http://informahealthcare.com/cts ISSN: 0300-8207 (print), 1607-8438 (electronic) Connect Tissue Res, 2014; 55(S1): 9–14 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/03008207.2014.923876
Peritubular dentin, a highly mineralized, non-collagenous, component of dentin: isolation and capture by laser microdissection Jason R. Dorvee, Alix Deymier-Black, Lauren Gerkowicz, and Arthur Veis Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
Abstract
Keywords
We demonstrate the capability and technique to perform microdissection and isolation of select regions of untreated, mineralized dentin using laser capture. Dentin is a complex, non-homogeneous tissue comprised of a mineralized collagenous matrix (intertubular dentin [ITD]), odontoblastic processes (ODPs), a void space (tubules) that forms within the ITD left behind by the retraction of ODPs during dentin maturation, and a highly mineralized non-collagenous component that exists at the interface between the tubules and ITD known as peritubular dentin (PTD). PTD forms as the dentin matures. The ODPs retract toward the direction of the pulp; leaving very little PTD at either the DEJ or near the pulp. Statistical analysis of thin cross-sections of coronal bovine dentin imaged by light microscopy reveal that the area occupied by PTD 450%. To examine the nature of PTD and its relation to both the tubules and ITD, we devised a series of steps to carefully prepare sections of coronal bovine dentin so that areas of the dentin tissue could be cut and isolated for further analysis. We demonstrate that it is possible to selectively isolate targeted regions of dentin for analysis and that high resolution analysis of such sections can be performed using electron microscopy. Results show that the mineralized PTD has a different texture than mineralized ITD and that there is a distinct boundary between the PTD and the ITD. Selective isolation of mineralized tissue components for further analytical study opens the door for the investigation of similar enigmatic mineralized structures.
Bovine, ITD, molar, osteocytes, PTD, texture
Introduction Polarized, mature, odontoblasts (OD) deposit a collagenous dentin matrix at the dentin-enamel junction (DEJ). The forming intertubular dentin matrix (ITD) pushes the main OD cell body away from the DEJ, leaving behind a thin connecting cell process attached to the DEJ as the OD moves in the direction of the pulp chamber. The process eventually retracts partially in the pulpal direction, leaving a tubule surrounded by collagen. Thus, the collagenous dentin is penetrated by a dense network of tubules, connected to the odontoblast layer, and partially filled by the OD processes. The perimeter of the tubule space becomes filled with a material called peritubular dentin (PTD), a non-collagenous component of the dentin tissue that is hyper-mineralized, with respect to the ITD. A number of questions arise with regard to this enigmatic, mineralized, non-collagenous, biological material: what is the
Correspondence: Prof. Arthur Veis, Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, 303 E. Chicago Avenue, Chicago, IL 60611, USA. Tel: (312) 503-1355. Fax: (312) 503-2544. E-mail: [email protected]
History Received 8 November 2013 Revised 28 March 2014 Accepted 7 April 2014 Published online 11 August 2014
distribution of PTD in bovine dentin, why is it produced in the dentin tubules, what is its role, what is its composition, how does the structure form? These questions extend beyond the architecture of the tooth. It is well known that there is a high density of tubules in bovine dentin as had been described and established in textbooks (1). Orban also points out that there exists ‘‘a thin organic membrane, high in gucosamine glycan and similar to the lining of lacunae in cartilage and bone,’’ (1) which separates PTD from the ITD. Thus the significance of the nature of and function of PTD is not isolated to the tooth, the importance of the question ‘‘what is PTD’’ extends to the structure and function of osteocytes in bone. In answer to these questions about PTD, we begin with an examination of the distribution and percent area of PTD found in dentin. Using the information from established textbooks (1), we make preliminary calculations of the assumed distribution of PTD in bovine dentin. We then compare these calculated results to additional calculations made using data from more recent studies of dentin cross-sections (2–4). This information provides a general and assumed picture of PTD distribution within dentin. To build a more complete picture we randomly select sections of bovine dentin taken from randomly selected bovine molars and measure the distribution and percentage area of PTD in cross-sections of
10
J. R. Dorvee et al.
bovine dentin. This analysis yields an empirical rather than assumed, picture of PTD distribution in dentin. To begin to answer the remaining questions we develop a methodology that can be used to probe the composition of PTD. In order to determine the composition and structure of PTD one critical step is to first isolate small sections of dentin, which we know to have PTD, in order to study this hypermineralized, non-collagenous, material as compared to the ITD matrix. Our strategy is to use laser microdissection to target, isolate, and capture areas of dentin tissue that contain PTD and make those sections available for analysis. As an example, we demonstrate how the isolated sections of dentin can be analyzed with tools such as electron microscopy.
Connect Tissue Res, 2014; 55(S1): 9–14
This same procedure was used separately to determine the percentage area of the tubules after the inside of each tubule was individually selected and blacked out, this created a baseline value of 0 for the pixels within the tubule (Supplementary Figure S1a), the inflection in the profile plot that deviated from this value indicated the edge of the tubule (Supplementary Figure S1b). The results of the percentage of cross-sectional area for each of these components from examining four randomly selected sections using this pixel threshold technique were: 2.93 ± 0.45% tubules, 43.4 ± 5.88% ITD and 53.68 ± 5.79% PTD for the midsection of coronal bovine dentin. SEM analysis
Results
The presence of tubules in the coronal dentin of a molar from an 18 mo. old Black Angus cow was confirmed through the inspection of polished dentin samples taken from just above the cervical line of the molar and examined using an FE-SEM (Supplementary Figure S2).
