A Laser Capture Microdissection Protocol That Yields High Quality RNA from Fresh-frozen Mouse Bones

Tissues are made up of heterogeneous and spatially distributed cell types. Different cell types in a given tissue may respond differently to the same signal. Therefore, it is essential being able to isolate specific cell populations for the assessment of the role of different cell types in both physiological and pathological conditions. Laser capture microdissection (LCM) offers a relatively rapid and precise method for isolating and removing specified cells from complex tissues¹. LCM systems use the power of a laser beam to separate the cells of interest from histological tissue sections without the need for enzymatic processing or growth in culture. This means that the cells are in their natural tissue habitat, and that tissue architecture including the spatial relationship between different cells is retained. Morphology of both the captured cells and the residual tissue is well preserved, and several tissue components can be sampled sequentially from the same slide. Isolated cells can then be used for subsequent analysis of their RNA, DNA, protein or metabolite content^2,3.

In order to analyze gene expression in different cell populations, or after different treatments, it is necessary to obtain mRNAs of sufficient quality and quantity for the subsequent analysis^4,5. In contrast to DNA, RNAs are more sensitive to fixation and the use of frozen tissue is recommended when the objective is to study RNA. Since mRNAs are quickly degraded by ubiquitous ribonucleases (RNase), stringent RNase-free conditions during specimen handling and preparation and avoiding storage of samples at room temperature are required. In addition, rapid techniques without any prolonged aqueous phase steps are crucial to prevent RNA degradation⁶. The RNA yield and integrity can also be affected by the LCM process and the LCM system used^7,8. Currently, four LCM systems with different operating principles are available². The method of RNA extraction can also be important, since different RNA isolation kits have been tested with significant differences in RNA quantity and quality^7,8.

Any tissue preparation method requires finding a balance between obtaining good morphological contrast and maintaining RNA integrity for further analyses. For preparing frozen sections from bone, an adhesive film was developed and continuously improved⁹. The bone sections are cut and stained directly on the adhesive film. This adhesive film is applicable to many types of staining, and can be employed to isolate the cells of interest from bone cryosections using LCM^{9,10,11,12,13,14}. All the steps including the surgical removal, embedding, freezing, cutting and staining can be completed within less than one hour. Importantly, cells such as osteoblasts, bone lining cells, and osteoclasts can be clearly identified^{9,10,11,12,13,14}. This method has the advantage of being rapid and simple. An alternative method for generating bone cryosections is to use the tape transfer system¹⁵. However, the latter technique is more time-consuming and requires additional instrumentation, since the sections have to be transferred from the adhesive tape onto precoated membrane slides by ultraviolet (UV) cross-linking. Although the tape transfer system has been successfully coupled with LCM^16,17,18,19, it should be noted that the cross-linked coating can create a background pattern that can interfere with cell-type identification²⁰.

Typically, only small amounts of RNA are extracted from microdissected cells, and RNA quality and quantity are often assessed by micro-capillary electrophoresis²¹. A computer program is used to assign an index of quality to RNA extracts called RNA integrity number (RIN). A RIN value of 1.0 indicates completely degraded RNA, whereas a value of 10.0 suggests that the RNA is fully intact²². Usually, indexes over 5 are considered sufficient for RNA studies. Gene expression patterns in samples with an RIN value of 5.0−10.0 have been reported to correlate well with each other²³. Although the sensitivity of this method is high, since as little as 50 pg/µL of total RNA can be detected, it can be very difficult to obtain a quality assessment if the RNA concentration in the sample is very low. Therefore, in order to assess RNA quality, the tissue section remaining after the LCM is often used to extract RNA, by pipetting buffer onto the slide²⁴.

Although LCM has been used extensively on different frozen tissues, RIN values of extracted RNAs are rarely reported. Furthermore, there are no comparative studies to clarify the most appropriate method to study RNA in mouse bones. In the present study, frozen sections from adult mouse femurs were used to optimize sample preparation, LCM protocol and RNA extraction in order to obtain high quality RNAs. The present protocol was optimized particularly for the LCM system that uses gravity for sample collection.

