Rapid Golgi Stain for Dendritic Spine Visualization in Hippocampus and Prefrontal Cortex

Summary

The protocol describes a modification of the rapid Golgi method, which can be adapted to any part of the nervous system, for staining neurons in the hippocampus and medial prefrontal cortex of the rat.

Abstract

Golgi impregnation, using the Golgi staining kit with minor adaptations, is used to impregnate dendritic spines in the rat hippocampus and medial prefrontal cortex. This technique is a marked improvement over previous methods of Golgi impregnation because the premixed chemicals are safer to use, neurons are consistently well impregnated, there is far less background debris, and for a given region, there are extremely small deviations in spine density between experiments. Moreover, brains can be accumulated after a certain point and kept frozen until further processing. Using this method any brain region of interest can be studied. Once stained and cover slipped, dendritic spine density is determined by counting the number of spines for a length of dendrite and expressed as spine density per 10 µm dendrite.

Introduction

The method of using potassium dichromate and silver nitrate to label neurons was first described by Camillo Golgi^1,2 and subsequently used by Santiago Ramon y Cajal to produce an immense body of work differentiating neuronal and glial subtypes. A recently published book with his illustrations is now available³. Following Ramon y Cajal's studies, which were published more than 100 years ago, very little Golgi impregnation was used. Golgi impregnation is a laborious process that allows three-dimensional visualization of neurons with a light microscope. There have been numerous modifications of the Golgi method over the years to make the method easier and the staining more consistent⁴. In 1984, Gabbott and Somogyi⁵ described the single section Golgi impregnation procedure which allowed for more rapid processing. This Golgi impregnation method requires perfusion with 4% paraformaldehyde and 1.5% picric acid, post-fixation followed by vibratome sectioning into a bath of 3% potassium dichromate. Sections are mounted onto glass slides, the four corners of coverslips glued so that when immersed in silver nitrate, diffusion is gradual. Coverslips are then popped off, sections are dehydrated, and eventually cover slipped permanently with mounting medium. This technique was successfully used to label neurons and glia^6,7,8 in the hippocampus. The rapid Golgi method described here is an improvement because there is far less exposure to both potassium dichromate and silver nitrate and no paraformaldehyde and picric acid are used. In addition, although cells that were impregnated using modifications of the Gabbott and Somogyi⁵ method could be analyzed, often the sections were over-or under-exposed or fell off the slides during the dehydration step and generally, several experiments had to be pooled to have enough cells for analysis.

The present protocol describes the use of the Golgi staining kit (see Table of materials) to label dendrites and dendritic spines in the hippocampus and medial prefrontal cortex (mPFC) of the rat. The advantages of this method over previous ones are that it is rapid, there is less exposure to noxious chemicals for the researcher and there is consistent staining of neurons. The protocol described below has been used with minor modifications to assess dendritic spine density in the hippocampus and mPFC of the rat in many studies^{9,10,11,12,13,14,15}.

Protocol

All experimental procedures are approved by the Sacred Heart University Institutional Animal Care and Use Committee and are in accordance with the NIH Guide for the Care and Use of Animals.

1. Isolation and infiltration of brain tissue

1. Premix solutions A and B of the Golgi staining kit 24 h prior to use and keep in dark bottles and/or in dark. Make approximately 80 mL of solution A and B mix which is sufficient to change the solution after 24 h. Store in airtight bottles.
   NOTE: Perfusion with either saline or paraformaldehyde is not necessary.
2. Sacrifice rats by guillotine following carbon dioxide euthanasia and remove brains within seconds. Rinse brains in saline if required but is not necessary.
   1. Place the brain, cortex down so that the hypothalamus is visible because the cuts are made anterior and posterior to it, on a non-porous surface. Cut into an anterior (contains the prefrontal cortex) and posterior (contains the hippocampus) block as shown in Figure 1.
3. Place blocks into the premixed solutions of A and B, making sure that they are well immersed in the solution. Ensure that the volume of solutions A and B is sufficient to immerse the blocks. Store blocks in either brown bottles or clear bottles covered with foil (to keep light out) in the dark at room temperature.
4. Replace solution after 24 h and keep in the dark for another 13 days at room temperature.
   NOTE: If used with mouse brains, there is no need to block, and the entire brain can be placed in the solution. If staining brain areas, which are different in size, the time in solutions A and B may have to be determined by trial and error.
5. After two weeks the tissue is infiltrated well with solutions A and B, transfer the blocks to the cryoprotectant solution (solution C in the Golgi staining kit), and leave them for 48-72 h at 4 °C.
   NOTE: After cryoprotection, blocks may be frozen until further processing. Also, note that all solutions from the kit are collected and disposed of as hazardous waste.
6. Freeze brain, cortex down, on a glass slide on dry ice. Once frozen either cut on the cryostat or store at -80 °C until sectioning using a cryostat.
   ​NOTE: Do not freeze in liquid nitrogen as this produces cracks in the tissue.

