Simultaneous fluorescence imaging of distinct nerve and blood vessel patterns in dual Thy1-YFP and Flt1-DsRed transgenic mice

Abstract

Blood vessels and nerve tissues are critical to the development and functionality of many vital organs. However, little is currently known about their interdependency during development and after injury. In this study, dual fluorescence transgenic reporter mice were utilized to observe blood vessels and nervous tissues in organs postnatally. Thy1-YFP and Flt1-DsRed (TYFD) mice were interbred to achieve dual fluorescence in the offspring, with Thy1-YFP yellow fluorescence expressed primarily in nerves, and Flt1-DsRed fluorescence expressed selectively in blood vessels. Using this dual fluorescent mouse strain, we were able to visualize the networks of nervous and vascular tissue simultaneously in various organ systems both in the physiological state and after injury. Using ex vivo high-resolution imaging in this dual fluorescent strain, we characterized the organizational patterns of both nervous and vascular systems in a diverse set of organs and tissues. In the cornea, we also observed the dynamic patterns of nerve and blood vessel networks following epithelial debridement injury. These findings highlight the versatility of this dual fluorescent strain for characterizing the relationship between nerve and blood vessel growth and organization.

Introduction

The vascular and nervous systems are complex networks that frequently course together to provide sensation and nutrition to vital tissues throughout the body. Anatomically, nerves and vessels travel together within the neurovascular bundle to supply many metabolic, highly active target tissues. On a more microscopic scale, the two systems are intimately interconnected, as the blood vessels composing the vasa nervorum are essential for the metabolic upkeep of peripheral nerve bundles, while autonomic nerves regulate circulation through their effect on the constriction and dilation of blood vessels [1, 2]. Although the exact factors and mechanisms regulating the interrelationships between these two networks remain poorly understood, recent studies indicate that the two systems may share more similarities than previously assumed [3–5].

The two systems are intimately tied during development, as the processes of angiogenesis and neural development are controlled by complex and highly regulated mechanisms to ensure that both networks conform to the structural and functional demands of specific tissues. In addition to their anatomic overlap in many regions, the two systems share common genetic pathways and principles for growth and navigation of cells to their intended targets [6]. Andreone et al. proposed that the patterning of both neural and vascular systems is essential in meeting the metabolic requirements for proper functioning of the central nervous system (CNS) [7]. This occurs also in other organs and tissues, further establishing the importance of the interdependency of both the vascular and nervous systems, i.e., the ‘neurovascular congruency’ [4, 5, 8–10]. In addition, neuronal growth cones and endothelial tip cells are instructed by shared molecular guidance cues [4–6, 8, 9, 11]. These highly-regulated shared processes are particularly important in the development of the CNS [7]. In the eye alone, multiple links have been shown to exist between blood vessel and nerve development. Within the retina, Okabe et al. hypothesized using a murine retinal model that neuronal vascular endothelial growth factor receptor 2 (VEGFR2) may act as a scavenger for vascular endothelial growth factor A (VEGF-A), which is primarily secreted by astrocytes in the retina, creating a gradient reduction in VEGF-A for blood vessel growth [5]. In the cornea, it is thought that blood vessels and nerves may inhibit the growth of one another. Ablation of trigeminal nerves leads to increased neovascularization of the cornea, whereas increased neovascularization induced by implantation of a basic fibroblast growth factor (b-FGF) pellet correlates with the absence of normal innervation [12]. There is also experimental evidence that VEGF mediates nerve regeneration in the cornea. VEGF-A blocking antibodies, for example, reduce corneal angiogenesis but can also increase the likelihood of corneal nerve recession, presumably because VEGF-A has pro-neural functions in the eye that are inhibited [13]. Pan et al. showed in vitro that VEGF-A enhances neurite growth of isolated trigeminal neurons, a phenomenon suppressed by VEGFR1-, VEGFR2-, and Neuropilin receptor 1-neutralizing antibodies, as well as VEGFR2 inhibitors [14]. They also demonstrated increased nerve regeneration in superficially injured corneas after implantation of low-dose VEGF-infused pellets.

