-MPIO injection. (A) Cresyl-violet stain revealed binding of MPIO to areas on the vascular wall (arrows). (B) Binding of LIBS-MPIO to platelets or platelet thrombi was confirmed using immunohistochemistry for platelet-specific CD41: 2 MPIOs appearing in different focus and therefore of different shape were recognized at areas of platelet aggregation (arrows). (C) Similarly, MPIOs were detected on platelets and platelet thrombi in animals with CM (inset, high-power view of the region indicated by arrow). (D) Platelet-positive elements per injected hemisphere at 6, 12, and 24 h after intracerebral injection of TNF (diamonds) or saline (squares) in mice. (E) Quantification of the LIBS-MPIO–induced signal void in animals 6, 12, and 24 h after intracerebral TNF injection. Note the correlation between the number of platelet-positive elements over time and the LIBS-MPIO–induced signal void (r² = 0.9503).

The number of platelets bound to the brain endothelium after the microinjection of TNF was highest at 12 h both in rats and in mice. To examine the relationship between the hypointensity volume and the number of platelets adherent in the vasculature, we compared the volume of MRI-detectable hypointensities with the number of adherent platelets determined immunohistochemically at 6, 12, and 24 h after microinjection of TNF in mice. Binding of the contrast agent to the microvasculature was maximal at 12 h as detected by MRI (Figure 6, D and E), and there was a correlation between the number of platelets present in the brain vasculature and the MPIO-dependent signal loss (r² = 0.9503). This suggests that increased platelet load over time is directly related to increased LIBS-MPIO–induced signal reduction. In an animal injected with LIBS-MPIO 24 h after intracerebral TNF injection and imaged at both 2 and 10 h after LIBS-MPIO injection, the initial signal from the LIBS-MPIO present on the first scan was almost entirely absent in the later scan (Supplemental Figure 2). This finding suggests degradation of the contrast agent over the intervening period.

Cytokines and platelet aggregation. To establish whether the early expression of cytokines within the brain parenchyma is responsible for platelet adhesion, TNF, LT-α, or IL-1β was injected directly into the brain parenchyma. These cytokines are all known to be expressed within the brain parenchyma in CM, but to our knowledge their role in platelet adherence was hitherto unknown. GPIIb-positive platelets became adherent to the luminal portion of the vasculature from 2 h after microinjection of TNF into the rat brain parenchyma (Figure 7, A and B). The number of platelet-positive elements peaked at 12 h, but was still significantly increased 24 h after injection of TNF. This effect was also observed in mouse brain parenchyma, where platelet adhesion at the 12-h time point reached the maximum compared with the 6- or 24-h time points (Figure 6D). In contrast, injection of IL-1β, LT-α, or vehicle did not increase the number of platelet-positive elements (Figure 7B). Thus it appears that platelet binding to the brain endothelium is cytokine specific and that TNF is likely the principal mediator of endothelial platelet binding in CM. Interestingly, leukocytes, identified by the leukocyte common antigen, were recruited to the brain parenchyma or became adherent to the luminal portion of the vasculature over the 24-h period following injection of TNF into the striatum, but after the appearance of platelets in the microvasculature (Figure 7C). The recruited leukocytes were principally ED-1–positive cells, and no neutrophils were observed (data not shown). The recruited ED-1–positive cells were distinguished as 2 separate populations: parenchymal ED-1–positive cells, which had diapedesed into the parenchyma (Figure 7D), and vessel-associated ED-1–positive cells, which were associated with the luminal portion of the brain vasculature and appeared to be unable to diapedese into the parenchyma (Figure 7E). The number of ED-1–positive cells detected in the brain parenchyma increased from 12 h after injection (P < 0.05) and peaked at 24 h (P < 0.05) compared with vehicle-injected controls. A larger number of vessel-associated ED-1–positive cells was observed after 6 h (P < 0.05), again increasing over time to peak at 24 h (P < 0.05), compared with vehicle-injected controls. Representative histology demonstrated the lack of cellular recruitment in the meninges (Figure 7F) and parenchyma (Figure 7G) of vehicle-injected controls and the presence of ED-1–positive cells in the parenchyma after TNF microinjection into the striatum (Figure 7H). The ED-1–positive cells adherent to the vessel lumen were visualized using double-labeling immunohistochemistry (Figure 7I): vessels were located using an antibody to GLUT-1, a marker whose expression has been established on both the luminal and abluminal portions of the brain vasculature (22), together with an antibody against ED-1. These findings were confirmed using immunofluorescence and laser-scanning confocal microscopy (Figure 7J). Interestingly, the GLUT-1 confocal microscopy revealed that the platelets were binding to intact endothelium, which was unexpected.

