Background
Considered for decades a passive process, calcification is now seen as an active mechanism sharing many similarities with embryonic bone formation [
1,
2]. The most recent mechanism proposed to elucidate arterial calcification is the possible role of resident or circulating stem cells that differentiate into chondro-osteogenic cells [
3]. Nowadays, the clinical impact of arterial calcification is still unclear [
4]; indeed the extent of calcification is associated with either a good or bad prognosis. For example, in the coronary arteries small calcifications increase the risk of plaque rupture whereas bigger deposits seem to stabilize the plaque [
5]. On the other hand, calcified carotid plaques are considered a low athero-embolic risk [
6]. In our recent study, we observed that plaques with massive calcifications showed the same incidence of histological complications but without influencing clinical symptomatology [
7].
A big issue in analysing calcified plaques is the deterioration of the carotid tissues due to the strong pre-treatment used to dissolve minerals (ethylenediaminetetraacetic acid (EDTA) or chloride acid (HCl) prior to histopathological processing [
8].
Previous studies attempted to process calcified tissue with alternative decalcification solutions or treatments as ultrasound, however the major issue was maintaining the whole cellular morphology together with an intact DNA content [
9,
10]; indeed the use of strong acids hydrolyses the DNA molecule. We hereby apply a parallel approach to standard histology methods to preserve and study the cellular content of heavily calcified plaques. Our aim was to validate and optimized the techniques for the study of calcified arterial tissues. Additionally to calcified plaques, the technique was also tested on a wide series of tissues from soft to hard (bone).
Methods
Tissue sampling
All samples were obtained under a protocol approved in advance by the Massachusetts Institute of Technology, Committee on the Use of Humans as Experimental Subjects. No persons, including minors/children, were enrolled in MIT studies. Only surgical discards from patients anonymous to MIT researchers were received from collaborating hospitals. These comprised fixed tissues from a variety of organs. The Massachusetts Institute of Technology IRB reviewed and approved the several collaborative relationships specifically indicating that the arrangements were in the US NIH “exempt” category for the requirement for informed consent (COUHES approval number 0804002679). Anonymous surgical discards were fixed immediately upon surgical removal in Carnoy’s solution as described and stored under refrigeration in 70 % ethanol [
11]. A total of 41 tissue samples were finally collected from 26 cases, 16 fetuses and 10 adults.
Experimental parameters were optimized for two different types of samples, soft and hard. Hard adult samples were defined as “
undissectable” with standard procedures without previous decalcification, commonly in Osteodec, Bio-optical (EDTA, calcium’s chelates, 40 to 60 min dependent on the tissue dimension). Of the 41 tissue samples retrieved, 22 were hard and 19 were soft, sorted in:
-
8 hard fetal samples; 3 femoral, 1 nasal bone, 1 elbow, 1 hip bone and 1 knee bone, and 1 chondroid tissue.
-
9 fetal soft samples; 3 guts, 3 brains, 2 muscles, and 1 eye retinae.
-
14 hard adult samples; 13 calcified carotid plaques and 1 calcified abdominal aortic aneurism.
-
10 soft adult samples; 9 uncalcified carotid plaques and 1 uncalcified abdominal aortic aneurism.
Spreading method
All 41 tissue samples were fixed in cold Carnoy at 4 °C (3 ml Ethanol 100 % and 1 ml glacial acetic acid per <0.25 cm
2 within 30 min from surgical removal. Carnoy solution was changed after 60 and 120 min for a total of 3 hours fixation. Sample were then stored at -20 °C in 70 % ethanol [
12]. Subsequent enzymatic tissue treatments with collagenase, type II (Clostridium histolyticum, ≥200 collagenase units/mg, Calbiochem, Merck Chemicals Ltd, UK) were tested with different incubation times and concentrations at 37 °C to define satisfactory conditions for various tissue types (Table
1). The most notable requirement is the need for up to 24 h of collagenase treatment to achieve unfolding of the hard tissues.