Preliminary calculations to estimate the area of the components of dentin
Laser microdissection and capture
Methods For methodology please see Supplementary materials.
Based on the sample calculations described in the methods (Supplementary materials) and assuming a perfectly circular shape, a single tubule would occupy an area of 3.14 mm2 or 3.14 106 mm2, at 70 000 tubules/mm2 (1) the total area occupied by the tubules is 0.22 mm2 or 21.98%. Assuming a wall thickness for the peritubular dentin of 0.75 mm (1) the entire tubule-PTD complex would have an area of 9.62 mm2 or 9.62 106 mm2, removing the percentage area of the tubules from this we obtain a value of 6.48 106 mm2, and a total area of PTD at 45.36%. Subsequently, the remaining crosssectional area of 32.66% is ITD. These are our target values that we will use to compare to empirical data. Area determination by image analysis To make the determination of total area occupied by individual dentin components using real samples, randomly selected pieces of coronal bovine dentin, pre-cleaned and lyophilized, were encased in Epon and sectioned from the predentin–dentin junction to the DEJ, in 1-mm thick sections. At a thickness of 1 mm, these sections were thin enough to be imaged on a light microscope. Figure 1 illustrates an example of how sections were analyzed using the values of light attenuation as it passed through the sample. As seen in Figure 1 the areas of higher contrast registered a unique pixel value when measured in terms of light and dark (255 ¼ most white, 0 ¼ most black). A region of interest selected in ImageJ, shown as a red solid line in Figure 1(b), was plotted by pixel value as a function of distance as described in the ‘‘Methods’’ section (Supplementary materials). The point of inflection where the pixel values begin to plateau indicate the edge of this contrast difference and delineates the pixel threshold value for this contrast difference marked by a dashed blue line in both the profile plot and on the corresponding location of the image (Figure 1b). This value was then used to threshold the image, all values less (darker) than this value were blocked out and the resulting area counted using the particle analyzer tool in ImageJ (Figure 1c).
Making 1-mm thick cross-sections from the same bovine (examined earlier with FE-SEM) tooth and recording the location of each section, we were able to maintain registry of the tubule locations within the tooth. The section shown in Figure 2(a) is 1048 mm from predentin–dentin junction. Direct capture of the isolated dentin sections on to a TEM grid (Figure 2a) was accomplished by placing the grid next to the dentin cross-section on the PALM laser capture slide. Samples were cut and catapulted without the capture tube in place. Samples are ballisticaly catapulted from the sample and captured on the neighboring TEM grid (Figure 2a). Once captured, the isolated dentin sections are examined with electron microscopy (Figure 2b and c). Further registry of the location of the captured tubules within a section is tracked by making laser cuts of the dentin using distinct, easily identifiable shapes. Figure 2(ai) shows a series of geometrically cut shapes made from the dentin section such as circles, triangles, L’s, and trapezoids. Figure 2(aii) shows the capture of a chevron, a star, and a rectangle from the dentin section shown in Figure 2(ai). The captured dentin sections found on the TEM grid have been correlated with the original dentin section using image overlays, and confirm the original location of each captured section. Examination of these samples by TEM show material of low-electron density on the inside edge of cut tubules (Figure 2b), which do not show the characteristic banding of mineralized collagen. High resolution TE images from an STEM, of an intact tubule show an asymmetric region of high electron density immediately around the tubule which is surrounded by medium electron dense material (Figure 2c). The high and medium density materials are separated by small patches of low-density material/gaps (Figure 2ci–ii) and further shown by Z-contrast imaging (Figure 2ciii).
Discussion Calculations using values from the literature (1) indicate that the majority component (by cross-sectional area) of bovine
DOI: 10.3109/03008207.2014.923876
Isolation of peritubular dentin by laser microdissection
11
Figure 1. An example of the image analysis performed on the cross-sectional area occupied by peritubular dentin. A light microscope image taken at 100 (a) is analyzed with a profile plot (b) along a line trace (red solid line) drawn from one side of tubule containing PTD to the other. From this profile plot, a threshold for the maximum pixel value for PTD is determined (blue dashed line) (b) and that value is used to create a black and white image delineating all material that is PTD from all material that is not PTD (c). This image is then processed with ImageJ using the Analyze Particle feature in ImageJ (National Institute of Health) (Size (pixel^2)¼ 0–Infinity, Circularity ¼ 0.00–1.00, including holes), the value of the tubule area was then subtracted from this value to yield the percentage of area covered by PTD.