The investigation of neural and blood vessel interactions in postnatal tissues may shed light on many pathological conditions and may further elucidate the developmental mechanisms governing neural and vascular cooperation. Common insults like traumatic brain injury inflict damage to both the vascular and neural networks of the brain, both of which must be monitored over extended time periods to assess the extent of tissue damage and recovery after injury. Use of this animal model may provide further insight into the cooperative physiological development and the responses of the neural and blood vessel systems to injury. Furthermore, such studies may establish a platform for the development of therapeutic interventions for blood vessel and nervous system regeneration in pathological conditions.

Thymocyte antigen 1 (Thy1 or CD90), a glycosylphosphatidylinositol (GPI)-anchored, cell surface protein, is expressed in various cell types in mice, including neuronal and immune lineages [15]. Deletion of intron 3 produces selective but widespread expression in a variety of peripheral nervous system (PNS) and CNS neurons, especially for mature neurons [15]. Alic et al. found that at day E14.5 in Thy1-yellow fluorescent protein (YFP)-16 transgenic mice, 22% of isolated CNS neurons were positive for Thy1-YFP, but after 1 month of postnatal development, the percentage of Thy1-YFP-positive neurons in cortex tissue was 50%. Thy1-YFP expression was noted to increase within multiple CNS and PNS tissues over the course of development [16]. Other previous studies have suggested that Thy1 inhibits axonal outgrowth and thus may regulate and stabilize neural connectivity. As a result, Thy1 is expressed most abundantly in mature, but not developing, neurons [17–20]. Thy1 is an excellent promoter for expressing fluorescent reporter transgenes for the purpose of monitoring neuronal structure and development in postnatal neural tissues. Thy1-YFP mice have been used in multiple in vitro and in vivo studies for the visualization of both the CNS and PNS [16, 21–23].

In addition to enabling observation of neural development in vivo, Thy1-YFP neuronal visualization provides a new approach to investigating neurological disease and responses to injury. Beirowski et al. utilized in vivo and in vitro explant approaches to observe that YFP fragmentation in the peripheral nerves of Thy1-YFP-H mice corresponds to Wallerian degeneration of the axons [24]. Namavari et al. performed surgical transection of corneal nerves in Thy1-YFP mice and observed corneal nerve regeneration. The YFP fluorescence in vivo demonstrated that corneal neurons assume a different regeneration pattern following injury [23]. Within the CNS, YFP fluorescence allowed researchers to monitor the differentiation of Thy1-YFP neural stem cells after transplantation to stroke-affected areas [25]. In addition to these advances, studies using Thy1-YFP-16 in vivo imaging have identified members of the VEGF family as potential therapeutic agents for the treatment of peripheral nerve injury [14, 26].

These studies employing Thy1-YFP highlight the versatility of this gene as a reporter for visualizing changes in nerve patterns during development and repair after injury. In this study, we crossbred the Thy1creERT2-YFP strain with the Flt1-DsRed strain to generate a double transgenic reporter strain, Thy1creERT2-YFP/Flt1-DsRed. The Thy1creERT2-YFP transgene is designed with two separate copies of the modified mouse Thy1 promoter region (each containing the sequences required for neuronal-specific expression) driving expression of a creERT2 fusion protein and YFP [27, 28]. This mouse line is referred to as Thy1-YFP/Flt1-DsRed (TYFD) mouse in this paper.

In order to generate a fluorescent reporter for vascular cells, Flt1 was chosen as the promoter for DsRed fluorescence in endothelial cells. VEGFR1 is a protein encoded by the Flt1 gene, and activation of VEGFR1 is important in the processes of angiogenesis [29, 30] vasculogenesis, and hematopoiesis [31–33]. VEGFR1 is expressed in vascular endothelial cells, hematopoietic stem cells, monocytes, and macrophages [34, 35]. Previously, we have extensively used Flt1-DsRed mice to successfully visualize blood vessels in vivo with red fluorescence [36–38], demonstrating the versatility and robustness of Flt1-DsRed mice in marking vascular endothelial cells.