Figure 7

Platelet and leukocyte accumulation in the brain following the intrastriatal injection of TNF. (A) Top: Platelets localized within blood vessel in the brain close to the injection site (brown, DAB), visualized by cresyl-violet counterstain. Bottom: Confocal triple-labeling immunocytochemistry identified platelets (red) and ED-1–positive cells (green) inside a vessel with an intact endothelium, as revealed by GLUT1 (blue) staining. (B) Number of GPIIb/IIIa-positive elements in the injected hemisphere following microinjection into the striatum of 1 μg TNF (diamonds), 1 ng IL-1 (triangles), or saline (squares) in a volume of 1 μl for rats or 0.5 μl for mice. (C–E) Total brain leukocytes, as identified by leukocyte common antigen (LCA; C), ED-1–positive recruited monocytes and activated microglial cells present in the brain parenchyma (D), and ED-1–positive recruited monocytes associated with the luminal portion of the brain vasculature (E), detected by immunohistochemistry over 24 h following intrastriatal injection of TNF (squares) or saline vehicle (diamonds). (F–H) Representative photographs showing the absence of leukocytes in the meninges (F) or parenchyma (G) and their presence in the parenchyma following injection of TNF (H) after immunohistochemical labeling for leukocyte common antigen. (I–L) Double-labeling immunocytochemistry (I and K) and confocal imaging (J and L) of ED-1–positive cells and the brain vasculature (as identified by GLUT-1) highlight the differences in mononuclear cell recruitment patterns in the parenchyma (I and J) and meninges (K and L). Note in the meninges the presence of large numbers of ED-1–positive cells (blue, VIP; red, FITC) that appear to have free passage from the vasculature (brown, DAB; blue, AF-636) compared with the same cells in the parenchyma that appear vessel associated. (K, inset) High-power view of extravascular ED-1–positive cells. ELISA results are expressed as pg MCP-1/mg total protein ± SEM. Cell numbers in brain are expressed per mm² ± SEM. *P < 0.05 versus vehicle control. Scale bars: 20 μm (A and F–J); 40 μm (K and L); 10 μm (K, inset).

In the meninges, recruited ED-1–positive mononuclear cells were observed as early as 2 h after intrastriatal TNF injection and increased dramatically over the 24-h period (data not shown). The movement of ED-1–positive mononuclear cells across the meningeal vasculature appeared unrestricted, as large numbers of ED-1–positive cells were observed abluminally (Figure 7K). These findings were supported by immunofluorescence and confocal microscopy (Figure 7L). Although not observed in the brain parenchyma, neutrophils were found in the meninges. These findings were also true of the choroid plexus, where a similar acute inflammatory response was observed after 2 h and increased to 24 h (data not shown).

We report the application of a functional activation-specific platelet contrast agent for the detection of vascular platelet adhesion in vivo. The use of this targeted contrast agent, which we believe to be novel, has contributed significantly to our understanding of the mechanism of platelet activation in the pathology of CM. The findings in this study are multiple: (a) Using the activation-specific LIBS-MPIO contrast agent, diagnosis of the life-threatening pathology of CM involving platelet aggregation was possible at an earlier stage than either clinical findings or conventional MRI allow. The functional approach used here enables imaging of activated platelets and platelet thrombi that cannot be detected by routine MRI sequences. (b) Platelet binding to the brain endothelium was cytokine specific, and TNF is likely to be the principal mediator in CM. (c) After the appearance of platelets in the microvasculature, TNF, but not IL-1β, induced the adherence of mononuclear cells to the cerebral vasculature in a manner very reminiscent of CM pathology. (d) These results demonstrate that the molecular imaging strategy applying LIBS-MPIO may be used for the detection of pathologies involving platelets, constituting a noninvasive in vivo examination with high sensitivity and specificity.