coronal dentin tissue is peritubular dentin at 45.36%, followed by the collagenous ITD at 32.66%, and dentin tubules at 21.98%. However the values used, had assumed an average uniformity in the statistical values across the calculated crosssection. Additionally, the calculations were made assuming both uniform and simple geometries for both tubules and peritubular dentin. It is easy to conclude that using a tubule density value of 70 000/mm2 is an over estimation, especially when more recent articles have shown that the density of tubules in the crowns of both human and bovine molars can be as low as 17 000/mm2 to 20 000/mm2 (4,5). Using such low values of tubule density with the same assumptions with regard to tubule and peritubule diameters, the values of peritubular percentage drop to 11–13%, the values of tubule percentage drop to 5–6%, and the ITD becomes the majority component at 81–84%. Other works have both measured and calculated the density of tubules within the middle coronal
dentin to be 32 000/mm2 to 38 000/mm2 (2,3). Again, based on our assumptions these values would yield 20–25% PTD, 10–12% tubules and 63–70% ITD. All of these variations in reported tubule density lead to large discrepancies in the assumed PTD distribution in dentin and these values can vary even more depending on the area of the tooth one chooses to consider. Rather than rely on these rough calculations we randomly selected sections of coronal dentin from bovine molars and measured the PTD content and distribution by area. As shown in Figures 1 and 2, and Supplementary Figures S1 and S2, the distribution and the shape of both the dentin tubules and peritubular dentin are not uniform. Examination of 1-mm thick sections of dentin using light microscopy shows the irregularity and asymmetry of the peritubluar dentin structures. As light passed through the thin microtomed samples, regions of greater density or differences in index of refraction
12
J. R. Dorvee et al.
Connect Tissue Res, 2014; 55(S1): 9–14
Figure 2. Demonstration of the laser microdissection direct capture technique used on cross-sections of bovine dentin. A light microscope image at 20 shows the distinct shapes that can be cut from dentin (1048 mm from the predentin–dentin junction) and serve as fiduciary markers to track the registry of the tubules as they are isolated from the dentin tissue (a). Once samples are captured directly onto TEM grids (a & b [light microscope image far left]) they can be examined with electron microscopy (b). By transmission electron microscopy the inside edges of the dentin tubules appear to have a thin organic film that is not banded as seen in collagen. The tubules shown in 2bi and 2bii–iii are two separate tubules. (c) Further examination of other isolated tubules via high resolution STEM shows a discontinuous transition between PTD and ITD as shown by arrows. (ci–iii.) Closer examination of this transition zone (via TE (cii) and Z-Contrast (ciii)) shows a fibrous region of distinct banding (likely collagen) and regions of high density material that is not banded.
DOI: 10.3109/03008207.2014.923876
attenuated the transmitted light, producing an area of greater contrast than the surrounding material. Once identified, the pixel value for this differing contrast was used to assign a threshold value that could then be used as a division line between material that attenuated light as PTD, and material that did not. Using this distinction, these individual components could be counted using ImageJ, the result for four randomly selected midsections of coronal bovine dentin was that PTD is 53.68 ± 5.79% of the cross-sectional area for these samples. This is an astonishing number, especially since the upper limit of our conservative estimates (45.36% PTD) is well below this measured value. In contrast our conservative upper limit estimate of the percentage area of tubules was 22% while the lower limit estimate was 5% based on values from the literature (1–5), but the measured percentage area tubules was 2.93 ± 0.45%. These discrepancies between the calculated estimations and the measured data point to a flaw in the assumptions made when performing our calculations. This is expected. Pashley made similar calculations for the area occupied by tubules, PTD and ITD, based on the data from Bra¨nnstro¨m and Garberoglio (3,6). In that review, Pashey recognized that both tubule diameter and density of tubules per mm2 changed with depth with respect to either the predentin–dentin junction or the dentin enamel junction. But Pashley assumed that the diameter of the PTD changed proportionally with the changing diameter of the void space (tubules). As a result Pashely grossly overestimated the presence of PTD near the pulp (36.58–66.25% area) and as current data reflects, underestimated the amount of PTD in the remaining dentin (3). Our assumptions, just like Pashley’s are just that, assumptions. At least one of our assumptions can be eliminated by directly assessing the tubule distribution in these random sections. A survey of all four sections reveals that the tubule density is 19 750 ± 1482 per mm2. This value is consistent with the works of Schilke and Dutra-Correa (4,5), using this value with our other assumptions, our calculated values for tubule percentage, and PTD percentage, come up short with respected to the measured data. This data indicates that our assumption of tubule diameters 2 mm was an over estimation while our assumption with regard to the PTD thicknesses of 0.75 mm was an under estimation. In the end this evidence demonstrates that both care and proper justification must be taken when making such calculations and extrapolations. We have shown that our measured data, regardless of calculations based on generalized assumptions, and consistent with values found in the literature (2,5), show that by area, peritubular denitin is the majority component of dentin. Compared to these randomly selected samples, results from further investigation may vary from region to region within the tooth, this is a fact that must be taken into account when considering the structural makeup of dentin. Given this information, we proceeded to devise a method to target and isolate this enigmatic material. The first step to find PTD within an intact tooth was to use established histological techniques to determine the presence of PTD in freshly harvested coronal dentin from a bovine molar. Knowing that PTD is predominantly found in the crown of the tooth, two 2.5-mm thick cross-sections were cut above the cervical line of a single bovine molar. From the section closest to the cervical line two blocks of the tooth