Bone tissue from mice was used in strict accordance with prevailing guidelines for animal care and all efforts were made to minimize animal suffering.

1. Animals and Freeze Embedding

1. House animals in conditions of constant room temperature (RT; 24 °C) and a 12 h light/12 h dark cycle with free access to food and water.
   NOTE: Bone tissues in this study were obtained from 3-month-old male wild type C57BL/6 mice.
2. At necropsy, exsanguinate mice from the abdominal vena cava under general anesthesia (ketamine/xylazine, 100/6 mg/kg intraperitoneal).
   NOTE: In order to avoid RNA degradation, use RNase-free instruments and materials, wear gloves and work quickly avoiding storage of samples at RT throughout the whole experiment.
3. Remove whole femurs rapidly, clean them of surrounding soft tissues using scalpel and paper towels. Pour optimal cutting temperature (OCT) compound into the embedding mold, and put the femur into the bottom of the embedding mold. Snap-freeze the samples in liquid nitrogen. OCT is transparent at RT and white when frozen.
4. Once entirely frozen, wrap the samples in foil and transfer them on dry ice to -80 °C. Store samples at -80 °C until further processing.
   NOTE: The protocol can be paused here.

2. Section Preparation

1. Set the temperature in the cryostat to -19 °C and of the block holder to -17 °C. Wipe the cryostat interior using 70% ethanol. Place the disposable blade for hard tissues, the glass slides and a fitting tool into the cryostat to cool. Keep them inside the cryostat for the duration of sectioning.
2. Transfer the frozen tissue block on dry ice to the cryostat and allow it to equilibrate for at least 10 min.
   NOTE: Avoid repetitive thawing and frequent cycling of one block from -80 °C to -19 °C for cryosectioning.
3. Press the bottom of the embedding mold to push the OCT block out of the mold. Apply enough OCT medium to the block holder to adhere the block to it. Wait until the OCT medium is fully frozen.
4. Place the block holder in the object holder and tighten it in place. Adjust the blade position and trim the block using 15 µm cutting increments to remove the OCT covering the sample. If the sample has been cut earlier and the surface exposed to air, discard the first 2−3 tissue sections from the block.
5. Adjust the cryostat to generate 8 µm sections and cut 2−3 cryosections that will be discarded (the first few will typically be thicker than 8 µm). Place the adhesive film on the block and use a fitting tool to adhere the film to the block. Make the cut slowly and at a constant speed holding the section by the film.
6. Place the film (sample facing up) immediately on a precooled glass slide (on the cryobar within the cryostat) to avoid thawing of the sample. Use tape to fix the film to the glass slide for easier staining. Proceed immediately with the staining protocol.

3. Rapid Staining Protocol

1. Prepare 40 mL of the following solutions in 50 mL tubes and put them on ice: 95% ethanol, RNase-free water, 100% ethanol, 100% ethanol and 100% xylene. Perform all steps on ice (except the staining). For each experimental day, prepare all aqueous reagents fresh with RNase-free water.
   CAUTION: Work in the hood to avoid intoxication by xylene.
2. Incubate sections for 30 s in 95% ethanol, then dip sections 30 s in RNase-free water to remove OCT carefully and completely which may interfere with LCM and downstream applications.
3. Dispense 50 µL of commercial LCM frozen section stain (Table of Materials) onto the section, incubate for 10 s at RT and drain it by placing the edge of the slide on absorbent tissue paper. Remove excess stain by rinsing 30 s in 100% ethanol.
4. Immerse bone sections in a second tube with 100% ethanol for 30 s and transfer them to 100% xylene for 30 s.
5. Put adhesive film (sample facing up) on dry glass slide as a support. Take care that the film is not impacted and placed as flat as possible. Do not allow the sample to dry completely.
6. Place a PET membrane frame slide on it. Shortly press a gloved finger on the membrane to attach it to the film. The sample is then sandwiched between the membrane and the adhesive film. The adhesive film should not be folded or wrinkled and there should be no air bubbles between the film and the membrane. Proceed immediately with the LCM.