Recently there has been increasing interest in studying the autonomic nervous system, and in particular, one of its divisions—the enteric nervous system [39, 40]. However, limited models exist for visualizing nerves together with blood vessels in vivo in the CNS, PNS, and autonomic nervous system. One example is imaging the enteric nervous system of Wnt1-cre:tdTomato mice, in which a subset of neurons is labeled with a red fluorescent marker, together with transient injections of the green FITC-dextran to label the vasculature [41]. We propose that our model can be used to image a variety of nervous and vascular structures, and in the present study, we tested the utility of our dual transgenic TYFD mouse strain for visualizing both vascular and nervous networks in various tissues ex vivo without staining. These included CNS- and PNS-supplied tissues, such as various sites within the brain, retina, optic nerve, trigeminal and dorsal root ganglia, skeletal muscle and tail, as well as tissues that receive autonomic innervation including heart, lungs, thymus, liver, kidney, stomach, small intestine, cecum, colon, and bladder.

Finally, we also demonstrate an in vivo application of our dual transgenic TYFD mouse strain for monitoring the neuroregenerative and angiogenic response to injury in the cornea. The cornea, as the transparent and refractive tissue on the outer surface of the eye, is a valuable site for the study of angiogenesis and neural regeneration because corneal vessels and nerves can be readily imaged in vivo. Previously, we utilized a dual fluorescence animal model, Prox1-GFP/Flt1-DsRed (PGFD) and Prox1-GFP/Flk1-mCherry (PGFM) mice, to characterize the regeneration of corneal lymphatic and blood vessels [38, 42, 43]. Here, we propose that the TYFD strain can be used not only for the direct visualization of blood vessels and nerves but also as a platform for observing angiogenesis and corneal nerve regeneration. Time-lapse fluorescence microscopy can be used to observe how these two systems grow and organize together in specific organs, which may shed light on the currently unexplored interrelationship between vessel and nerve growth post injury in adulthood. Our model introduces a novel strategy for investigating the poorly understood link between two well-known processes: nerve regeneration and angiogenesis.

Methods

Animals

All procedures were performed as per the US National Institutes of Health Guide for the Care and Use of Laboratory Animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The Institutional Animal Care and Use Committee of University of Illinois at Chicago approved this study’s experimental protocols.

Organ imaging

Both male and female TYFD mice were used to harvest various tissues and organs. At least three animals were imaged for each tissue type and time point. The tissues were dissected immediately post sacrifice of the TYFD mice and imaged using the Zeiss LSM 710 Series confocal microscope (Zeiss Microscopy, Oberkochen, Germany). The following tissues were imaged at the postnatal day 20 (P20) time point: brain, retina, cornea (whole-mount), optic nerve, trigeminal ganglion, dorsal root ganglion, heart, lung, thymus, stomach, small intestine, cecum, colon, liver, kidney, thigh muscle, bladder, and tail.

Two-photon microscopy

Image acquisition was carried out using a Bruker Prairie Ultima two-photon microscope (Bruker, Billerica, MA) equipped with an Olympus XLUMPlanFL N 20x/1.00 W objective (Olympus, Waltham, MA). YFP and DsRed were simultaneously excited by a Coherent Chameleon Ultra II two-photon laser tuned to 1000 nm. Emission signals from fluorophores were collected at 530/50 nm (YFP) and 595/50 nm (DsRed). PrairieView software was used for image acquisition and Imaris (Bitplane, Version 7.2) for image reconstruction.

Corneal injury model

An Algerbrush was used for debridement of the superficial epithelium of the whole cornea with sparing the limbal area (3.10±0.3 mm diameter) in TYFD mice [36]. The procedure was done under general anesthesia with intra-peritoneal injection of a mixture of ketamine-xylazine and topical anesthesia with proparacaine eye drops. Eyes were rinsed with saline and given topical erythromycin ointment post debridement. The injured eyes were visualized under an Axio Zoom V16 fluorescence microscope (Zeiss Microscopy, Oberkochen, Germany) at baseline (before injury) and on days 0, 4, 7, 10, and 14 after injury.

Results

Imaging of neurovasculature in various tissues/organs of TYFD mice

To characterize the expression of Thy1-YFP–labeled nerves and to demonstrate the anatomical relationship between nerves and blood vessels in TYFD mice, we imaged a variety of neural and non-neural organs at the P20 time point. Our results support that Thy1-YFP is consistently and reliably expressed both in the neuron cell body and axons. All imaged vascularized organs exhibited Flt1-DsRed signal, highlighting the ubiquity of Flt1 expression in blood vessels.