Isolation of peritubular dentin by laser microdissection
13
were excised and polished starting from the enamel side of the tooth and working toward the direction of the pulp. The methodology we developed for this polishing and sample preparation was designed to produce the minimal amount of artifacts when examined with electron microscopy. Imaging of these polished blocks using FE-SEM showed PTD around the dentin tubules (Supplementary Figure S2) as seen in the literature (7). These FE-SEM images show that PTD is not only irregular in shape, but that it also has a different texture than the surrounding ITD and the interior lining is punctuated by holes. The porosity visible through the FE-SEM images may provide for the transport of ions and other components between the OD processes and the dentin matrix (2,7,8). Once we established that there was indeed PTD present above the cervical line in this molar, two additional blocks were excised from the second cross-section taken from this molar; but this time the blocks were encased in Epon and microtomed for laser microdissection. The 1-mm thick slices were placed on a PEN-MembraneSlide 0.17-mm slide with TEM grids surrounding the sample to catch the catapulted sections. We have demonstrated that a laser microdissection system can be used to target and isolate selected regions of dentin tissue. Additionally, we have shown by using uniquely shaped cuts of the sample that isolated tissue fragments can be traced to a region of interest on the targeted dentin section (Figure 2a). Once isolated, the samples were examined with transmission election microscopy. The TEM images reveal a distinct texture to the PTD mineral and an organic component that is not banded as seen in mineralized collagen (Figure 2b). TEM images show that PTD is contains a higher density of material than ITD and that there exists a distinct discontinuous boundary layer between PTD and ITD (Figure 2c). At this point, without mechanical testing it is difficult to suggest any mechanical function for the PTD, but we speculate that due the hypermineralized, non-collagenous structure, and the visible separation that is occurring between that PTD and ITD, that the apparent and distinctly different composition of PTD from ITD would result in different mechanical properties between the two.
Conclusion We have demonstrated that sections of coronal bovine dentin can be target, isolated and captured with laser microdissection. Preliminary results from inspection of isolated sections reveal that the texture of the PTD mineral is different than that of the ITD mineral, and the PTD is porous. It is evident that the PTD collar around the OD processes are higher in mineral content than the surrounding ITD, as shown by many others. However, the PTD is irregular and not uniformly placed within the tubule and such geometry and structure must be considered when making extrapolations with regard to PTD distribution in dentin. We have also determined by examination of randomly selected samples that in these sections PTD is the majority of the cross-sectional area of midsection coronal bovine dentin. Through the STEM examination of isolated dentin sections, we have found a discontinuous boundary between peritubular dentin and ITD. Thus, PTD is quite distinct from ITD, and is not a uniform material.
14
J. R. Dorvee et al.
Connect Tissue Res, 2014; 55(S1): 9–14
Declaration of interest Imaging work was performed at the Northwestern University Cell Imaging Facility: generously supported by CCSG P30 CA060553 awarded to the Robert H Lurie Comprehensive Cancer Center. Microdissection was performed on Zeiss PALM microdissection system purchased with the support of NCRR 1S10RR025624-01. FE-SEM and STEM work was performed at the Northwestern Electron Probe Instrument Center (EPIC), part of Northwestern University’s Atomic and Nanoscale Characterization Experimental Center (NUANCE). This work is funded by National Institute of Health, National Institute for Dental Research R01-DE001374-53.
References 1. Orban BJ, Bhaskar SN. Orban’s oral histology and embryology. 10th ed. St. Louis (MO): C.V. Mosby; 1986:102–10.
2. Mjor IA, Nordahl I. The density and branching of dentinal tubules in human teeth. Arch Oral Biol 1996;41:401–12. 3. Pashley DH. Dentin – a dynamic substrate – a review. Scan Microsc 1989;3:161–76. 4. Dutra-Correa M, Anauate-Netto C, Arana-Chavez VE. Density and diameter of dentinal tubules in etched and non-etched bovine dentine examined by scanning electron microscopy. Arch Oral Biol 2007;52:850–5. 5. Schilke R, Lisson JA, Bauss O, Geurtsen W. Comparison of the number and diameter of dentinal tubules in human and bovine dentine by scanning electron microscopic investigation. Arch Oral Biol 2000;45:355–61. 6. Garberoglio R, Brannstrom M. Scanning electron-microscopic investigation of human dentinal tubules. Arch Oral Biol 1976;21: 355–62. 7. Weiner S, Veis A, Beniash E, Arad T, Dillon JW, Sabsay B, Siddiqui F. Peritubular dentin formation: crystal organization and the macromolecular constituents in human teeth. J Struct Biol 1999; 126:27–41. 8. Bertassoni LE, Stankoska K, Swain MV. Insights into the structure and composition of the peritubular dentin organic matrix and the lamina limitans. Micron 2012;43:229–36.