2. Sectioning of brain tissues

1. Place a small amount of tissue medium on a pre-cooled cryostat chuck. Mount blocks on cryostat chucks by thawing one side of the block slightly in hand (gloved) and placing it on the tissue medium.
   NOTE: It is not necessary to embed the tissue. The block containing the hippocampus is somewhat easier to cut than the block with the prefrontal cortex because the latter is anterior to the corpus callosum and the two halves are separate.
2. Cut 100 µm sections on a cryostat at -22 °C and mount onto subbed slides. Sections up to 150 µm of thickness can be used. Try to mount 3-4 coronal sections per slide as this decreases the number of slides that require processing. Use one of the following techniques to keep the sections frozen till they get on the slide.
   NOTE: These sections are not fixed in the usual way and therefore they tend to melt quickly. Allow the sections to melt on the slide. If they begin to melt before being placed on the slide, it is impossible to get them to be flat on the slide. There are several techniques to do this.
   1. Use a freezing spray on the knife and the block. Depending on the temperature and humidity in the room this may be necessary for every section. Use either the antiroll plate or a brush to keep the section flat while cutting.
      NOTE: Make sure to have enough freezing spray. Several cans can be used when cutting 20 blocks.
   2. Thaw mount, if possible, by using a room temperature slide and quickly appose it to the section. With practice, one can mount several sections at once. If this doesn't work, keep slides in the cryostat so they are very cold and transfer the section with cold forceps or a cold paintbrush and then thaw mount.
3. Clean the knife with paper wipes between slices. If the knife requires more cleaning, use 100% ethanol (EtOH), and let it dry before cutting the next slice.
4. Once mounted onto the slide, place the slide flat on cardboard slide trays and allow sections to dry at room temperature (for several hours to a maximum of 48 h) in the dark. Do not cover the slide tray. Ensure that the sections are completely dry before Golgi impregnation. Store flat, in the dark, on slide trays until Golgi impregnation.
   ​NOTE: Drying the sections does not cause any damage.

3. Staining and dehydration of brain tissue

1. Perform staining in glass staining dishes. Mix solutions D and E immediately prior to staining. As per the protocol, add one part of solution D, one part of solution E, and two parts of distilled water. For glass staining dishes, make a 200 mL solution to completely submerge the sections. For Koplin jars, this requires less volume.
2. Place slides in racks spaced far enough apart to allow solution access to sections.
3. Place sections in distilled water for 4 min (2x) before placing them into the Golgi impregnation staining solution for 10 min. Change the staining solution every 70 sections. Change the distilled water when it turns yellow.
   NOTE: The timing for the staining step is critical. Too short does not allow enough staining and too long causes over staining, making the dendrites hard to separate when doing analysis. Once out of the staining solution the timing is less critical.
4. Dehydrate the sections as follows: 70% EtOH (5 min), 95% EtOH (5 min; 2x), 100% EtOH (5 min; 2x), clearing agent (5 min; 3x). Change all solutions often, especially the 100% EtOH and clearing agent to ensure the sections remain anhydrous.
   NOTE: It is not necessary to go through a 50% EtOH step. Counterstaining is not needed for cell visualization because the Golgi impregnation provides sufficient contrast. The use of other clearing agents has not worked as well as the one used here (see Table of materials).
5. Coverslip with glass coverslips that are 60 mm long with a generous amount of mounting medium. Make sure that there are, as few air bubbles as possible. If needed, carefully remove, and redo the coverslip (can be done even weeks later). Do this very slowly in order not to damage the section (can be done because of the large amount of mounting medium).
   NOTE: A large amount of mounting medium is somewhat messy but still necessary because enough mounting medium must be used to cover the thick 100 µm sections.
6. To ensure that only one coverslip is placed on the sections, separate the coverslips in advance of the process.
7. Once cover slipped, dry slides flat on any non-porous paper for 3-5 days, moving them slightly, especially after the first day, to avoid sticking. After 3-5 days transfer the slides to slide holders and, ideally, dry slides for at least 3 weeks before examining them. Keep slides flat to decrease the possibility of air bubbles forming.

4. Determination of dendritic spine density

1. For analysis of dendritic spine density in pyramidal neurons of both the mPFC and the CA1 region of the hippocampus, examine the most lateral secondary basal dendrites and the most lateral tertiary apical dendrites as described in step 4.1.1 (Figure 2).
   1. Choose a dendrite, measure the length of the dendrite using an image analysis program, count the spines on the dendrites using a hand counter, and record both length and number of spines.
2. Study and analyze six cells per region (mPFC, CA1) per brain. Quantify a minimum of six brains per group as previously described^7,8. Choose neurons that meet the following criteria for analysis: cell bodies and dendrites are well impregnated; dendrites are distinguishable from adjacent cells and are continuous.
3. Count the spines at 1000x (oil-immersion) by hand counting with a light microscope and measure dendritic length using an image analysis program. Calculate spine density by dividing the spine number by the length of the dendrite and express data as the number of spines/ 10 µm dendrite.
   NOTE: There are far more sophisticated methods for differentiating spine subtypes and dendritic architecture that can be used, but hand counting with a light microscope at 1000x can give a rapid result that can then determine if further investigation is necessary. Although every effort is made to consistently sample similar dendrites, there are variations in thickness that may affect the counting.