The dorsal, ventral, and sagittal aspects of the brain were imaged via confocal microscopy (Figures 1 & 2). As the main organ of the central nervous system, both Thy1 and Flt1 are extensively expressed in the brain. In both the dorsal (Figure 1a) and ventral (Figure 1c) aspects, a vast network of blood vessels and nervous system were seen throughout the cerebral hemispheres, olfactory bulb, cerebellum, and brainstem. An enlarged view of the medulla showed the basilar artery with splitting vertebral arteries in the central area of the ventral brain, along with laterally oriented parallel strands of Thy1-expressing nerve fibers (Figure 1d). Another enlarged view emphasized the pituitary area, which was densely vascularized and surrounded by Thy1-expressing nerve fibers (Figure 1e). Imaging of a sagittal section of the brain also showed a vascular network extending across the brain, along with significant Thy1 expression throughout the whole brain, but interestingly weak expression within the cortex (Figure 2). Many of these Thy1-expressing axons traveled in parallel projections that marked their axonal tracts within the brain.

Figure 1.

Confocal imaging of the dorsal and ventral aspects of the TYFD mouse brain.

Figure 2.

Confocal imaging of the sagittal aspect of the TYFD mouse brain.

Imaging of a corneal whole-mount without injury revealed a network of nerves and vessels distributed within the limbal environment (Figure 3d). The blood vessels were limited to the limbal area, while nerves continued to invade the entirety of the cornea. Image stratification was performed to reveal the general morphology of corneal nerves. Epithelial nerves (Figure 3b) were fine, dense, and gave rise to the whorl-like pattern near the apex of the cornea; on the other hand, deeper stromal nerves (Figure 3c) were less dense but thicker in width.

Figure 3.

Fluorescent imaging of whole-mount TYFD mouse corneas.

Imaging of the optic nerve revealed a major trunk of the blood vessel running along the course of the optic nerve with several small branches from the sides. The optic nerve fibers ran in a parallel pattern until they entered the globe (Figure 4a).

Figure 4.

Confocal imaging of the optic nerve, trigeminal, and dorsal root ganglion of TYFD mice.

Trigeminal and dorsal root ganglion imaging exhibited groups of nerve cell bodies condensed intertwined with tightly packed vascular plexus (Figures 4b, 4c, & 5).

Figure 5.

Two-photon imaging of the trigeminal and dorsal root ganglion of TYFD mice.

As expected, the retina was comprised of a multilayered heterogeneous population of neurons that could not be appreciated fully with collapsed z-stack images. By stratifying the images, we could appreciate these layers and cell types of the retina and the blood vessel plexus, penetrating and supplying each layer of the retina. We observed that the nerves and the blood vessels have distinct patterns from each other (Figure 6).

Figure 6.

Confocal z-stack imaging of the TYFD mouse retina.

In the heart, we observed that the nerves and their branches started from the base and trailed along the axis of the heart, with dense DsRed-expressing populations surrounding them (Figure 7a).

Figure 7.

Confocal imaging of the TYFD mouse heart, lung, and thymus.

Some organs did not have Thy1-expressing populations or either showed weak or non-specific expression, such as the lung (Figure 7b), thymus (Figure 7c), liver (Figure 8a), and kidney (Figure 8b). On the other hand, the digestive system organs (Figure 9), such as the stomach (Figure 9a), small intestine (Figure 9b), cecum (Figure 9c and 9d), and colon (Figure 9e) exhibited strong expression of both blood vessels and nerves with very distinctive patterns. Thicker Thy1-expressing dots found in the small intestine, cecum, and colon, presumably represent the group of nerve cell bodies that form the ganglion. The cecum had a very different blood and nerve pattern versus the colon. The cecum had a fine parallel course of blood vessels and nerves with occasional branches and also numerous cell bodies, whereas the colon had a thicker cobblestone-patterned blood vessel architecture with numerous larger nerve ganglia. Skeletal muscle (thigh) also exhibited Thy1-expressing nerve organization (Figure 10a), and further magnification revealed Thy1-expressing neuromuscular junction (Figure 10b). The bladder was diffusely innervated and vascularized with minimal pattern overlapping (Figures 10c & 10d). Non-distinct expression of Thy1 was also observed along the superficial aspect of the tail, while a ladder-like organization of blood vessels and some nerve filaments were visualized in the core of the tail (Figure 10e).

Figure 8.

Confocal imaging of the TYFD mouse liver and kidney.