Imaging cerebrovascular platelet aggregation with LIBS-MPIO. Conjugation of superparamagnetic particles to antibodies or peptidomimetic peptides has previously been used to selectively image cellular receptors expressed in various disease conditions (11, 23, 24). Gadolinium-based contrast agents shorten the T₁ relaxation time of protons and provide positive contrast on T₁-weighted images (25). Sipkins et al. have previously imaged ICAM-1 expression in rat brain ex vivo using paramagnetic liposomes (26), and we have targeted E-selectin expression in rat brain in vivo using a Sialyl Lewis^x mimetic moiety conjugated to Gd-DTPA (9). The major limitation of gadolinium-based agents is the small quantity of gadolinium that can be delivered to an endothelial monolayer, which limits the resulting contrast effects. Shapiro et al. recently described the use of MPIOs for cellular imaging and tracking, which provide substantially greater contrast than do gadolinium-based agents (14). The detection of MPIOs in single cells in animal-based in vivo models has been described by the same group recently (27). Therefore, the use of micron-sized particles for the detection of sparse and difficult-to-access functional epitopes was highly attractive, and the excellent contrast properties improved the feasibility of this approach. We have recently described a similar method for detecting VCAM-1 expression on the vascular endothelium in vivo early in CNS inflammation, using an anti–VCAM-1 antibody conjugated to MPIO (15).

An important advantage of the current study is the use of a single-chain antibody in the LIBS-MPIO contrast agent, thus enabling binding to function-specific targets such as LIBS on activated GPIIb/IIIa. Single-chain antibodies are also less immunogenic than IgG-sized proteins because of the lack of the Fc-regions and, in spite of small protein size, allow the attachment of micron-sized iron oxide particles to target epitopes, as shown in the present study.

Differential binding patterns of the LIBS-MPIO contrast agent were evident in each experimental model. In TNF-injected animals, agent binding was observed in both hemispheres encompassing both cortical and striatal regions, although more contrast was evident in the injected (left) hemisphere. We have previously shown widespread (cortical and striatal) effects of TNF following a single intrastriatal bolus injection into rat brain (28). Immunohistochemically, adhesion molecule expression and platelet adhesion was found to be bilateral at the time point used in the current study, although — in accord with the MRI findings — more staining was present in the injected hemisphere. Conversely, in CM animals cortical binding was more pronounced than subcortical contrast changes. Maximum binding of LIBS-MPIO to platelets in the microvasculature was highest at 12 h, as detected by MRI, correlating to the number of platelets present (r² = 0.9503). This suggests that increased platelet load over time is directly reflected by increased LIBS-MPIO–induced signal void.

The concept of functional GPIIb/IIIa targeting opens up the possibility of detecting and imaging in vivo platelet aggregation, which has only been observed in postmortem tissue from the brains of patients with CM, multiple sclerosis, HIV-dementia, and bacterial meningitis. We have previously described the application of this contrast agent for the detection of activated platelets in a mouse model of platelet aggregation on mechanically injured endothelium using ex vivo MRI (19), and the potential applications of this or similar contrast agents are far reaching. In particular, the use of such an agent for monitoring antiplatelet therapy may have considerable utility across a number of diseases, including stroke, thrombosis, and cardiovascular disease, as well as in CM as described here.

Another important property for purposes of therapy monitoring is the biodistribution of the particles as well as the duration of contrast agent binding to the targeted receptors. In animals imaged for the first time 24 h after intracerebral TNF injection, the signal from the LIBS-MPIO in the brain detected 2 h after agent injection was almost entirely lost 10 h later. This suggests that the contrast agent is degraded and is not sequestered in the brain, which is an important prerequisite for serial imaging. Concerning the biodistribution of MPIO after injection, we have recently shown that MPIOs were identified in lung, kidney, liver, and spleen after 30 min with redistribution to the spleen and, to a lesser extent, the liver after 24 h (29). Importantly, there was no evidence of infarction, hemorrhage, edema, inflammation, or MPIO extravasation in any of these organs at either time point. The liver and spleen signals decayed over a week-long period, but the route of excretion is unclear (data not shown).

Cytokines and platelet aggregation. A broad spectrum of proinflammatory agents rapidly upregulates adhesion molecules on cerebral endothelium, in particular ICAM-1 and the selectins, in a time course comparable to that demonstrated on non-CNS endothelium. Notably, leukocyte recruitment is often negligible (30), and little is known about adhesion of platelets in CM models. IL-1β and TNF both upregulate ICAM-1 (31, 32) and selectins (33, 34) on the brain endothelium, and therefore, the differential binding of platelets to the brain endothelium following the microinjection of these cytokines into the brain was surprising. However, non–adhesion molecule–related vascular events, such as cytokine-induced volume changes, may play a role. We have previously shown that following intrastriatal injection of TNF there is a decrease in regional cerebral blood volume (rCBV) (28), which is in contrast to the increased rCBV induced by IL-1β (35). TNF-induced reduction in rCBV may promote the adhesion of platelets and subsequent recruitment of monocytes. It is of interest to note that monocyte recruitment follows platelet adhesion and suggests that the adhesion of platelets provides a scaffold for the subsequent recruitment of monocytes, which happens at a time when regional blood flow has returned to normal (28).