Representative Results

Using the rapid Golgi method, cells are consistently well impregnated so that there are plenty of cells to analyze. This is a marked improvement over prior methods where experiments had to be pooled to have enough data for analysis. Therefore, more samples can be processed at once and brains can be stored frozen until processing. Examples of Golgi impregnated cells in the CA1 region of the hippocampus are shown at low and high power in Figure 3. Counting of spines in a given region yields consistent results with small standard errors. This is also important because one can make comparisons between experiments. Figure 4 illustrates an experiment in which basal dendritic spine density was increased on pyramidal cells in adolescent male and female rats after environmental enrichment (EE) in both CA1 and the mPFC. Briefly, male and female rats were weaned at postnatal day (PND) 21 and assigned to control or enriched groups. The EE group spent 2 h/day in enriched housing from PND 24-42 while the control group was housed in ordinary cages during this time. EE induced, in adolescents of both sexes, an increase in basal dendritic spine density in both CA1 and the mPFC. Note the small standard error of the mean (SEM) in all the dendritic spine values.

Figure 1: Ventral surface of rat brain. Photograph of the ventral surface of a fresh rat brain indicating where to cut into anterior and posterior blocks before submerging blocks into initial solutions. Please click here to view a larger version of this figure.

Figure 2: Typical Pyramidal Cell. Schematic of a pyramidal cell, illustrating apical and basal dendrites which are analyzed for dendritic spine density. Please click here to view a larger version of this figure.

Figure 3: Golgi Impregnated Neuron. Examples of Golgi impregnated neurons in the CA1 region of the rat hippocampus. Left: Several impregnated pyramidal cells. Scale bar = 25 µm. Upper right: Basal dendrites. Scale bar = 12.5 µm. Lower right: Example of a secondary basal dendrite. Arrows denote spines. Scale bar = 5 µm. Please click here to view a larger version of this figure.

Figure 4: Spine Density. A data set illustrating basal and apical dendritic spine density in the mPFC and CA1 region of the hippocampus (CA1) following environmental enrichment (EE) compared to control (CON) in male (M) and female rats (F). Histograms show the average number of spines/10 µm dendrite + SEM. Data were analyzed using statistical analysis software (see Table of materials). Two-way (sex X EE) ANOVAs were used to test for group differences and Fisher's LSD tests were used for post-hoc analysis. Significant effects are p<0.05. t denotes a significant difference. Please click here to view a larger version of this figure.

Discussion

The present protocol describes a method of Golgi impregnation that allows for rapid simultaneous processing of many sections. It is an improvement over previously described⁵ more labor-intensive methods and consistently yields impregnated neurons for analysis. In addition, there is less exposure to toxic chemicals used in Golgi impregnation. The most challenging part of the process is getting the sections to be flat on the slides, which takes considerable practice. Keeping everything as cold as possible with the use of freezing spray is essential.

Once slides are dry and analysis can be performed, it is very important to be consistent in selecting the cells that are counted. Hippocampal pyramidal cells are chosen from CA1. For the mPFC, which has several subparts, pyramidal cells from the infralimbic cortex are used. For reasons that are not clear, fewer cells in the mPFC are stained than in the CA1 region of the hippocampus. In addition, it is possible to combine experiments when all the sections cannot be processed together for logistical reasons. For consistency, the same person should count a given set of cells.

The limitations of this method are similar to all Golgi impregnation methods. The fact that only a small number of cells are stained is an advantage. The small number of cells impregnated allows for visualization of the entire cell in three dimensions. The disadvantage of the Golgi method is that it is not clear which subset of cells is labeled. Therefore, in experiments, one must assume that the cells impregnated for both control and experimental groups are the same. Even though this method results in far better staining than previous methods, there are always cells that cannot be analyzed because they are covered by debris, an air bubble, or have broken dendrites.

In conclusion, the rapid Golgi method described here is a fast, safe method for consistent and reproducible labeling of neurons that can be used for any brain region. In addition to labeling cells in the prefrontal cortex^17,18 and hippocampus^10,12,19, it has also been used in the amygdala²⁰, cerebellum²¹, and cortex²². Assuming one is familiar with the anatomy and can quickly identify cells, hand counting of spine density can provide a quick result, but it does not provide information on dendritic subtypes which requires more sophisticated methods of analysis.

Materials

Name	Company	Catalog Number	Comments
Cardboard slides trays	Fisher Scientific	12-587-10
Coverslips 24 x 60mm	Fisher Scientific	12-545-M
FD Rapid GolgiStain kit	FD Neurotechnologies	PK 401	Stable at RT in the dark for months; Golgi staining kit
Freezing Spray	Fisher Scientific	23-022524
HISTO-CLEAR	Fisher Scientific	50-899-90147	clearing agent
NCSS Software	Kaysville, UT, USA
Permount	Fisher Scientific	SP-15-100	mounting medium
Superfrost Plus Microscope slides	Fisher Scientific	12-550-15
Tissue Tek CTYO OCT Compound	Fisher Scientific	14-373-65	Used to mount brains on cryostat chuck