In-Vivo Calcium Imaging of Sensory Neurons in the Rat Trigeminal Ganglion

Summary

Genetically encoded calcium indicators (GECI) enable a robust, population-level analysis of sensory neuron signaling. Here, we have developed a novel approach that allows for in vivo GECI visualization of rat trigeminal ganglia neuron activity.

Abstract

Genetically encoded calcium indicators (GECIs) enable imaging techniques to monitor changes in intracellular calcium in targeted cell populations. Their large signal-to-noise ratio makes GECIs a powerful tool for detecting stimulus-evoked activity in sensory neurons. GECIs facilitate population-level analysis of stimulus encoding with the number of neurons that can be studied simultaneously. This population encoding is most appropriately done in vivo. Dorsal root ganglia (DRG), which house the soma of sensory neurons innervating somatic and visceral structures below the neck, are used most extensively for in vivo imaging because these structures are accessed relatively easily. More recently, this technique was used in mice to study sensory neurons in the trigeminal ganglion (TG) that innervate oral and craniofacial structures. There are many reasons to study TG in addition to DRG, including the long list of pain syndromes specific to oral and craniofacial structures that appear to reflect changes in sensory neuron activity, such as trigeminal neuralgia. Mice are used most extensively in the study of DRG and TG neurons because of the availability of genetic tools. However, with differences in size, ease of handling, and potentially important species differences, there are reasons to study rat rather than mouse TG neurons. Thus, we developed an approach for imaging rat TG neurons in vivo. We injected neonatal pups (p2) intraperitoneally with an AAV encoding GCaMP6s, resulting in >90% infection of both TG and DRG neurons. TG was visualized in the adult following craniotomy and decortication, and changes in GCaMP6s fluorescence were monitored in TG neurons following stimulation of mandibular and maxillary regions of the face. We confirmed that increases in fluorescence were stimulus-evoked with peripheral nerve block. While this approach has many potential uses, we are using it to characterize the subpopulation(s) of TG neurons changed following peripheral nerve injury.

Introduction

Somatosensation, the neural encoding of mechanical, thermal, and chemical stimuli impinging on the skin or other bodily structures, including muscles, bone, and viscera, starts with activity in primary afferent neurons that innervate these structures¹. Single unit based electrophysiological approaches have provided a wealth of information about the afferent subtypes involved in this process as well as how their stimulus-responses properties may change over time^1,2,3. However, while there remains strong evidence in support of the labeled line theory, which suggests specific sensory modalities are conveyed by specific subpopulation(s) of neurons, the ability of many subpopulations of neurons to respond to the same types of mechanical, thermal, and chemical stimuli suggests the majority of somatosensory stimuli are encoded by multiple subpopulations of neurons^4,5. Thus, a better understanding of somatosensation will only come with the ability to study the activity of 10's, if not hundreds, of neurons simultaneously.

Advances in optical approaches with the relatively recent advent of confocal and, subsequently, multiphoton and digital imaging techniques have facilitated the ability to perform relatively non-invasive population-level analyses of neuronal activity^6,7. One of the last hurdles in the application of this technology has been the development of tools to enable the optical assessment of neural activity. Given the speed of an action potential that can start and end in less than a millisecond, a voltage-sensitive dye with the capacity to follow changes in membrane potential at the speed of an action potential would be the ideal tool for this purpose. But while there has been tremendous progress in this area^7,8,9,10, the signal-to-noise ratio for many of these dyes is still not quite high enough to enable a population analysis of hundreds of neurons at the single cell level. As an alternative approach, investigators have turned to monitoring changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i). The limitations with this strategy have been clear from the start and include the fact that an increase in [Ca²⁺]_i is an indirect measure of neural activity¹¹; that an increase in [Ca²⁺]_i may occur independently of Ca²⁺ influx associated with the activation of voltage-gated Ca²⁺ channels (VGCCs)^12,13; that the magnitude and duration of a Ca²⁺ transient may be controlled by processes independent of VGCC activity^11,12,14; and that the time-course of Ca²⁺ transients far exceeds that of an action potential¹⁵. Nevertheless, there are a number of significant advantages associated with the use of Ca²⁺ as an indirect measure of neural activity. Not the least of these is the signal-to-noise ratio associated with most Ca²⁺ indicators, reflecting both the magnitude of the change in intracellular Ca²⁺ and the fact that the signal is arising from the three-dimensional space of the cytosol rather than the two-dimensional space of the cell membrane. Furthermore, with the development of genetically encoded Ca²⁺ indicators (GECI's), it is possible to take advantage of genetic strategies to drive the expression of the Ca²⁺ indicators in specific subpopulations of cells, facilitating population-level analyses in intact preparations (e.g., see¹⁶).

