THE ANATOMICAL RECORD 295:1101–1108 (2012)
The Viability of Perilabyrinthine Osteocytes: A Quantitative Study Using Bulk-Stained Undecalcified Human Temporal Bones SUNE LAND BLOCH,* SØREN LUND KRISTENSEN, AND MADS SØLVSTEN SØRENSEN Department of Otorhinolaryngology Head and Neck Surgery, Otopathological Laboratory, Rigshospitalet, University Hospital of Copenhagen, Denmark
ABSTRACT Bone remodeling is highly inhibited around the inner ear space, most likely by the anti-resorptive action of the inner ear cytokine osteoprotegerin (OPG) entering perilabyrinthine bone through the lacuno-canalicular porosity (LCP). This extracellular signaling pathway depends on the viability of individual osteocytes. The objective of this study was to evaluate the patency of the LCP at different ages. Sixty-five bulk-stained undecalcified human temporal bones and 19 ribs were selected to span the ages from the 30th gestational week to 95 years. Osteocytes from inside a 2-mm wide perilabyrinthine zone of bone were identified by 3D vector calculations and the numerical densities estimated with an optical dissector and compared to age-matched ribs. From a high fetal count of 90,000 cells/mm3, the density of viable capsular osteocytes declined rapidly to 73,000 cells/mm3 at three years of age, and non-viable osteocytes increased inversely. After 3 years, this decline/increase continued at a much slower rate. The densities of viable as well as non-viable osteocytes and the rates of change were much higher in perilabyrinthine bone compared to ribs. Only after the age of 80 years had the density of viable capsular osteocytes declined to the level of ribs. The bi-phasic osteocyte kinetics reflects different development stages. The high initial density of viable osteocytes may secure a life-long anatomical route for inner-ear OPG despite the unique accumulation of non-viable osteocytes. Clustering of non-viable osteocytes may cause local aberrations in the signaling sysC 2012 Wiley tem by closure of the LCP. Anat Rec, 295:1101–1108, 2012. V Periodicals, Inc.
Key words: bone remodeling; development; osteoprotegerin; bone repair; otosclerosis
INTRODUCTION The human skeleton adapts and maintains its structure by bone remodeling at a turnover rate of around 10% per year. This is a coordinated process in which old bone is removed by osteoclasts and replaced with new bone formed by osteoblasts. However, studies on time-labeled undecalcified human and animal temporal bones have revealed how normal otic capsular bone resorption and remodeling is centripetally inhibited around the inner ear space to a mere 0.1% per year in perilabyrinthine bone (Frisch et al., C 2012 WILEY PERIODICALS, INC. V
*Correspondence to: Sune Land Bloch, MD, PhD, Department of Otorhinolaryngology Head and Neck Surgery, F2074, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark. E-mail:
[email protected] Received 10 June 2011; Accepted 5 April 2012. DOI 10.1002/ar.22492 Published online 11 May 2012 in Wiley Online Library (wileyonlinelibrary.com).
1102
BLOCH ET AL.
1998, 2000). The messenger responsible for this is most likely osteoprotegrin (OPG), a potent anti-resorptive cytokine detected in high levels inside the labyrinthine space (Zehnder et al., 2005). OPG knockout mice exhibit abnormal perilabyrinthine bone remodeling, stapes fixation and progressive loss of hearing (Kanzaki et al., 2006; Zehnder et al., 2006). OPG is assumed to enter the bony otic capsule through intercellular gaps of the inner ear lining and diffuse towards the capsular periphery via the lacuno-canalicular porosity (Sørensen et al., 2005; Zehnder et al., 2005). This branching, fluid-filled extracellular space accommodates the osteocytes and their connecting processes, and forms an ‘‘osseous functional network’’ involved in matrix and cellular metabolic traffic and signaling activity (Knothe Tate, 2003). The patency of the osseus functional network depends on the viability of individual osteocytes (Frost, 1960; Knothe Tate, 2003). Several studies have discussed the general viability of the bony otic capsule (Mayer, 1917; Eckert-M€ obius, 1926; Nager, 1947; Kakizaki et al., 1971). Most have described high numbers of empty lacunae in celloidine-embedded temporal bones. However, dead osteocytes may be overestimated in decalcified thin H&E sections (Eckert-M€obius, 1926; Frost, 1960). Recently, the spatial distribution of viable and non-viable otic capsular osteocytes was measured in undecalcified bulk-stained human materials (Bloch and Sørensen, 2010). The present study follows the changing viability of perilabyrinthine osteocytes over time and monitors the condition of the perilabyrinthine signaling pathway for inner ear OPG during development and aging of the human labyrinthine capsule. Otic capsular data are compared to an extracapsular rib bone reference with a normal rate of bone remodeling. Costal segments, which are easy to collect and process for microscopy, were chosen as reference because previous bulk-staining data exist (Frost, 1960). The results are discussed in a context of classical temporal bone morphology and novel temporal bone dynamics.