4. Laser Capture Microdissection

1. Clean the stage and cap holder from RNase using surface decontaminant (Table of Materials). Load the slide and the caps in the slide holder and the cap holder, respectively.
2. Adjust the focus and acquire slide overview with 1.25x objective. Change to the 40x objective and adjust the focus. Choose the area of interest using the slide overview. Adjust laser parameters as follows: aperture, 7; laser power, 60; speed, 5; pulse frequency, 201; specimen balance, 20. Optimize these parameters for each objective.
3. If the laser fails to cut the sample, increase the laser power. Alternatively, the laser can be applied more than once. In addition, inspect the target for spots of incomplete cuts and re-drawn the line in these spots using the Move and cut option. Save laser settings for the dissection of subsequent slides.
4. Select osteoblasts, osteocytes and bone lining cells in distal femoral cancellous or cortical bone based on morphologic criteria. Draw a line for laser path further away from the target cells to minimize the damage by the UV laser. Perform all microdissections within less than 1 h after the staining.
5. Collect each cell type in a separate cap of a 0.5 mL tube.
   NOTE: Dry capturing instead of liquid recovery may avoid RNA degradation. In addition, very small volumes of buffer contained in the cap of the microcentrifuge tube can either evaporate or crystallize (depending on its composition) during the LCM.
6. Confirm capture success by observation of the collection tube cap after the LCM where applicable. Proceed immediately with the RNA extraction.

5. RNA Extraction

1. Dispense 50 µL of the lysis buffer containing β-mercaptoethanol into the cap of the collection tube and lyse the sample by pipetting up and down in the cap for 1 min. Spin down the lysate and add 300 µL of the lysis buffer containing β-mercaptoethanol into the tube. If several caps are going to be pooled, take care that the total volume of buffer is as recommended for the RNA extraction kit (Table of Materials).
   CAUTION: β-mercaptoethanol must be added to protect RNA from degradation, but it is considered toxic. Wear protective clothes and gloves and work in the hood.
2. In addition, for each slide prepare one labeled 1.5 mL microcentrifuge tube with 350 µL of the lysis buffer containing β-mercaptoethanol. Use the sections remaining after the LCM to extract RNA. Carefully separate the film from the membrane and lyse the sample by slowly pipetting the lysis buffer onto the section several times.
   NOTE: Total amount of lysis buffer should be 350 µL. Do not use all of it at once for digestion as it will flow off the slide.
3. Put the lysate samples on dry ice and store them at -80 °C.
   NOTE: The protocol can be paused here.
4. Thaw the lysates at RT. Transfer lysates from LCM-harvested cells from collection tubes into the 1.5 mL microcentrifuge tubes. If more than one cap was used to harvest the sample, pool several lysates.
5. Extract RNA according to the manufacturer’s instructions. Treat columns with DNase I to remove genomic DNA. Elute RNA with 14 µL of RNase-free water, resulting in a 12 µL eluate. Store RNA at -80 °C.
   NOTE: The protocol can be paused here.

6. Measurement of RNA Yield and Integrity

1. Thaw RNA on wet ice. Measure the yield and integrity of isolated RNA using micro-capillary electrophoresis. Load total RNA (1 µL per sample) into a chip (Table of Materials) according to the manufacturer’s instructions.

An LCM protocol was developed to obtain sufficient quantity of high-quality RNA for gene expression analysis in bone cells of mouse femurs. In the optimized protocol, 8 µm-thick frozen bone sections were cut on an adhesive film and stained using a rapid protocol for a commercial LCM frozen section stain. The sample was sandwiched between the PET membrane and the adhesive film. Mouse bone cells were microdissected using an LCM system that uses gravity for sample collection. A column-based RNA extraction method was used to obtain a high-quality RNA of sufficient yield. The yield and integrity of isolated RNA was measured using micro-capillary electrophoresis (Figure 1).