Table 1
Specifications for satisfactory spreading of soft versus calcified tissue samples in adult vascular tissues and fetal samples
Adult hard | 24 h | 10U/1 ml | 30 min | 0.1 mm |
Adult soft | 18 h | 10U/1.5 ml | 10 min | 0.3 mm |
Fetal hard | 12 h | 10U/1.5 ml | 20 min | 0.2 mm |
Fetal soft | 1–4 h | 10U/1.5 ml | 10 min | 0.3 mm |
Collagenase-treated tissue samples were subsequently macerated in 500 ul of 45 % acetic acid (Table
1). Macerated tissue samples were minced with surgical scissors creating pieces no greater than 0.3 mm in length. Each piece was transferred to a slide in a drop of 45 % acetic acid. To spread the macerated tissue into a monolayer, several thicknesses of filter paper were placed on the coverslip and a tweezer handle was moved steadily in one direction with a delicate pressure. Intermittent microscopic examination during the spreading process was used to determine when satisfactory spreading had been achieved. Increased pressure on the stacked filter paper was required for the harder bone-containing samples. Cover slips were removed after direct immersion in liquid nitrogen, and slides were air dried for 1 h. Spread slides were stored at 4 °C, in the dark, and subsequently used for different forms of analyses: Von Kossa, Feulgen-Giemsa staining, immunohistochemistry, Feulgen reaction for DNA quantification and RNA extraction.
Histochemistry
Histochemistry included Von Kossa staining for the identification of calcified matrix: sections were hydrated, treated with 1.5 % silver nitrate (AgNO3) solution and placed in front of a 100 W lamp for 1 h. Sections were therefore rinsed in distilled H2O (dH2O), counterstained with nuclear-fast Red, dehydrated and mounted.
The Feulgen-Giemsa staining was used to test the preservation of the nuclei after the collagenase digestion and the subsequent DNA quantification. Briefly, the slides were placed in a Coplin jar filled with HCl 1 N at 60 °C for 8 min, for partial hydrolysis of the macromolecules and DNA depurination. Hydrolysis was stopped by washing the slides in cold dH2O. After carefully drying, sections were placed in Schiff’s reagent (#3952016, Sigma-Aldrich, St. Louis MO, US) for 1 h at room temperature, washed twice with 2x standard saline citrate (trisodium citrate 8.8 g/L, sodium chloride 17.5 g/L), washed again with dH2O and thereafter counterstained with 1 % Giemsa solution (#GS500, Sigma-Aldrich, St. Louis MO, US) for 5 min. Slides were dehydrated in graded step of alcohol and mounted with Canada Balsam (Sigma- Aldrich C1795).
Feulgen DNA quantification
Feulgen densitometry relies on the premise that the amount of stain bound to DNA is proportional to the amount of DNA present in the nucleus with a 1:1 ratio. Feulgen reaction is specific for DNA and does not stain RNA [
11,
13].
DNA quantification using Feulgen reaction (without Giemsa) was carried out in 3 fetal and 4 adult samples, each equally distributed as hard and soft areas (Table
2, raw data are included in the Additional file
1). Nuclei of chicken red blood cells (cRBC, #IC05-0810, Innovative Research, Novi, MI, US) were used as an internal standard for DNA quantification (one nucleus of RBC contains a diploid quantity of DNA equivalent to 2C = 2.5 pg). Slides are Feulgen stained in groups (25 to 100 slides) with the same reagents, lots, conditions and standards. Image acquisition and segmentation were performed with the Carl Zeiss Vision AxioVision software: these are the critical steps for a precise DNA quantification. All steps are presented in details in the Additional file
2. The measure features for each nucleus are the densitometric sum, the densitometric area and the densitometric standard deviation. The quantity of stain is determined based on the absorbance (Optical Density OD) evaluated by the transmitted light (T).
Table 2
Descriptive statistics for Feulgen-DNA content and mean nuclear area
1 | Adult | Hard | 137 | 0.4 | -0.6 | 6.0 | 3.1 | 72.7 | 36.9 | 50.8 % | 3.2 | 66.5–79.0 |
2 | Adult | Soft | 306 | 1.1 | 1.8 | 4.3 | 2.1 | 57.5 | 25.9 | 45.2 % | 1.5 | 54.6–60.4 |
3 | Adult | Soft | 48 | 1.4 | 2.8 | 5.6 | 2.2 | 66.1 | 24.7 | 36.8 % | 3.5 | 59.0–73.1 |
4 | Adult | Soft | 92 | 0.6 | 0.5 | 7.0 | 2.8 | 68.4 | 27. | 40.4 % | 2.9 | 62.7–74.1 |
5 | Fetus | Hard | 145 | 1.6 | 2.7 | 3.1 | 1.6 | 46.9 | 25.3 | 53.9 % | 2.1 | 42.7–51.0 |
6 | Fetus | Hard | 46 | 1.5 | 1.7 | 1.7 | 0.9 | 31.4 | 16.8 | 53.5 % | 2.5 | 26.4–36.3 |
7 | Fetus | Soft | 120 | 1.1 | 0.5 | 4.8 | 2.4 | 63.8 | 32.7 | 51.2 % | 2.9 | 57.9–69.7 |
Total | | | 894 | | | | 59.2 | 30.2 | 51.1 % | 1.0 | 57.2–61.2 |
cRBC | Control | | 339 | -0.1 | 1.2 | 2.5a
| 25.9 | 5.9 | 22.6 % | 0.3 | 25.3–26.5 |
$$ OD\kern0.37em =\kern0.37em lo{g}_{10}\left( 1/T\right) $$
(1)
As nuclear DNA stain is heterogeneous, a single point (OD) would not be representative. Thus it is necessary to evaluate the whole nucleus as the integrated optical density (IOD):
$$ IOD={\displaystyle {\sum}_{i=1}^nlo{g}_{10}\left( 1/Ti\right)} $$
(2)
The scale of each image was 95.04 pixel/ μm2. Additionally to nuclear IOD, the measured features were the nuclear area (μm2) and the total number of nuclei evaluated.