Supplementary material available online Supplementary Figures S1 and S2, and Methods (contained in one document) Supplementary Figure S1: An example of the image analysis performed on the cross-sectional area occupied by tubules in dentin. Supplementary Figure S2: Sections J11M3R1Aa and J11M3R1Ab, isolated from just above the cervical line of a bovine molar from an 18 mo. Black Angus cow. Supplementary Methods: Preliminary Calculations to Estimate the Area of the Components of Dentin Supplemental material available at informahealthcare.com/cts
Copyright of Connective Tissue Research is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Peritubular dentin, a highly mineralized, non-collagenous, component of dentin: isolation and capture by laser microdissection Jason R. Dorvee, Alix Deymier-Black, Lauren Gerkowicz, and Arthur Veis Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
Abstract
Keywords
We demonstrate the capability and technique to perform microdissection and isolation of select regions of untreated, mineralized dentin using laser capture. Dentin is a complex, non-homogeneous tissue comprised of a mineralized collagenous matrix (intertubular dentin [ITD]), odontoblastic processes (ODPs), a void space (tubules) that forms within the ITD left behind by the retraction of ODPs during dentin maturation, and a highly mineralized non-collagenous component that exists at the interface between the tubules and ITD known as peritubular dentin (PTD). PTD forms as the dentin matures. The ODPs retract toward the direction of the pulp; leaving very little PTD at either the DEJ or near the pulp. Statistical analysis of thin cross-sections of coronal bovine dentin imaged by light microscopy reveal that the area occupied by PTD 450%. To examine the nature of PTD and its relation to both the tubules and ITD, we devised a series of steps to carefully prepare sections of coronal bovine dentin so that areas of the dentin tissue could be cut and isolated for further analysis. We demonstrate that it is possible to selectively isolate targeted regions of dentin for analysis and that high resolution analysis of such sections can be performed using electron microscopy. Results show that the mineralized PTD has a different texture than mineralized ITD and that there is a distinct boundary between the PTD and the ITD. Selective isolation of mineralized tissue components for further analytical study opens the door for the investigation of similar enigmatic mineralized structures.
Bovine, ITD, molar, osteocytes, PTD, texture
Introduction Polarized, mature, odontoblasts (OD) deposit a collagenous dentin matrix at the dentin-enamel junction (DEJ). The forming intertubular dentin matrix (ITD) pushes the main OD cell body away from the DEJ, leaving behind a thin connecting cell process attached to the DEJ as the OD moves in the direction of the pulp chamber. The process eventually retracts partially in the pulpal direction, leaving a tubule surrounded by collagen. Thus, the collagenous dentin is penetrated by a dense network of tubules, connected to the odontoblast layer, and partially filled by the OD processes. The perimeter of the tubule space becomes filled with a material called peritubular dentin (PTD), a non-collagenous component of the dentin tissue that is hyper-mineralized, with respect to the ITD. A number of questions arise with regard to this enigmatic, mineralized, non-collagenous, biological material: what is the
Correspondence: Prof. Arthur Veis, Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, 303 E. Chicago Avenue, Chicago, IL 60611, USA. Tel: (312) 503-1355. Fax: (312) 503-2544. E-mail: [email protected]
History Received 8 November 2013 Revised 28 March 2014 Accepted 7 April 2014 Published online 11 August 2014
distribution of PTD in bovine dentin, why is it produced in the dentin tubules, what is its role, what is its composition, how does the structure form? These questions extend beyond the architecture of the tooth. It is well known that there is a high density of tubules in bovine dentin as had been described and established in textbooks (1). Orban also points out that there exists ‘‘a thin organic membrane, high in gucosamine glycan and similar to the lining of lacunae in cartilage and bone,’’ (1) which separates PTD from the ITD. Thus the significance of the nature of and function of PTD is not isolated to the tooth, the importance of the question ‘‘what is PTD’’ extends to the structure and function of osteocytes in bone. In answer to these questions about PTD, we begin with an examination of the distribution and percent area of PTD found in dentin. Using the information from established textbooks (1), we make preliminary calculations of the assumed distribution of PTD in bovine dentin. We then compare these calculated results to additional calculations made using data from more recent studies of dentin cross-sections (2–4). This information provides a general and assumed picture of PTD distribution within dentin. To build a more complete picture we randomly select sections of bovine dentin taken from randomly selected bovine molars and measure the distribution and percentage area of PTD in cross-sections of
10
J. R. Dorvee et al.
bovine dentin. This analysis yields an empirical rather than assumed, picture of PTD distribution in dentin. To begin to answer the remaining questions we develop a methodology that can be used to probe the composition of PTD. In order to determine the composition and structure of PTD one critical step is to first isolate small sections of dentin, which we know to have PTD, in order to study this hypermineralized, non-collagenous, material as compared to the ITD matrix. Our strategy is to use laser microdissection to target, isolate, and capture areas of dentin tissue that contain PTD and make those sections available for analysis. As an example, we demonstrate how the isolated sections of dentin can be analyzed with tools such as electron microscopy.
Connect Tissue Res, 2014; 55(S1): 9–14
This same procedure was used separately to determine the percentage area of the tubules after the inside of each tubule was individually selected and blacked out, this created a baseline value of 0 for the pixels within the tubule (Supplementary Figure S1a), the inflection in the profile plot that deviated from this value indicated the edge of the tubule (Supplementary Figure S1b). The results of the percentage of cross-sectional area for each of these components from examining four randomly selected sections using this pixel threshold technique were: 2.93 ± 0.45% tubules, 43.4 ± 5.88% ITD and 53.68 ± 5.79% PTD for the midsection of coronal bovine dentin. SEM analysis
Results
The presence of tubules in the coronal dentin of a molar from an 18 mo. old Black Angus cow was confirmed through the inspection of polished dentin samples taken from just above the cervical line of the molar and examined using an FE-SEM (Supplementary Figure S2).
