High‐resolution in vivo imaging of microglia using a versatile nongenetically encoded marker

Microglial cells are the innate immune cells of the CNS, whose main role is to monitor the integrity of and to react to any disturbances of brain homeostasis. As such, microglial cells are involved in a large number of CNS insults (e.g. acute CNS injury, brain tumors, apoptosis, infection, ischemia, neurodegenerative diseases) and their engagement can be either neurotoxic or neuroprotective 1, 2. Despite their critical role in ameliorating or exacerbating disease progression, little is known about the in vivo functional properties of these cells. So far high resolution in vivo studies of microglial function were conducted in mice with genetically labeled microglia 3, 4. These analyses revealed the surveillance function of microglia in the healthy brain 5, 6, and its involvement in the remodeling of synaptic circuits during ischemia and sensory deprivation 7. However, because of the lack of fractalkine receptor (CX3CR1 mice) or low expression levels of GFP (Iba1-GFP mice), these mouse lines are less suitable for studying the role of microglia under pathological conditions 2, 8-10. Therefore, there is a need for a nongenetically encoded, easy to use microglial marker, enabling high-quality staining of microglia (similar to the quality obtained in GFP-expressing mice), but applicable to any mouse strain at any experimental age. Here we utilize a well-known histological marker tomato lectin (from Lycopersicon esculentum 11-13) for high resolution in vivo imaging of microglia. A brief pressure injection of tomato lectin conjugated with a fluorescent dye DyLight® 594 (TL) into the mouse cortex resulted in robust staining of microglial cells (Fig. 1A) and blood vessels (arrowhead). The latter, however, were easily distinguished from microglia based on their morphological appearance. A single injection of TL labeled a roughly spherical tissue volume with a diameter of 80–130 μm. Multiple, equally spaced injections from the same pipette were used to increase the dimensions of the labeled area. To test the reliability of the in vivo staining protocol, we applied it to the CX3CR1GFP/+ mice previously described 3. A total of 100% of all GFP-positive cells within the stained area were also TL positive (n = 103 cells in four mice; Fig. 1A). Noteworthy, this staining protocol enabled visualization of thin microglial processes at a high spatial resolution (Fig. 1B). In vivo, such image quality was so far only attainable when using GFP-expressing cells. Comparable image quality was obtained when using fluorescein-conjugated tomato lectin (Fig. 1C). It was possible to image the same cells repetitively over 5–6 h (n = 17 cells in two mice; Supporting Information Fig. 1). With time, however, more TL-labeled material accumulated inside the cells (arrow; probably due to the turnover of the cell membrane), making the identification of cell somata more difficult. Next, we tested whether the TL positive in vivo-labeled cells can be reidentified in the fixed tissue processed for immunohistochemistry. The quality of TL-based labeling was well preserved in the fixed tissue (Fig. 1D). A total of 100% of TL-positive cells was also stained with antibodies recognizing a well-known microglial marker Iba1 (n = 53 cells in two mice). As shown in Fig. 1D (arrows), TL labeled the entire cell down to the very tips of microglial processes. The protocol used also allowed to label microglia in aged mice (Fig. 2A). In contrast to previously published techniques (14, Supporting Information Fig. 2), it also enabled visualization of microglia in a mouse model of Alzheimer's disease. In the latter mice, TL labeled the cells located both in the vicinity of amyloid plaques and in plaque-free areas (Fig. 2B and inset; Supporting Information Fig. 3; n = 127 cells in three mice). To test whether TL can be combined with other cell type specific and/or activity markers, we labeled neurons and astrocytes with a calcium-sensitive dye Oregon Green BAPTA-1 AM (OGB-1) as shown in Figure 2C using a previously described multicell bolus loading technique (MCBL) 15. In addition, sulforhodamine 101 (SR101, 16) and TL were used to label astrocytes and microglia, respectively (see also Supporting Information Fig. 4). In stained preparations, neurons appeared green, astrocytes yellow, and microglia red, because OGB-1, when applied using MCBL, is not taken up by microglia 14. It was recently reported, however, that SR101 is a highly bioactive substance inducing long-lasting changes in the efficacy of synaptic transmission 17, 18. Therefore, we tested whether TL-based labeling influences activity of surrounding neurons. Layer 2/3 cortical neurons were stained with OGB-1 using MCBL. Amplitudes and frequencies of spontaneous Ca2+ transients in these cells 19 were monitored before and after TL injection (Supporting Information Fig. 5). There was no significant change in either of these parameters upon injection of TL. Similar to GFP-expressing microglia 6, TL-labeled cells in WT mice rapidly moved their processes toward an ATP-containing pipette (Fig. 2D and E). Neither the average extension rate of the processes, nor the final diameter of the spherical containment surrounding the pipette differed significantly between GFP-expressing and TL-labeled cells (Fig. 2F and G; p = 0.26 and 0.07, respectively, Mann–Whitney test). After loading with OGB-1, TL-labeled microglial cells readily responded with Ca2+ transients to ATP and to the agonists of ionotropic/metabotropic P2 receptors Bz-ATP/UDP (Fig. 2H and I). There was no significant difference between the normalized areas under the Ca2+ transients (see Supporting Information Materials and methods for details) recorded from GFP-expressing and TL-labeled microglial cells (Fig. 2J). In conclusion, our results establish dye-conjugated tomato lectin as a versatile microglial marker for in vivo studies. Using the protocol described in this study, microglia can be visualized in juvenile, adult, and aged brain tissue. To our knowledge, this is the only nongenetically encoded marker enabling imaging of microglia in mouse models of Alzheimer's disease. The labeling procedure is fast and can be easily combined with other experiments, as for example, imaging of intracellular Ca2+ dynamics. This approach will facilitate in vivo analyses of microglial function in general, as well as the understanding of its specific role in aging and neurodegenerative diseases. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (GA 654/1-1) to O.G. The authors declare no financial or commercial conflict of interest. Disclaimer: Supplementary materials have been peer-reviewed but not copyedited. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.