Calculation of genome size DNA
To convert IOD in genome size in pg we used the primary standard of RBC (1 C = 1.25 pg). Smears were processed with the same way as samples and were included in each experiment. To calculate DNA content a standard curve (IOD
vs known C-value) was generated and used to verify that the stain was accurate:
$$ \mathrm{D}\mathrm{N}\mathrm{A}\ \mathrm{pg}\ \mathrm{per}\ \mathrm{nuclei}\ \mathrm{in}\ \mathrm{sample} = 2.5\ \mathrm{pg}\ /\ \mathrm{mean}\ \mathrm{I}\mathrm{O}\mathrm{D}\ \mathrm{standard}\ \mathrm{x}\ \mathrm{I}\mathrm{O}\mathrm{D}\ \mathrm{sample} $$
Immunohistochemistry technique on Carnoy-fixed spread tissue
Table
3 summarizes the antibody characteristics and the blocking/antigen unmasking procedures used. Indeed, due to the Carnoy fixation, some modifications in the antigen retrieval procedures were optimized.
Table 3
Technical characteristics of the antibodies used for imunohistochemistry
Ki-67 | Mouse Monoclonal | Perichromosomal layer protein. Identifies cells in G1-S-G2-M phases. | Dako A/S, Copenhagen, Denmark | 10q26.2 MIB-1 | 1:100 | 20 min | Citrate Buffer, heat mediated (4 cycles 750 W, 5 min each) |
Osteocalcin | Rabbit Polyclonal | Bone specific protein. Synthesized by osteoblasts, accumulates in the bone matrix. | Millipore USA | 1p22 | 1:100 | 30 min | Tryton 0.5 % 4 min |
Desmin | Mouse Monoclonal | Muscle specific intermediate filament type III protein. | Millipore USA | 2q35 D33 | 1:100 | 20 min | Citrate Buffer, heat mediated (4 cycles 750 W, 5 min each) |
Vimentin | Mouse Monoclonal | Intermediate filament type III protein expressed in mesenchymal cells. | Millipore USA | 10p13 V9 | 1:80 | 20 min | Not required |
In addition to Von Kossa staining, immunostaining for Ostecalcin (OCN) was used to spot the main components of the bone matrix as well as the osteoblasts.
Samples were immunostained for Ki-67 antigen in order to assess the chromatin preservation after the procedure and the cell proliferation. Finally, in order to verify the preservation of specific cytoplasmic antigens, Desmin and Vimentin immunostaining was performed. Carnoy-fixed spread tissue was rehydrated through graded steps of ethanol absolute (100 %, Methanol 3 % H2O2, 95 %, 70 %, H2O). Endogenous peroxidase activity was neutralized with 3 % H2O2 in absolute methanol at room temperature (rt), in the dark. Antigen-antibody reaction was developed with the NovoLink® Polymer Detection Kit (Novocastra, Newcastle, UK). To reduce nonspecific antibody bindings, hydrophobic binding sites were blocked with Casein 0.4 % (Novolink Protein Block). Sections were incubated 1 h in a wet chamber at rt with the primary antibodies.
To detect primary antibody bound to tissue, sections were incubated with NovoLink® Polymer (8 mg/L) for 30 min at rt. Spreadings were further incubated with chromogen/substrate, 3,3 - diaminobenzidine NovoLink® (DAB pre-diluted 1:20), 30 s for fetal tissue and 2 min for adult tissue, at rt, Cell nuclei were counterstained with Mayer’s haematoxylin (Sigma Chemicals). Fetal tissue was used as positive controls [
12]. As negative controls primary antibodies were omitted. Washing steps were all performed with PBS. After dehydration in graded steps of ethanol, samples were mounted onto glass slides using Canada Balsam (Sigma-Aldrich C1795). The sections were observed under a light microscope (Axio Imager Z1 Zeiss, Germany) connected with a charge-coupled device (CCD) camera.