Given the number of genetic tools now available in mice, it should be no surprise that GECI's have been used most extensively in this species. Mouse lines with constitutive GECI expression in subpopulations of sensory neurons have been developed^7,16,17. With the development of mouse lines expressing recombinases in specific cell types, it is possible to use even more sophisticated strategies to control GECI expression¹⁵. However, while these tools are ever more powerful, there are a number of reasons why other species, such as rats, might be more appropriate for some experimental questions. These include the larger size, facilitating a number of experimental manipulations that are difficult, if not impossible, in the smaller mouse; the ease of training rats in relatively complex behavioral tasks; and at least some evidence that biophysical properties and expression patterns of several ion channels in rat sensory neurons may be more similar to that observed in human sensory neurons than are the same channels in mouse relative to the human¹⁸.

While the transduction of somatosensory stimuli generally occurs in the peripheral terminals of primary afferents, the action potential initiated in the periphery must pass through the structure that houses primary afferent somata, referred to as dorsal root (DRG) or trigeminal (TG) ganglia before reaching the central nervous system¹⁹. While there is evidence that not every action potential propagating along a primary afferent axon will invade the cell body²⁰, a consequence of the fact the primary afferent somata are connected to the main afferent axon via a T-junction¹⁹, the majority of action potentials initiated in the periphery appear to invade the soma²¹. This confers three experimental advantages when using GECIs to assess population coding in primary afferents: the large size of the cell body relative to the axons further increases the signal to noise when using [Ca²⁺]_i as an indirect measure of afferent activity; the DRG are generally easy to access; and assessing activity at a site that is spatially remote from the afferent terminals minimizes the potential impact of the surgery needed to expose the ganglia on the stimulus-response properties of the afferent terminals. However, because TG are located beneath the brain (or above the palette), they are far more difficult to access than DRG. Furthermore, while there are many similarities between DRG and TG neurons, there is a growing list of differences as well. This includes the roughly somatotopic organization of neurons in the TG²², unique structures innervated, different central terminal termination patterns^23,24,25,26, and now a growing list of differences in both gene expression^27,28 and functional receptor expression²⁹. In addition, because we are interested in the identification of peripheral mechanisms of pain, the relatively large number of pain syndromes that appear to be unique to the trigeminal system (e.g., migraine, trigeminal neuralgia, burning mouth syndrome) that appear to involve aberrant activity in primary afferents^30,31,32, suggests that the TG needs to be studied directly.

Thus, while stimulus-response properties of TG neurons have been studied with GECIs in the mouse¹⁶, because the reasons listed above suggest that the rat may be a more appropriate species to address a variety of experimental questions, the purpose of the present study was to develop an approach to use GECIs to study TG neurons in the rat. To achieve this, we utilized a viral approach to drive the expression of the GECI GCaMP6s in the peripheral nervous system. We then removed the forebrain to allow access to the TG. Finally, mechanical and thermal stimuli were applied to the face while neuronal responses were assessed under fluorescent microscopy. Together, these data support a role for utilizing the rat to investigate changes in the TG under many states, expanding the toolkit for investigators interested in sensory coding in the trigeminal system.

Protocol

All experiments involving the use of animals in research were performed in accordance with standards put forth by the National Institutes of Health and the International Association for the Study of Pain and were approved by the University of Pittsburgh Institutional Animal Care and Use Committee (protocol #22051100). At the end of each experiment, rats were euthanized via exanguination with cardiac perfusion of ice-cold phosphate-buffered saline (PBS), an approach approved by the American Veterinary Medical Association and the University of Pittsburgh IACUC.