MATERIALS AND METHODS Materials From a total of >400 anonymized, bulk-stained human temporal bones routinely obtained in agreement with Danish medical legislation and ethical rules at the Otopathological Laboratory in Copenhagen, 65 temporal bones and 19 age-matched 4 cm rib segments were selected to span the ages from the 30th gestational week to 95 years. No other selection criteria were applied and no history of ear disease was indicated a priori. The series included 35 male and 20 female temporal bones, and nine male and 10 female rib segments.
Tissue Preparation Entire tissue blocks of undecalcified human temporal bones and ribs were bulk-stained by immersion in 62% ethanol with 1% basic fuchsin (Certistain fuchsin, Merck) for 2–4 months and embedded in methyl methacrylate. Horizontal sectioning with an Accutom-2 milling machine (Struers, Copenhagen, Denmark) or a KDG95 microslicer (BioScan BV, Netherlands) into 50–80-lmthick sections produced 20–35 sections per specimen.
The sections were slide-mounted, cover-slipped, and studied without any further staining.
Stereological Setup The computer-assisted setup consisted of an Olympus BX50 light microscope (Olympus, Japan) equipped with a motorized X–Y stage (M€ arzh€ auser, Germany), a microcator device for accurate measurement of stage movement in the z axis (Heidenheim, Germany), and a digital camera (ColorView II, Olympus, Japan) connected to a PC fitted with a Cintiq15x touch screen (Wacom, Saitama, Japan) and CAST-grid software (Visiopharm, Hørsholm, Denmark).
Criteria for Classification of Viable and NonViable Osteocytes in Bulk-Stained Sections We applied a modified version of the basic viability criteria used by Frost (Frost, 1960) to evaluate the in vivo state of osteocytes in two types of lacunae: (1) Osteocyte lacuna containing a living cell (obvious recognition of basic fuchsin in a well-defined starshaped lacuna with canaliculi, Fig. 1A) and (2) Osteocyte lacuna containing no viable cell (lack of basic fuchsin inside lacuna and canaliculi, or smooth lacuna with no canaliculi that may contain some cellular debris, Fig. 1B).
The Optical Dissector The optical dissector is a ‘‘virtual’’ 3D counting probe, in which counting relies on optical sectioning through a thick section by the creation of focal planes with a thin depth-of-field (Gundersen, 1986). The volume of the 3D counting probe is defined by the area of an unbiased counting frame and the z-distance in which objects are counted. Since the total volume of tissue scanned in this manner is known, the optical dissector counts the true number of objects in a known volume of tissue. To avoid oversampling bias, half of the surfaces of the 3D counting probe are designated as exclusion surfaces and the other half as inclusion surfaces. In practice, only objects that touch the inclusion lines or the horizontal inclusion plane, or fall entirely within the dissector frame, are counted when scanning though a thickness of 20 l of the section. The first 15 l from the upper and lower surface of a section are assigned as ‘‘guard zones’’ in which no counting is done. This avoids artifacts such as ‘‘lost caps,’’ that is, objects cut or torn out in the path of the knife blade, which artificially changes the number of objects appearing at the section surface. In most cases, cells contain unique subunits (landmarks, e.g., nucleus or nucleolus) that serve as the counting items for the objects. However, given the diversity of lacunar morphology, no unambiguous and recognizable landmarks exist and the ‘‘leading edge’’ (top of the encountered object) was assigned as the unique sampling point. Although the ‘‘leading edge’’ is not a unique geometrical feature of the object itself, there will be only one leading edge per osteocyte lacunae regardless of the plane of sectioning.