RNA content: RT-PCR
The RNA was extracted from Carnoy fixed spread sections (monolayer). A commercial kit was used for RNA extraction (RNeasy® FFPE, Cat No. 73504, QIAGEN GmbH, Hilden, Germany); the protocol was slightly modified to increase RNA yield and purity. Spread tissue was recovered from the slides with a scalpel by adding 5 μl of Proteinase K buffer (PKD), under RNase free conditions. Retrieved tissues were combined into a single nuclease-free screw cap tube, dissolved in 245 μl of PKD and followed by centrifugation for 1 min at 10 000 rpm. For cell lysis, we added 20 μL of protease K; the mixture was incubated for 30 min at 56 °C, mixed at 850 rpm, and then incubated again at 80 °C for 15 min. The protocol was followed as stated by manufacturer’s instruction. Finally the RNA was eluted in 22 μl of RNAse free-water, heated at 65 °C for 5 min in order to be denatured and to inactivate RNases, and stored at -80 °C until used. RNA integrity and concentration were measured by using an ND-1000 spectrophotometer (NanoDrop, Fisher Thermo, Wilmington, DE, USA). The ratio of the readings at 260 nm and 280 nm (A260/A280) provides an estimate of the purity of RNA with respect to the contaminants that absorb in the UV spectrum, such as protein. Pure RNA has an A260/A280 ratio of 1.9–2.1.
RT-PCR assay
To assess the possibility of extracting RNA from Carnoy-fixed spread tissue as template for retrotranscription and gene expression, we amplified by RT-PCR the housekeeping gene beta-glucuronidase (GUSB) for both femur bone and intestine. GUSB is considered an housekeeping gene required for the maintenance of basic cellular function and expressed in all cells [
14].
We also amplified the transcription factor Osterix (OSX) essential for osteoblast differentiation and bone formation [
15] and the carcinoembryonic antigen-related cell adhesion molecule 5 (CEACAM 5) normally produced in gastrointestinal tissue during fetal development [
16].
The reverse transcription assay was performed using 2 μg of total RNA per 20 μl of mix, following the manufacturer’s protocol (High capacity cDNA Archive kit, Applied Biosystem). The cDNA was stored at -20 °C until RT-PCR was performed. RT-PCR was carried out following MasterMix TaqMan® Protocol (TaqMan Univ PCR MasterMix, Applied Biosystems). Two μl of neat cDNA were amplified using specific probes for GUSB (NM_000181.3), CEACAM5 (NM_004363.2) and Osterix (NM_001173467.1) in the RT-PCR mix (TaqMan® Gene Expression Assay, Applied Biosystems, respective ID assay: Hs00939627_m1, Hs00237075_m1, Hs00541729_m1). Reactions were run on ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). The cycling conditions were performed as follows: 10 min at 95 °C, 45 cycles at 95 °C for 15 s and 60 °C for 60 s. Each assay was carried out in triplicate and the transcription level was normalized using GUSB as a reference gene. The threshold was set at 0.2 in order to be in the exponential phase. The values were expressed as DCT (= > CT Target − CT GUSB).
Statistics
To present the basic features of the F-DNA analysis we used descriptive statistics: mean, standard deviation, coefficient of variation, standard error, confidence interval stated at 95 %, skewness and kurtosis. Differences in DNA content in pg and fluctuations of DNA (expressed as DNA content variation, DCV) [
17] between samples were analysed with Kruskal-Wallis (non parametric-unmatched) and Dunn’s post tests. The value of
P < 0.05 was considered significant.
Discussion
The spreading technique showed a good reproducibility and accuracy in the evaluation of cell protein expression and single-cell DNA content in calcified carotid arteries. The IHC markers were selected in order to check the integrity of the different cellular compartments: nucleus (Ki-67) and cytoplasm (Desmin, Vimentin, Osteocalcin). To verify the reliability of the techniques we used fetal tissues sample as positive control since this tissue was already validated [
12]. As standard for the Feulgen DNA quantification we chose cRBC, since they are convenient, easy to isolate and they have a constant known DNA content [
18,
19]. Our results confirmed the reliability of the standard as cRBC genome size was proportional to the area of the cells and CV% was low (23 %). The digestion times with collagenase requested by the spreading technique varied considerably according to the different tissues (see Table
1). Nevertheless, the adult and fetal DNA content distribution curves were all Gaussian with the same trend, showing that collagenase did not influence the DNA content, regardless the digestion times. The Carnoy-fixed spread tissue also allows to extract preserved RNA, without any contaminations by proteins or phenols, and the sigmoid-shaped curves obtained at RT-PCR showed a typical PCR gene amplification.