1. GCaMP induction

1. Order time-pregnant Sprague Dawley rats so that pups can be injected at the appropriate time after birth.
2. Spray the gloves with 70% EtOH before handling each rat pup. This will deter the dam from cannibalizing the young.
3. Swab rat pups (P1-2) with 70% EtOH and anesthetize on ice for 3 min.
4. Inject 15 µL of AAV9-CAG-WPRE-GCaMP6s-SV40 (Addgene) intraperitoneally using a 25 µL sterile, gas-tight Hamilton syringe.

2. Trigeminal ganglion exposure surgery

1. Administer anesthetic cocktail (55 mg/kg ketamine, 5.5 mg/kg xylazine, 1 mg/kg acepromazine) intraperitoneally based on body weight to 6-8 week-old rats (approximately 150-200 g).
   NOTE: This is generally sufficient to maintain a surgical plane of anesthesia as assessed by the absence of a withdrawal reflex to noxious pinch of a hindpaw. However, supplement with isoflurane via a nose cone if the animals become light.
2. Once fully anesthetized, shave the hair and whiskers of the head and face.
3. Mount the rat to a stereotaxic frame with ear bars and place a heating pad (~37 ^oC) underneath to maintain body temperature.
4. Monitor vital signs (heart rate, respiration rate, blood oxygen saturation) with a mouse oximeter or comparable.
   NOTE: Body temperature is monitored by a rectal probe and maintained by placing rats on a feedback controlled circulating water blanket.
5. Place gauze dipped in ice-cold saline on the head to constrict blood vessels to minimize bleeding. Using a size 15 scalpel, make a midline incision of the skin and muscle over the skull.
6. Use blunt dissection of the skin and muscle to expose the skull.
7. Using a ¼ round drill bit, carefully perforate the skull cap to expose the forebrain. Then, use rongeurs (2.5 mm cup) to carefully cut through the skull.
8. Using a size 15 scalpel, make an incision into the brain (Bregma: -3.80) and the olfactory bulb.
   NOTE: Cutting more caudally past this point will result in rat death
9. Use a spatula to carefully disconnect the dura from the skull and gently lift the severed brain to reveal the TG and the base of the skull.
   NOTE: Additional use of ice-cold saline perfusion throughout the extraction will minimize bleeding and optimize the following dissection field.
10. Using a cautery pen, stem any bleeding resulting from the extraction.
    ​NOTE: Cutting the dura will result in inevitable bleeding. It is best to prepare ice-cold aCSF (119 mM NaCl, 26.2 mM NaHCO₃, 2.5 mM KCl, 1 mM NaH₂PO₄, 1.3 mM MgCl₂, 10 mM glucose, 2.5 mM CaCl₂) to bathe the skull cavity to help constrict blood vessels while maintaining neuronal health.

3. GCaMP6s imaging

NOTE: Given the size and density of these neurons, the imaging and data acquisition system used (objective, microscope, light source, camera) will determine the number of GECI⁺ cells visualized. The light source, objective, and camera will also determine the parameters used for image acquisition, including exposure time and image capture rate. While multiphoton and confocal techniques can be used depending on the experiment's parameters, epifluorescence microscopy may be sufficient to resolve many cells. Any image acquisition package can be used. Ideally, the stimulus application is time-locked with the image acquisition software package.