The difference in RNA quality and quantity obtained using different lysis protocols can be seen in the representative gel and electropherograms. When the sample was lysed by pipetting up and down in the cap for 1 min, it was possible to isolate approximately 8.5 ng of RNA from 1 mm² microdissected bone tissue (8 µm-thick section). RIN value was 8.60 (Figure 2).

Alternatively, LCM was performed using an LCM system that uses de-focused laser pulse, which catapults the material into the overhanging adhesive cap. RNA quality and quantity can be estimated in the representative gel and electropherograms. For fresh frozen bones, it was possible to isolate approximately 1.6 ng of RNA from 1 mm² microdissected bone tissue (8 µm-thick section). The RIN value was 1 (Figure 3).

Figure 1: Flowchart of laser capture microdissection protocol for fresh-frozen bones. Please click here to view a larger version of this figure.

Figure 2: Representative gel (top) and electropherograms (bottom) of RNA samples. LCM was performed using an LCM system that uses gravity for sample collection. RNA samples were retrieved from LCM-harvested tissue (samples 1, 3, 5, 7, and 9) and corresponding control sections (samples 2, 4, 6, 8, and 10, respectively). One RNA sample was retrieved from the control section that was stained but was not used for LCM (sample 11). A total area of 1 mm² was microdissected, and different lysis protocols were used. Sample 1: 350 µL of the lysis buffer containing β-mercaptoethanol were added into the tube. The cap was closed carefully, and the tube inverted. LCM-harvested cells were lysed by vortexing 1 min, incubating at RT for 10 min and vortexing 1 min. Sample 3: 350 µL of the lysis buffer containing β-mercaptoethanol was added into the collection tube, the cap was carefully closed, and the tube inverted. LCM-harvested cells were lysed by incubating collection tubes upside down for 30 min at RT. Sample 5: 50 µL of the lysis buffer containing β-mercaptoethanol was added into the cap of the collection tube and the sample was lysed by pipetting up and down in the cap for 1 min. The lysate was centrifuged and 300 µL of the lysis buffer containing β-mercaptoethanol was added into the tube. Sample 7: 350 µL of the lysis buffer containing β-mercaptoethanol was added into the tube. The cap was closed carefully, and the tube inverted. LCM-harvested cells were lysed by vortexing and inverting several times. Sample 9: 50 µL of the lysis buffer containing β-mercaptoethanol was added into the cap of the collection tube, and the sample was lysed by incubation for 5 min at RT in the cap. The lysate was centrifuged, and 300 µL of the lysis buffer containing β-mercaptoethanol was added to the tube. Please click here to view a larger version of this figure.

Figure 3: Representative gel (top) and electropherograms (bottom) of RNA samples. LCM was performed using an LCM system that uses de-focused laser pulse, which catapults the material into the adhesive cap positioned above the section. RNA samples were retrieved from LCM-harvested tissue (samples 5 and 6) and corresponding control section (sample 7). One RNA sample was retrieved from the control section that was stained but was not used for LCM (sample 8). A total area of 0.5 mm² or 1 mm² was microdissected (samples 5 and 6, respectively). 350 µL of the lysis buffer containing β-mercaptoethanol was added into the collection tube, the cap was carefully closed and the tube inverted. LCM-harvested cells were lysed by incubating collection tubes upside down for 30 min at RT. Please click here to view a larger version of this figure.

Both RNA quality and quantity can be affected negatively at all stages of the sample preparation such as tissue manipulation, LCM process, and RNA extraction. Therefore, an LCM protocol was developed to obtain sufficient amount of high-quality RNA for subsequent gene expression analysis.

For LCM, most laboratories use sections 7−8 µm thick². Thicker sections would allow more material to be harvested. However, if they are too thick, this could reduce the microscopic resolution and the laser may not be able to cut through. It is likely that optimal tissue thickness depends on the LCM system (and the laser and slide type) used, as well as on the tissue in question^8,25. In the present study, cryosections of different thicknesses (4, 8 or 12 µm) were tested; 8 µm-thick sections were found to be ideal for LCM, while 12 µm bone sections were more difficult to cut with the laser and 4 µm sections gave lower amounts of RNA.