The technique could also give new information regarding the study of calcification and osteogenesis. Usually DNA quantity is evaluated in the whole tissue; unfortunately one disadvantage is the loss of information on the single cell profile. Moreover standard decalcification procedures degrade the DNA content in calcified arteries. First this protocol avoids the use of formalin to fix the arteries and also avoids the use of strong acid to decalcify the arteries. This allows extracting directly from the slides of spread tissue the RNA and DNA and evaluates the gene expression in calcified tissues from a very small quantity of sample. Moreover the protocol preserved the morphology of the nucleus and allowed the visualization of the nuclear organization. As already reported, some cases of atherosclerotic carotid tissues contain bone lacunae-like mature structure in development with lamellar bone. In these cases, the protocol is of great help to open the “core” of the bone structure and visualize the cellular content. It is very difficult with the standard histological techniques to obtain a monolayer of cells. Surprisingly our data revealed and confirmed that the calcified part of arteries contains a heterogeneous population of cells composed by smooth muscle cells, osteoclasts-like cells and osteoblasts. Moreover the IHC revealed that some of the cells, composing the core of the calcification, were actively dividing. The observation is in line with other studies suggesting that arterial calcification is an active process [
5].
Feulgen on calcified arteries spread allows spotting the peculiarity of the single cell. For example, we observed that two cases of fetal bone (#5 and #6) showed the lowest DNA absolute quantities. By reviewing the literature only a few groups spotted this paradox in DNA content. Solari et al. reported that osteoclasts undergo amitotic division by a budding mechanism in vitro [
20]. Then Sundaram in 2004 demonstrated the existence of a set of cells dividing through neosis [
21] and in 2014, Thilly et al. showed that the cells in syncytia divide by an amitotic mechanism with a dsRNA/DNA intermediate [
22]. As Feulgen only stains DNA and not RNA [
11,
13] we can hypothesize that the nuclei in the syncytia of fetal bone tissue undergo dsRNA/DNA intermediate during replication, thus explaining the lower mean quantity of DNA (1 pg to 3 pg) and the negativity to Ki-67.
Of fluctuations of DNA, defined as an increased range of DNA content, were observed both in the cell populations and within the single cells. The lowest CV variation was found in the healthy soft sample (#3) characterized by a homogenous population of SMC. Conversely the major CV in adult tissue was counted in the calcified tissue. This is likely to reflect a heterogeneous cell population in the fetal and adult pathologic tissues, compared to the cytological homogeneity of the smooth muscle cells in healthy arteries.
In conclusion, thanks to the preservation of the nuclear morphology the technique allowed to disclose osteoclasts-like cells and others cells type with a bell-shaped nucleus, recognised as metakaryotic cells [
23]. These particular cells are very difficult to identify due to the loss of nuclear morphology when using strong acid for decalcification. Mainly studied in the progression of tumours [
12], we can hypothesis that these cells are also a key player in atherosclerosis [
24].
Conclusion
The spread gently preserves the original morphology of each single cell in the calcified vascular tissues. The presented protocol will be used to study the calcification or osteogenic process occurring during atherosclerosis and to evaluate the single cell DNA content on a large series of calcified plaques. The opportunity to extract and amplify intact nucleic acid from the cells of heavily calcified arteries opens the possibility to study calcification on a larger scale. This is not doable when arteries are treated with decalcification reagents.
Abbreviations
CEACAM 5, carcinoembryonic antigen-related cell adhesion molecule 5; cRBC, chicken red blood cells; CT, threshold cycle; CV, content variation; DCV, DNA content variation; EDTA, ethylenediaminetetraacetic acid; F-DNA, Feulgen DNA; GUSB, beta-glucuronidase; HCl, chloride acid; IOD, the integrated optical density; OCN, osteoclacin; OD, optical density; OSX, osterix; PKD, proteinase K buffer; RT, room temperature
Acknowledgements
Research at MIT (SF, EVG, WGT) was supported by a grant from United Therapeutics Corporation (UTC) (
www.unither.com), Silver Springs Md, USA, to study metakaryotic biology under MIT No. 6915390.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.