Preliminary calculations to estimate the area of the components of dentin
Laser microdissection and capture
Methods For methodology please see Supplementary materials.
Based on the sample calculations described in the methods (Supplementary materials) and assuming a perfectly circular shape, a single tubule would occupy an area of 3.14 mm2 or 3.14 106 mm2, at 70 000 tubules/mm2 (1) the total area occupied by the tubules is 0.22 mm2 or 21.98%. Assuming a wall thickness for the peritubular dentin of 0.75 mm (1) the entire tubule-PTD complex would have an area of 9.62 mm2 or 9.62 106 mm2, removing the percentage area of the tubules from this we obtain a value of 6.48 106 mm2, and a total area of PTD at 45.36%. Subsequently, the remaining crosssectional area of 32.66% is ITD. These are our target values that we will use to compare to empirical data. Area determination by image analysis To make the determination of total area occupied by individual dentin components using real samples, randomly selected pieces of coronal bovine dentin, pre-cleaned and lyophilized, were encased in Epon and sectioned from the predentin–dentin junction to the DEJ, in 1-mm thick sections. At a thickness of 1 mm, these sections were thin enough to be imaged on a light microscope. Figure 1 illustrates an example of how sections were analyzed using the values of light attenuation as it passed through the sample. As seen in Figure 1 the areas of higher contrast registered a unique pixel value when measured in terms of light and dark (255 ¼ most white, 0 ¼ most black). A region of interest selected in ImageJ, shown as a red solid line in Figure 1(b), was plotted by pixel value as a function of distance as described in the ‘‘Methods’’ section (Supplementary materials). The point of inflection where the pixel values begin to plateau indicate the edge of this contrast difference and delineates the pixel threshold value for this contrast difference marked by a dashed blue line in both the profile plot and on the corresponding location of the image (Figure 1b). This value was then used to threshold the image, all values less (darker) than this value were blocked out and the resulting area counted using the particle analyzer tool in ImageJ (Figure 1c).
Making 1-mm thick cross-sections from the same bovine (examined earlier with FE-SEM) tooth and recording the location of each section, we were able to maintain registry of the tubule locations within the tooth. The section shown in Figure 2(a) is 1048 mm from predentin–dentin junction. Direct capture of the isolated dentin sections on to a TEM grid (Figure 2a) was accomplished by placing the grid next to the dentin cross-section on the PALM laser capture slide. Samples were cut and catapulted without the capture tube in place. Samples are ballisticaly catapulted from the sample and captured on the neighboring TEM grid (Figure 2a). Once captured, the isolated dentin sections are examined with electron microscopy (Figure 2b and c). Further registry of the location of the captured tubules within a section is tracked by making laser cuts of the dentin using distinct, easily identifiable shapes. Figure 2(ai) shows a series of geometrically cut shapes made from the dentin section such as circles, triangles, L’s, and trapezoids. Figure 2(aii) shows the capture of a chevron, a star, and a rectangle from the dentin section shown in Figure 2(ai). The captured dentin sections found on the TEM grid have been correlated with the original dentin section using image overlays, and confirm the original location of each captured section. Examination of these samples by TEM show material of low-electron density on the inside edge of cut tubules (Figure 2b), which do not show the characteristic banding of mineralized collagen. High resolution TE images from an STEM, of an intact tubule show an asymmetric region of high electron density immediately around the tubule which is surrounded by medium electron dense material (Figure 2c). The high and medium density materials are separated by small patches of low-density material/gaps (Figure 2ci–ii) and further shown by Z-contrast imaging (Figure 2ciii).
Discussion Calculations using values from the literature (1) indicate that the majority component (by cross-sectional area) of bovine
DOI: 10.3109/03008207.2014.923876
Isolation of peritubular dentin by laser microdissection
11
Figure 1. An example of the image analysis performed on the cross-sectional area occupied by peritubular dentin. A light microscope image taken at 100 (a) is analyzed with a profile plot (b) along a line trace (red solid line) drawn from one side of tubule containing PTD to the other. From this profile plot, a threshold for the maximum pixel value for PTD is determined (blue dashed line) (b) and that value is used to create a black and white image delineating all material that is PTD from all material that is not PTD (c). This image is then processed with ImageJ using the Analyze Particle feature in ImageJ (National Institute of Health) (Size (pixel^2)¼ 0–Infinity, Circularity ¼ 0.00–1.00, including holes), the value of the tubule area was then subtracted from this value to yield the percentage of area covered by PTD.