The endogenous RNase activity varies between different tissues. Therefore, staining protocols developed for tissues with very low RNase activity could be unsuitable for other tissue types. Nuclear fast red²⁶ and cresyl violet²⁴ were suggested to be the best in terms of preserving RNA integrity. In addition, alcohol-based methods were superior to aqueous stains for maintaining RNA integrity. However, they suffered from irreproducible staining intensity²⁷. Time is an important parameter to consider. On the one hand, the time required to search and identify cells of interest in the section, and to perform LCM is often relatively long. On the other hand, time available for LCM is limited, since, in some cases, 20% RNA degradation was reached after 30 min²⁸. Therefore, different methods were applied in order to stabilize RNA, such as exposure of sections to argon flux during the microdissection²⁸, or use of buffered alcohol-based cresyl violet staining and keeping the humidity level in the laboratory low²⁷. In addition, a maximum of 15 min for the microdissection step was suggested². In this study, frozen bone sections were stained using a rapid protocol for a commercial LCM stain, and even after 1 h at RT, RIN values decreased from 8.3 to 9.1. Therefore, all microdissections were performed within less than 1 h after staining. In addition, different protocols for staining were tested (with shorter incubations in aqueous reagents or without the xylene step). No significant improvement in RIN values was achieved.

In LCM studies, two different types of RNA extraction methods have been compared: a phase separation method vs. a column-based method. It was found that the RNA extraction method based on columns led to better RNA quality and higher yield compared to the phase separation method⁸. In another study, a commercial kit that uses minimal digestion time and temperature (5 min at RT) resulted in superior RNA quality compared to an RNA isolation kit that needs longer digestion time at higher temperature (30 min at 42 °C)⁷. In the present study, it was possible to extract high quality RNA (RIN > 8) of sufficient concentration from LCM samples using fresh-frozen mouse bones and a column-based RNA extraction method. Thorough lysis (1 min pipetting in the cap) is essential for good RNA yield. The effect of the RNA extraction method on the RNA yield and integrity obtained after LCM was investigated using several different RNA isolation kits according to the manufacturers’ instructions. However, better quality RNAs and increased quantity were obtained with the kit used in the optimized protocol. In the pilot study, microdissected tissue regions of 1 mm² final area were cut, and it was possible to isolate approximately 8.5 ng of RNAs using 8-µm-thick bone sections (approximately 800 pg/µL). Typically, osteoblasts, osteocytes, and bone lining cells were captured in 2−3 sections per sample.

Direct comparisons between different LCM instruments are scarce. In one study, two common LCM instruments were tested, which differ in the type of laser used (UV and infrared [IR], respectively) and method of capturing tissue (adhesive isolation cap vs. caps coated with transparent thermoplastic film, respectively). It was found that for thinly-sectioned fresh-frozen mouse brain sections, the IR system resulted in modestly higher RNA quality⁷. In the present study, two LCM systems that allow contact-free sample collection were compared. In one system, the microdissected sample falls by gravity into the cap placed just below²⁹. Another system uses a de-focused laser pulse, which catapults the material into the overhanging cap³⁰. For fresh frozen bones, higher RNA amounts and RIN values were obtained when gravity was used for tissue capturing. In addition, the catapulting technology is not compatible with the “sandwich” composed of adhesive film, cryo-section, and PET membrane, due to the additional weight of the PET membrane.

LCM requires physical access to the tissue surface. Therefore, mounting and glass covering of the specimen is inapplicable, with the consequence of impaired visualization of morphology. To overcome this limitation, a fluid cover medium was developed³¹. Alternatively, 10−15 µL of ethanol can be added directly on the tissue section, enabling better morphological analysis before evaporation². In the present study, frozen bone sections were cut on an adhesive film and, before the sample was allowed to dry completely, it was sandwiched between the PET membrane and the film. This “sandwich” method gave good morphological contrast and specific bone cells were clearly identified.