Figure 9.

Confocal imaging of digestive system organs of the TYFD mouse.

Figure 10.

Confocal imaging of skeletal muscle, neuromuscular junction, bladder, and tail of the TYFD mouse.

TYFD and dynamic nerve and vessel patterns following corneal injury

We utilized the TYFD mouse strain to assess the changes in corneal blood vessels and nerves in vivo in response to injury by whole corneal epithelium debridement (Figure 11). The superficial layer of the epithelium was removed using an Algerbrush, which typically removes the superficial fibers and also some of the subbasal plexus of the corneal nerve (Figure 13, day 4, black arrowheads). A decrease in basal nerve density was observed after injury (Day 0; Figures 11 & 12). No difference was seen in the corneal blood vessels of basal versus injured corneas, as they were absent due to the corneal angiogenic privilege that keeps the cornea optically clear (see also Figure 3) [44]. Angiogenesis was observed beginning at day 4 post injury and continually progressed until day 14, whereas the nerve regrowth was seemingly slower than that of blood vessels. At day 14, we observed that the nerves were approaching their pre-injury density, although it had not completely returned to the baseline level (Figure 12).

Figure 11.

Axio Zoom fluorescence in vivo imaging of the TYFD mouse cornea at uninjured baseline and over 14 days after whole cornea epithelium debridement.

Figure 13.

Enlarged view of the upper left quadrant region of the TYFD mouse cornea before and after whole cornea epithelium debridement. The white arrowheads indicate trunks of stromal nerve bundles, and the black arrowheads point to regenerated corneal nerve fibers. The white arrows highlight angiogenesis post debridement.

Figure 12.

Quantitated percent areas of hemangiogenesis and corneal nerve regeneration in the TYFD mouse cornea after whole cornea epithelium debridement.

Discussion

Angiogenesis, the formation and branching of new blood vessels from pre-existing vessels, involves coordination of a growing list of molecular signals that direct the proliferation, differentiation, and tubular formation of endothelial cells during tissue morphogenesis, cancer, and metathesis [45, 46]. Similarly, the nervous system is another highly organized network, in which neuronal axons can traverse long distances guided by a plethora of guidance molecules [47–49]. The structural and organizational pattern resemblance between these two expansive and highly branched systems begs the questions of whether there are intercommunications and commonalities in the molecular regulation and guidance processes of nerves and blood vessels and whether the two systems can influence each other’s development and organization. Some nerves and blood vessels follow very similar trajectories to distant tissues, an observation that may be accounted for by shared or interdependent signals. Evidence is now emerging that classical guidance molecules attributed to the development of one system can in fact influence the guidance of the other system [4–6, 8, 9, 11].

In this study, we utilized TYFD mice, which express yellow fluorescence in nerves due to expression of the YFP gene using the Thy1 promoter and red fluorescence in blood vessels due to endothelial expression of the DsRed gene using the Flt1 promoter. Our results showcase the application of TYFD to visualize nerves and vessels in a variety of organs and tissues, including the brain, retina, optic nerve, cornea, trigeminal and dorsal root ganglion, heart, lung, thymus, liver, kidney, digestive system (stomach, small intestine, cecum, and colon), bladder, skeletal muscle, and tail. We also demonstrated the close relationship of blood vessels and neurons in the trigeminal and dorsal root ganglia using this two-photon microscopy technique ex vivo. Flt1-DsRed expression consistently matched what appeared to be tubular networks of blood vessels, while the morphology of Thy1 expression varied between organs, ranging from diffuse (corresponding to either dense neuronal tissues or non-specific expression of Thy1-YFP) to organized and filamentous bands of nerves. Our results exhibited that Thy-YFP expression indeed consistently and robustly marks nervous system-specific tissues ranging from large axonal tracts in the CNS, trigeminal and dorsal root nerve ganglia, and nerve plexuses in the digestive system, down to individual neurons in the retina and nerve endings in neuromuscular junctions.