Limitations of LIBS-MPIO. In this work, commercially available MPIOs were used. The polystyrene coating of these specific particles may not be suitable for use in humans owing to their potential toxicity, although no adverse events were observed in any of the mice injected with MPIO-based contrast agents. Due to the histidine-tag conjugation technique, the bond between antibody and MPIO remained stable for approximately 4 h; after this time, a significant reduction in the efficacy of the agent was observed (data not shown). Research on more stable preparations is ongoing.

Conclusion. This study provides proof of concept that in experimental models of human neurological diseases, targeted contrast agents can be used to detect pathology earlier than conventional, clinically used MRI approaches or clinical assessment. Wall-adherent platelets were detected noninvasively in vivo in a model of CM, prior to the onset of clinical symptoms, using a functional MRI contrast agent that targeted LIBS on activated GPIIb/IIIa receptors. In contrast, conventional MRI techniques failed to reveal the presence of CNS pathology before the appearance of overt clinical signs. Owing to the high specificity and sensitivity of our contrast agent, we were able to identify TNF as the mediator responsible for platelet aggregation. Furthermore, we were able to demonstrate that due to disappearance of the contrast agent signal over time, serial imaging seems possible. These results demonstrate the potential of such agents in enabling the noninvasive assessment of pathological mechanisms in disease models for diagnostic purposes, mechanistic studies, and monitoring of therapy.

View Supplemental data

We thank Andrew Lowe, Nicole Bassler, and Irene Neudorfer for technical assistance and Tom Bannister for his help with the data quantification. This work was funded by grants from the Medical Research Council (to N.R. Sibson and D.C. Anthony), the Wellcome Trust (to R.P. Choudhury), and the NHMRC and National Heart Foundation of Australia (to K. Peter).

Address correspondence to: Daniel C. Anthony, Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, United Kingdom. Phone: 44-1865-281136; Fax: 44-1865-271853; E-mail: daniel.anthony@pharm.ox.ac.uk.

Nonstandard abbreviations used: CM, cerebral malaria; Gd-DTPA, gadopentetic acid; GPIIb/IIIa, glycoprotein IIb/IIIa; LIBS, ligand induced binding site(s); LT-α, lymphotoxin-α; MPIO, microparticle of iron oxide; TE, echo time; TR, repetition time.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.118:1198–1207 (2008). doi:10.1172/JCI33314

Constantin von zur Muhlen, Nicola R. Sibson, Karlheinz Peter, and Sandra J. Campbell contributed equally to this work.