1. Place the skull cavity underneath the objective and bring the TG into focus using visible light.
2. The excitation wavelength of GCaMP6s is 496 nm, and its emission wavelength is 513 nm; thus, use the appropriate dichroic and filter cubes to locate GCaMP⁺ cells. Adjust the focus to locate and resolve the area of the ganglia in which neurons are responding to stimuli applied to the receptive field of interest.
3. Use a 10x objective to enable visualizations of most neurons in the TG responsive to mechanical stimuli applied to a 1 cm² region of the face¹⁶. Use a 20x dry objective with a long working distance (10.8 mm) to increase resolution.
4. Use acquisition software (e.g., Metamorph) to collect fluorescence data over time and in response to stimulus application.
   1. To minimize photo-bleaching of the neurons, use as short exposure time as possible to detect baseline and evoked increases in fluorescence. With the 20x, 0.40 NA air objective, a 120 W mercury halide light source, and a complementary metal-oxide-semiconductor (CMOS) camera used in the present study, an exposure time of 300 ms, and an image acquisition rate of 3 Hz, obtain stable baseline recordings for >90 min.
      NOTE: 1) These image acquisition parameters must be determined empirically. 2) Mechanical (brush, punctate, vibration, pinch), thermal (heat and cold), and chemical (capsaicin, menthol, inflammatory mediators) may be applied to the receptive field. A relatively inexpensive subjective approach is to apply the stimuli by hand. More objective and reproducible approaches are recommended, however, where feedback-controlled actuators can be used for repeated application at a known force. While feedback-controlled Peltier devices are useful for controlled heating and cooling, they are not ideal for the thermal stimulation of a curved face. Feedback-controlled infrared light sources are an alternative for heating, and cold spray is a less than ideal alternative for cooling.

Results

Because we have previously had success with the AAV9 serotype for the infection of rat sensory neurons¹⁵, we used this serotype for the expression of GCaMP6s in rat TG neurons. We therefore first sought to assess the sensory neuron infection efficiency of AAV9-CAG-GCaMP6s-WPRE-SV40 (AAV9-GCaMP) when this virus was administered to neonatal rat pups²⁰. This virus utilizes the CAG promoter, which drives and maintains high levels of gene expression. Furthermore, AAV9 has been s...

Discussion

Here, we demonstrate a quick, non-invasive way of generating a GECI rat for imaging the TG. We chose a CAG promotor to drive and maintain high levels of gene expression. While previous studies suggest that other AAV serotypes may efficiently drive gene expression in DRG neurons³⁹, our results are consistent with a recent study involving intraperitoneal injection of AAV in neonates³², indicating that the AAV9 serotype is highly efficient in the infection of rat neonatal sens...

Materials

Name	Company	Catalog Number	Comments
AAV9-CAG-WPRE-GCaMP6s-SV40	Addgene	100844-AAV9	AAV9-GCaMP6s virus
ACEpromazine maleate	Covetrus	11695-0095-5	10 mg/mL
AnaSed (Xylazine) injection	AKORN Animal Health	23076-35-9	20 mg/mL
CTR5500 Electronics box	Leica	11 888 820	Power Supply
Cutwell burr drill bit	Ransom & Randolph		¼ round
DM 6000 FS	Leica	11 888 928	Base Stand
EL6000	Leica	EL6000	Light source with 120 W mercury bulb
Forceps	FST	11252-00	Dumont No. 05
Friedman rongeurs	FST	16000-14	2.5 mm cup size
Friedman-Pearson rongeurs	FST	16021-14	1 mm cup size
Heating pad (Temperature therapy pad)	STRYKER	8002-062-022
Ketamine hydrochloride	Covetrus	1695-0703-1	100 mg/mL
Plan Fluor 20x/0.40	Leica	MRH00105	20x objective, 0.4 NA10.8 mm WD
Power handle high-temp cautery pen	Bovie	HIT1	handheld Change-A-Tip cautery pen
Prime 95B	Photometrics	Prime 95B	CMOS Camera
Saline	Fisher Scientific	NC0291799	0.9% Sterile Saline
Scalpel blade	Fisher Scientific	22-079-701	size 15 disposable blade
Spatula	BRI	48-1460	brain spatula
Spring scissors	FST	91500-09	Student Vannas, 5 mm cutting edge
Spring scissors	FST	15012-12	Noyes, 14 mm cutting edge
STP6000 Smart touch panel	Leica	11 501 255	Control Panel
Syringe	Hamilton	80201	25 μL Model 1702 Luer Tip syringe
Water heater	Adroit	HTP-1500