In addition to simultaneously imaging nerves and blood vessels ex vivo in the organs of TYFD mice, we demonstrate the in vivo application of this animal model for monitoring corneal recovery from superficial epithelial debridement. The cornea serves as an ideal model for observing peripheral nerve recovery from injury, as its accessible location and transparency circumvent the necessity for repeated tissue harvesting and allow long-term monitoring of the same subject. The avascularity of the uninjured cornea also provides a unique model for the study of angiogenesis promoted by injury. In addition to the cornea’s utility for the observation of angiogenesis and nerve regeneration, the study of this tissue has major significance within the field of ophthalmology. Corneal neovascularization is a prevalent cause of blindness as well as a major risk factor and cardinal sign of corneal transplantation rejection [50]. Thus, understanding the molecular underpinnings of blood vessel growth and the association with other physiological entities, including nerves, may have vast implications in treating ophthalmic pathologies. In our experiments, the TYFD mouse corneas showed significant angiogenesis in response to scraping of the whole cornea, while corneal nerves exhibited regeneration albeit slower than the progression of angiogenesis [51]. Moreover, measurement of the percentage areas of blood vessels and nerves, separately, over the entire cornea allowed us to provide a semi-quantitative estimate of the progression of angiogenesis and neuronal regeneration.

By using this double fluorescent reporter model, we sought to understand the gross anatomical organization of blood vessels and nerves in various organs and to evaluate the structural interdependence between the two distinct yet, in some ways, similar networks. Our TYFD mouse model serves as a novel platform for the simultaneous in vivo imaging of nerve and vessel growth. The versatility of the dual fluorescence reporter mouse model for in vivo imaging can be extended beyond corneal imaging by utilizing deep tissue imaging techniques, such as imaging in the cranial or spinal cord windows and multi-photon imaging [52–58]. Two-photon microscopy, a microscopy technique utilizing two photons of infrared light that simultaneously excite a precise point within the sample tissue, has the advantage of penetrating deeper into tissue (up to 1 mm depth) with minimal light scatter [59]. This imaging technique is prevalent in studies of brain function, both physiologically and pathologically [54–58], but also has been used successfully for studies in a variety of other organs and tissues, including the kidneys, skin, skeletal muscle, lymph nodes, and tumors [60, 61]. Garaschuk et al. proposed a method for in vivo brain imaging using two-photon microscopy and a thinned skull window. By utilizing TYFD mice, one could isolate the expression of the neural cells while gaining the ability to simultaneously observe the vascular system [54]. Sigler and Murphy also demonstrated the use of two-photon microscopy to compare the normal versus post-stroke murine brain. They used XFP genetic labeling for neurons and fluorescent dextran to delineate the brain blood flow [55]. TYFD mice can also be utilized for the glioma model proposed by Ricard et al., broadening the observation of cellular dynamics between the glioma, neural, and vascular endothelial cells [58]. TYFD mice can also be potentially used to study the relationship of nerve and vascular endothelial cells in nerve reconstruction following traumatic nerve injury [62, 63].

Furthermore, our dual fluorescent animal model can be used for in vitro experiments requiring neuronal and endothelial co-cultures, i.e., a neurovascular unit as proposed by Lo et al. [64]. This neurovascular unit has been simulated in several in vitro studies; Xue et al. used rat neurons, astrocytes, and microvascular endothelial cells [65] and Chou et al. used human brain endothelial and neural stem cells [66]. Davies et al. carried out an experiment to observe the effect of nerve growth factor (NGF) on excised and then cultured mouse embryonic trigeminal ganglion [67]. They reported that the E10-E12 ganglion were more responsive to NGF than were E9 ganglion. NGF is also known to positively regulate angiogenesis [68–89]. Combined use of the TYFD mouse with trigeminal ganglion culture, such as that described by Davies et al., or co-culture of a neurovascular unit, could expand the novelty of these experiments, allowing for further elucidation of the factors and/or signaling mechanisms that regulate both nerve and blood endothelial cells and how their developmental behaviors affect each other. The endogenous YFP and DsRed signals would allow for simple and continuous visualization of the cellular interactions between the two cell types without the need for immunohistochemistry studies on separate sets of samples at multiple time points. Wang et al. presented a continuous ex vivo study of co-cultured vascular and lymphatic endothelial cells from dual fluorescent reporter mice [90].

In summary, we demonstrated that the TYFD neuro- and angio-fluorescent mouse model is an effective and expedient alternative to traditional immunostaining techniques for visualizing nerves and vasculature in concert at the scales of gross anatomy down to the microscopic level as well as continuously in the same animals during development or following injury.