1. Marsh, K., Snow, R.W. 1999. Malaria transmission and morbidity. Parassitologia. 41:241-246.
   View this article via: PubMed Google Scholar
2. Wassmer, S.C., Combes, V., Candal, F.J., Juhan-Vague, I., Grau, G.E. 2006. Platelets potentiate brain endothelial alterations induced by Plasmodium falciparum. Infect. Immun. 74:645-653.
   View this article via: CrossRef PubMed Google Scholar
3. Grau, G.E., et al. 1993. TNF-induced microvascular pathology: active role for platelets and importance of the LFA-1/ICAM-1 interaction. Eur. Cytokine Netw. 4:415-419.
   View this article via: PubMed Google Scholar
4. Grau, G.E., Lou, J. 1993. TNF in vascular pathology: the importance of platelet-endothelium interactions. Res. Immunol. 144:355-363.
   View this article via: CrossRef PubMed Google Scholar
5. Combes, V., et al. 2004. Pathogenic role of P-selectin in experimental cerebral malaria: importance of the endothelial compartment. Am. J. Pathol. 164:781-786.
   View this article via: PubMed Google Scholar
6. Grau, G.E., et al. 2003. Platelet accumulation in brain microvessels in fatal pediatric cerebral malaria. J. Infect. Dis. 187:461-466.
   View this article via: CrossRef PubMed Google Scholar
7. Lou, J., et al. 1997. Platelets play an important role in TNF-induced microvascular endothelial cell pathology. Am. J. Pathol. 151:1397-1405.
   View this article via: PubMed Google Scholar
8. Grau, G.E., Lou, J.N. 1995. Experimental cerebral malaria: possible new mechanisms in the TNF-induced microvascular pathology. Soz. Praventivmed. 40:50-57.
   View this article via: CrossRef PubMed Google Scholar
9. Sibson, N.R., et al. 2004. MRI detection of early endothelial activation in brain inflammation. Magn. Reson. Med. 51:248-252.
   View this article via: CrossRef PubMed Google Scholar
10. Hunt, N.H., Grau, G.E. 2003. Cytokines: accelerators and brakes in the pathogenesis of cerebral malaria. Trends Immunol. 24:491-499.
    View this article via: CrossRef PubMed Google Scholar
11. Spuentrup, E., et al. 2005. Molecular magnetic resonance imaging of coronary thrombosis and pulmonary emboli with a novel fibrin-targeted contrast agent. Circulation. 111:1377-1382.
    View this article via: CrossRef PubMed Google Scholar
12. Nahrendorf, M., et al. 2006. Noninvasive vascular cell adhesion molecule-1 imaging identifies inflammatory activation of cells in atherosclerosis. Circulation. 114:1504-1511.
    View this article via: CrossRef PubMed Google Scholar
13. Shapiro, E.M., Skrtic, S., Koretsky, A.P. 2005. Sizing it up: cellular MRI using micron-sized iron oxide particles. Magn. Reson. Med. 53:329-338.
    View this article via: CrossRef PubMed Google Scholar
14. Shapiro, E.M., et al. 2004. MRI detection of single particles for cellular imaging. Proc. Natl. Acad. Sci. U. S. A. 101:10901-10906.
    View this article via: CrossRef PubMed Google Scholar
15. McAteer, M.A., et al. 2007. In vivo magnetic resonance imaging of acute brain inflammation using microparticles of iron oxide. Nat. Med. 13:1253-1258.
    View this article via: CrossRef PubMed Google Scholar
16. Schwarz, M., et al. 2006. Conformation-specific blockade of the integrin GPIIb/IIIa: a novel antiplatelet strategy that selectively targets activated platelets. Circ. Res. 99:25-33.
    View this article via: CrossRef PubMed Google Scholar
17. Schwarz, M., et al. 2004. Single-chain antibodies for the conformation-specific blockade of activated platelet integrin alphaIIbbeta3 designed by subtractive selection from naive human phage libraries. FASEB J. 18:1704-1706.
    View this article via: PubMed Google Scholar
18. Stoll, P., et al. 2007. Targeting ligand-induced binding sites on GPIIb/IIIa via single-chain antibody allows effective anticoagulation without bleeding time prolongation. Arterioscler. Thromb. Vasc. Biol. 27:1206-1212.
    View this article via: CrossRef PubMed Google Scholar
19. von zur Mühlen, C., et al. 2007. Magnetic resonance imaging of platelets on wire-injured mouse femoral arteries using activation-specific anti-GP IIb/IIIa single chain antibodies conjugated to microparticles of iron oxide [abstract]. J. Am. Coll. Cardiol. 49(Suppl.):108A.
20. Sun, G., et al. 2003. Inhibition of platelet adherence to brain microvasculature protects against severe Plasmodium berghei malaria. Infect. Immun. 71:6553-6561.
    View this article via: CrossRef PubMed Google Scholar
21. van der Heyde, H.C., Nolan, J., Combes, V., Gramaglia, I., Grau, G.E. 2006. A unified hypothesis for the genesis of cerebral malaria: sequestration, inflammation and hemostasis leading to microcirculatory dysfunction. Trends Parasitol. 22:503-508.
    View this article via: CrossRef PubMed Google Scholar