coronal dentin tissue is peritubular dentin at 45.36%, followed by the collagenous ITD at 32.66%, and dentin tubules at 21.98%. However the values used, had assumed an average uniformity in the statistical values across the calculated crosssection. Additionally, the calculations were made assuming both uniform and simple geometries for both tubules and peritubular dentin. It is easy to conclude that using a tubule density value of 70 000/mm2 is an over estimation, especially when more recent articles have shown that the density of tubules in the crowns of both human and bovine molars can be as low as 17 000/mm2 to 20 000/mm2 (4,5). Using such low values of tubule density with the same assumptions with regard to tubule and peritubule diameters, the values of peritubular percentage drop to 11–13%, the values of tubule percentage drop to 5–6%, and the ITD becomes the majority component at 81–84%. Other works have both measured and calculated the density of tubules within the middle coronal
dentin to be 32 000/mm2 to 38 000/mm2 (2,3). Again, based on our assumptions these values would yield 20–25% PTD, 10–12% tubules and 63–70% ITD. All of these variations in reported tubule density lead to large discrepancies in the assumed PTD distribution in dentin and these values can vary even more depending on the area of the tooth one chooses to consider. Rather than rely on these rough calculations we randomly selected sections of coronal dentin from bovine molars and measured the PTD content and distribution by area. As shown in Figures 1 and 2, and Supplementary Figures S1 and S2, the distribution and the shape of both the dentin tubules and peritubular dentin are not uniform. Examination of 1-mm thick sections of dentin using light microscopy shows the irregularity and asymmetry of the peritubluar dentin structures. As light passed through the thin microtomed samples, regions of greater density or differences in index of refraction
12
J. R. Dorvee et al.
Connect Tissue Res, 2014; 55(S1): 9–14
Figure 2. Demonstration of the laser microdissection direct capture technique used on cross-sections of bovine dentin. A light microscope image at 20 shows the distinct shapes that can be cut from dentin (1048 mm from the predentin–dentin junction) and serve as fiduciary markers to track the registry of the tubules as they are isolated from the dentin tissue (a). Once samples are captured directly onto TEM grids (a & b [light microscope image far left]) they can be examined with electron microscopy (b). By transmission electron microscopy the inside edges of the dentin tubules appear to have a thin organic film that is not banded as seen in collagen. The tubules shown in 2bi and 2bii–iii are two separate tubules. (c) Further examination of other isolated tubules via high resolution STEM shows a discontinuous transition between PTD and ITD as shown by arrows. (ci–iii.) Closer examination of this transition zone (via TE (cii) and Z-Contrast (ciii)) shows a fibrous region of distinct banding (likely collagen) and regions of high density material that is not banded.
DOI: 10.3109/03008207.2014.923876
attenuated the transmitted light, producing an area of greater contrast than the surrounding material. Once identified, the pixel value for this differing contrast was used to assign a threshold value that could then be used as a division line between material that attenuated light as PTD, and material that did not. Using this distinction, these individual components could be counted using ImageJ, the result for four randomly selected midsections of coronal bovine dentin was that PTD is 53.68 ± 5.79% of the cross-sectional area for these samples. This is an astonishing number, especially since the upper limit of our conservative estimates (45.36% PTD) is well below this measured value. In contrast our conservative upper limit estimate of the percentage area of tubules was 22% while the lower limit estimate was 5% based on values from the literature (1–5), but the measured percentage area tubules was 2.93 ± 0.45%. These discrepancies between the calculated estimations and the measured data point to a flaw in the assumptions made when performing our calculations. This is expected. Pashley made similar calculations for the area occupied by tubules, PTD and ITD, based on the data from Bra¨nnstro¨m and Garberoglio (3,6). In that review, Pashey recognized that both tubule diameter and density of tubules per mm2 changed with depth with respect to either the predentin–dentin junction or the dentin enamel junction. But Pashley assumed that the diameter of the PTD changed proportionally with the changing diameter of the void space (tubules). As a result Pashely grossly overestimated the presence of PTD near the pulp (36.58–66.25% area) and as current data reflects, underestimated the amount of PTD in the remaining dentin (3). Our assumptions, just like Pashley’s are just that, assumptions. At least one of our assumptions can be eliminated by directly assessing the tubule distribution in these random sections. A survey of all four sections reveals that the tubule density is 19 750 ± 1482 per mm2. This value is consistent with the works of Schilke and Dutra-Correa (4,5), using this value with our other assumptions, our calculated values for tubule percentage, and PTD percentage, come up short with respected to the measured data. This data indicates that our assumption of tubule diameters 2 mm was an over estimation while our assumption with regard to the PTD thicknesses of 0.75 mm was an under estimation. In the end this evidence demonstrates that both care and proper justification must be taken when making such calculations and extrapolations. We have shown that our measured data, regardless of calculations based on generalized assumptions, and consistent with values found in the literature (2,5), show that by area, peritubular denitin is the majority component of dentin. Compared to these randomly selected samples, results from further investigation may vary from region to region within the tooth, this is a fact that must be taken into account when considering the structural makeup of dentin. Given this information, we proceeded to devise a method to target and isolate this enigmatic material. The first step to find PTD within an intact tooth was to use established histological techniques to determine the presence of PTD in freshly harvested coronal dentin from a bovine molar. Knowing that PTD is predominantly found in the crown of the tooth, two 2.5-mm thick cross-sections were cut above the cervical line of a single bovine molar. From the section closest to the cervical line two blocks of the tooth