1103
THE VIABILITY OF PERILABYRINTHINE OSTEOCYTES
tions. This produced an enormous data set of 60,000 counted osteocytes. Only data belonging to the innermost 2,000 l was extracted and numerical densities calculated as described below.
Numerical Density of Viable and Non-Viable Osteocytes The numerical density, NV (the number of objects in a unit reference volume) is given by the total sum of P objects counted P Q divided by the total volume of dissector probes Volsamp: P P Q Q P P NV ðobjectsÞ ¼ ¼ Volsamp h ða=pÞ P where h ¼ height of the dissector probe (20 lm), (a/p) ¼ area of the dissector frame (3055 lm2), and P ¼ total number of dissector-associated test points hitting the bone tissue.
Precision of the Stereological Estimates The precision of NV is affected by the stereological contribution to the error variance which can be expressed by the coefficient of error, CE (NV). This consists of the variance when sampling within a section (the ‘‘Noise’’ effect), and the variance between sections, SURS variation (systematic uniformly random sampling), given by: CE ðNvÞ ¼
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P P Varnoise Q þ VarSURS Q P Q
The contribution to error variance caused by the noise approximately equals: Fig. 1. A human rib (A) and perilabyrinthine bone (B) from a patient of 60 years of age. Note the well-stained and star-shaped lacunas of viable osteocytes (A) compared to the empty or mineralized non-viable lacunas with smooth borders (B). Bulk-stain with basic fuchsin; original magnification, 100. The superimposed counting frames are 55 55 l.
VarNoise
X Q ¼ Q
The contribution to error variance caused by the SURS variation is calculated by: X VarSURS ¼ 3 A Q 4B þ C =240
Sampling Procedure On the basis of a pilot study, a systematic sampling fraction of 1/3 of the total data set was chosen. At low magnification (4), the boundary profiles of the inner ear space were first delineated on the digitizer by a wireless pointing device. Next, an oil-immersion objective of 100 magnification was selected and each field of vision superimposed with an unbiased counting frame generated by the CAST-grid software. Under the meander sampling, the step length of the motorized X– Y stage was fixed in the x and y plane to enable uniform movement of counting frames over the total section area. The unbiased counting frame was moved to between 200 and 250 systematic-random locations on 5–8 systematic-random sections through the entire reference space. From the entire dataset the x, y, and z coordinates of the inner ear space boundary profiles and dissector probes were exported to an Excel worksheet, where the shortest distance from the dissector probes (and associated data) to the inner ear space boundary profiles was estimated by 3D vector calcula-
X
where A¼
n X ðQ i xQi Þ; i¼1
n X C¼ ðQ i xQiþ2 Þ;
B¼
n X ðQ i xQiþ1 Þ; i¼1
i¼1
is the total number of dissector-counted Q 1, osteocytes within the sum of dissector volumes in the ith section, and n the number of sections (Gundersen et al., 1999). The average CE (NV) for m subjects is given by: !1=2 m 1 X 2 x CEðNvÞ ¼ CEi m i¼1
1104
BLOCH ET AL.
Fig. 2. Perilabyrinthine bone at the level of the round window opening from a patient of 1 year of age (A and B) and 70 years of age (C and D). Note the dense and viable osteocyte network in the infant otic capsule (A and B) compared to the less stained perilabyrinthine zone (white halo) in the ageing otic capsule as a result of mineralized or
Statistical Analysis All data were tabulated using Microsoft Excel 2000. Statistical analysis was performed with GraphPad Prism Version 5.0 (GraphPad software, San Diego, CA) using two-way ANOVA and non-linear regression. Changes in data trends along the plots were analyzed by means of generalized linear models, and the levels of significance were tested with the Monte Carlo permutation method (Joinpoint regression software, Version 3.4.2 from the Surveillance Research Program of the US National Cancer Institute). Obtaining a P value of