22. Guerin, C., et al. 1990. The glucose transporter and blood-brain barrier of human brain tumors. Ann. Neurol. 28:758-765.
    View this article via: CrossRef PubMed Google Scholar
23. Winter, P.M., et al. 2003. Improved molecular imaging contrast agent for detection of human thrombus. Magn. Reson. Med. 50:411-416.
    View this article via: CrossRef PubMed Google Scholar
24. Lipinski, M.J., et al. 2006. MRI to detect atherosclerosis with gadolinium-containing immunomicelles targeting the macrophage scavenger receptor. Magn. Reson. Med. 56:601-610.
    View this article via: CrossRef PubMed Google Scholar
25. Runge, V.M., et al. 1985. Initial clinical evaluation of gadolinium DTPA for contrast-enhanced magnetic resonance imaging. Magn. Reson. Imaging. 3:27-35.
    View this article via: CrossRef PubMed Google Scholar
26. Sipkins, D.A., et al. 2000. ICAM-1 expression in autoimmune encephalitis visualized using magnetic resonance imaging. J. Neuroimmunol. 104:1-9.
    View this article via: CrossRef PubMed Google Scholar
27. Shapiro, E.M., Sharer, K., Skrtic, S., Koretsky, A.P. 2006. In vivo detection of single cells by MRI. Magn. Reson. Med. 55:242-249.
    View this article via: CrossRef PubMed Google Scholar
28. Sibson, N.R., et al. 2002. TNF-alpha reduces cerebral blood volume and disrupts tissue homeostasis via an endothelin- and TNFR2-dependent pathway. Brain. 125:2446-2459.
    View this article via: CrossRef PubMed Google Scholar
29. McAteer, M.A., et al. 2007. Magnetic resonance imaging of endothelial adhesion molecules in mouse atherosclerosis using dual-targeted microparticles of iron oxide. Arterioscler. Thromb. Vasc. Biol. 28:77-83.
    View this article via: CrossRef PubMed Google Scholar
30. Bell, M.D., Perry, V.H. 1995. Adhesion molecule expression on murine cerebral endothelium following the injection of a proinflammagen or during acute neuronal degeneration. J. Neurocytol. 24:695-710.
    View this article via: CrossRef PubMed Google Scholar
31. McHale, J.F., Harari, O.A., Marshall, D., Haskard, D.O. 1999. Vascular endothelial cell expression of ICAM-1 and VCAM-1 at the onset of eliciting contact hypersensitivity in mice: evidence for a dominant role of TNF-alpha. J. Immunol. 162:1648-1655.
    View this article via: PubMed Google Scholar
32. Proescholdt, M.G., et al. 2002. Intracerebroventricular but not intravenous interleukin-1beta induces widespread vascular-mediated leukocyte infiltration and immune signal mRNA expression followed by brain-wide glial activation. Neuroscience. 112:731-749.
    View this article via: CrossRef PubMed Google Scholar
33. Carvalho-Tavares, J., et al. 2000. A role for platelets and endothelial selectins in tumor necrosis factor-alpha-induced leukocyte recruitment in the brain microvasculature. Circ. Res. 87:1141-1148.
    View this article via: PubMed Google Scholar
34. Bernardes-Silva, M., Anthony, D.C., Issekutz, A.C., Perry, V.H. 2001. Recruitment of neutrophils across the blood-brain barrier: the role of E- and P-selectins. J. Cereb. Blood Flow Metab. 21:1115-1124.
    View this article via: PubMed Google Scholar
35. Blamire, A.M., et al. 2000. Interleukin-1beta -induced changes in blood-brain barrier permeability, apparent diffusion coefficient, and cerebral blood volume in the rat brain: a magnetic resonance study. J. Neurosci. 20:8153-8159.
    View this article via: PubMed Google Scholar
36. Schwarz, M., et al. 2004. Reversibility versus persistence of GPIIb/IIIa blocker-induced conformational change of GPIIb/IIIa (alphaIIbbeta3, CD41/CD61). J. Pharmacol. Exp. Ther. 308:1002-1011.
    View this article via: CrossRef PubMed Google Scholar
37. Wilcockson, D.C., Campbell, S.J., Anthony, D.C., Perry, V.H. 2002. The systemic and local acute phase response following acute brain injury. J. Cereb. Blood Flow Metab. 22:318-326.
    View this article via: PubMed Google Scholar
38. Blond, D., et al. 2002. Differential induction of interleukin-1beta and tumour necrosis factor-alpha may account for specific patterns of leukocyte recruitment in the brain. Brain Res. 958:89-99.
    View this article via: CrossRef PubMed Google Scholar
39. Andersson, P.B., Perry, V.H., Gordon, S. 1992. The acute inflammatory response to lipopolysaccharide in CNS parenchyma differs from that in other body tissues. Neuroscience. 48:169-186.
    View this article via: CrossRef PubMed Google Scholar
40. Matyszak, M.K., Perry, V.H. 1995. Demyelination in the central nervous system following a delayed-type hypersensitivity response to bacillus Calmette-Guerin. Neuroscience. 64:967-977.
    View this article via: CrossRef PubMed Google Scholar
41. Anthony, D., et al. 1998. CXC chemokines generate age-related increases in neutrophil-mediated brain inflammation and blood-brain barrier breakdown. Curr. Biol. 8:923-926.