Isolation of peritubular dentin by laser microdissection
13
were excised and polished starting from the enamel side of the tooth and working toward the direction of the pulp. The methodology we developed for this polishing and sample preparation was designed to produce the minimal amount of artifacts when examined with electron microscopy. Imaging of these polished blocks using FE-SEM showed PTD around the dentin tubules (Supplementary Figure S2) as seen in the literature (7). These FE-SEM images show that PTD is not only irregular in shape, but that it also has a different texture than the surrounding ITD and the interior lining is punctuated by holes. The porosity visible through the FE-SEM images may provide for the transport of ions and other components between the OD processes and the dentin matrix (2,7,8). Once we established that there was indeed PTD present above the cervical line in this molar, two additional blocks were excised from the second cross-section taken from this molar; but this time the blocks were encased in Epon and microtomed for laser microdissection. The 1-mm thick slices were placed on a PEN-MembraneSlide 0.17-mm slide with TEM grids surrounding the sample to catch the catapulted sections. We have demonstrated that a laser microdissection system can be used to target and isolate selected regions of dentin tissue. Additionally, we have shown by using uniquely shaped cuts of the sample that isolated tissue fragments can be traced to a region of interest on the targeted dentin section (Figure 2a). Once isolated, the samples were examined with transmission election microscopy. The TEM images reveal a distinct texture to the PTD mineral and an organic component that is not banded as seen in mineralized collagen (Figure 2b). TEM images show that PTD is contains a higher density of material than ITD and that there exists a distinct discontinuous boundary layer between PTD and ITD (Figure 2c). At this point, without mechanical testing it is difficult to suggest any mechanical function for the PTD, but we speculate that due the hypermineralized, non-collagenous structure, and the visible separation that is occurring between that PTD and ITD, that the apparent and distinctly different composition of PTD from ITD would result in different mechanical properties between the two.
Conclusion We have demonstrated that sections of coronal bovine dentin can be target, isolated and captured with laser microdissection. Preliminary results from inspection of isolated sections reveal that the texture of the PTD mineral is different than that of the ITD mineral, and the PTD is porous. It is evident that the PTD collar around the OD processes are higher in mineral content than the surrounding ITD, as shown by many others. However, the PTD is irregular and not uniformly placed within the tubule and such geometry and structure must be considered when making extrapolations with regard to PTD distribution in dentin. We have also determined by examination of randomly selected samples that in these sections PTD is the majority of the cross-sectional area of midsection coronal bovine dentin. Through the STEM examination of isolated dentin sections, we have found a discontinuous boundary between peritubular dentin and ITD. Thus, PTD is quite distinct from ITD, and is not a uniform material.
14
J. R. Dorvee et al.
Connect Tissue Res, 2014; 55(S1): 9–14
Declaration of interest Imaging work was performed at the Northwestern University Cell Imaging Facility: generously supported by CCSG P30 CA060553 awarded to the Robert H Lurie Comprehensive Cancer Center. Microdissection was performed on Zeiss PALM microdissection system purchased with the support of NCRR 1S10RR025624-01. FE-SEM and STEM work was performed at the Northwestern Electron Probe Instrument Center (EPIC), part of Northwestern University’s Atomic and Nanoscale Characterization Experimental Center (NUANCE). This work is funded by National Institute of Health, National Institute for Dental Research R01-DE001374-53.
References 1. Orban BJ, Bhaskar SN. Orban’s oral histology and embryology. 10th ed. St. Louis (MO): C.V. Mosby; 1986:102–10.
2. Mjor IA, Nordahl I. The density and branching of dentinal tubules in human teeth. Arch Oral Biol 1996;41:401–12. 3. Pashley DH. Dentin – a dynamic substrate – a review. Scan Microsc 1989;3:161–76. 4. Dutra-Correa M, Anauate-Netto C, Arana-Chavez VE. Density and diameter of dentinal tubules in etched and non-etched bovine dentine examined by scanning electron microscopy. Arch Oral Biol 2007;52:850–5. 5. Schilke R, Lisson JA, Bauss O, Geurtsen W. Comparison of the number and diameter of dentinal tubules in human and bovine dentine by scanning electron microscopic investigation. Arch Oral Biol 2000;45:355–61. 6. Garberoglio R, Brannstrom M. Scanning electron-microscopic investigation of human dentinal tubules. Arch Oral Biol 1976;21: 355–62. 7. Weiner S, Veis A, Beniash E, Arad T, Dillon JW, Sabsay B, Siddiqui F. Peritubular dentin formation: crystal organization and the macromolecular constituents in human teeth. J Struct Biol 1999; 126:27–41. 8. Bertassoni LE, Stankoska K, Swain MV. Insights into the structure and composition of the peritubular dentin organic matrix and the lamina limitans. Micron 2012;43:229–36.
Supplementary material available online Supplementary Figures S1 and S2, and Methods (contained in one document) Supplementary Figure S1: An example of the image analysis performed on the cross-sectional area occupied by tubules in dentin. Supplementary Figure S2: Sections J11M3R1Aa and J11M3R1Ab, isolated from just above the cervical line of a bovine molar from an 18 mo. Black Angus cow. Supplementary Methods: Preliminary Calculations to Estimate the Area of the Components of Dentin Supplemental material available at informahealthcare.com/cts
Copyright of Connective Tissue Research is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
