CONFOCAL MICROSCOPY
Portable Confocal Microscope Reveals
Fossil Hominid Microstructure
Timothy G. Bromage1, Alejandro Perez-Ochoa 2, Alan Boyde 3
1. New York University College of Dentistry, NY, USA 2. Dept Paleontology, Universidad Complutense de Madrid, Spain
3. Dental Biophysics, Queen Mary University of London, UK
BIOGRAPHY
Tim Bromage is a mineralized tissue biologist combining human health and
evolutionary sciences in
studies of development,
function, and environmental response of the
skeleton. Together with his co-authors, he
regards the development of practical solutions to technical problems of preparation
and imaging as an essential responsibility and
contribution. African and Middle Eastern
fieldwork complement his laboratory research and student training.
ABSTRACT
Fossil hominid bones and teeth are mainly
examined at the organ level. Microstructural
studies, typically investigated by light
microscopy of histological thin sections or by
scanning electron microscopy of bulk specimens, would enhance our understanding of
skeletal function and development. However, unique hominid fossils are not readily
available for sectioning, and bulk examination is restricted to surfaces. A practical alternative is confocal scanning light microscopy
(CSLM), which generates excellent surface
reflection images and provides basic details
of sub-surface bone and tooth histology. This
article describes a Nipkow disk-based CSLM
system’s applicability to fossils from 2-3 million year old hominid sites.
KEYWORDS
portable confocal microscope, confocal scanning light microscopy, hominid skeletal
microstructure, bone and tooth histology
ACKNOWLEDGEMENTS
INTRODUCTION
Most fossil specimens are either translucent or,
if they are surface reflective, are not flat. In
both cases, light interacts with the sample over
a considerable vertical range and is reflected
(or the fluorescent light emanates) from a
thick layer. We have found a solution in the
development of portable confocal microscopy
for the evaluation of rare and unique early
hominid fossils.
The principle of the confocal scanning light
microscope (CSLM) is to eliminate the scattered, reflected or fluorescent light from outof-focus planes. Only light originating from
the plane of focus of the objective lens contributes to image formation at the several conjugate focal planes (intermediate, eye point,
image recording device) in deference to light
that is eliminated from all out-of-focus planes.
In practice, an illuminated spot in the plane of
focus is scanned across the field of view and an
image is compiled. CSLM thus differs from conventional light microscopy, where light from
the focus plane of the objective lens, as well as
from all out-of-focus planes across the entire
field of view, is observed. The history and various technical achievements in confocal
microscopy are summarized in Boyde [1].
We employed a CSLM based on the Nipkow
disk technique [2], described in detail by
Petran and Hadravsky [e.g. 3] and first commercialized in the early 1980s. The Petran and
Hadravsky design uses a so-called two-sided
disk; the specimen is illuminated through an
array of pinholes on one side of the disk while
detected through a conjugate array of pinholes on the other (via a number of delicately
aligned mirrors). Applications of this technology to bone and tooth microanatomy were
demonstrated by Boyde et al. [4]. Another Nipkow disk design employs a single-sided disk in
which the illumination and detection pinholes
are the same [5]. This latter design trades
slightly improved quantum efficiency for a
robust construction able to tolerate our relatively extreme portable applications.
M AT E R I A L S A N D M E T H O D S
A one-sided Nipkow disk (Technical Instrument Co. K2S-BIO confocal module, Zygo
Corp., Sunnyvale, CA) was specifically configured to contend with challenging imaging
problems such as those encountered in paleoanthropology (Figure 1) [6]. Like other confocal scanning light microscopes, the image
derives from the plane in focus, thus eliminating the fog due to the halo of reflected, scattered or fluorescent light from all elements in
the sample above and below the plane of
focus, which is otherwise confounding image
content in conventional light microscopy.
An interesting feature of the single-sided
disk design by Kino [5] is the solution taken to
suppress internal non-image-related reflections that are otherwise a significant
problem in this type of system: the classical
method of illuminating with polarized light to
stop that light reflecting from within the optical system (e.g. from optical hardware within
the body of the microscope), but not the useful light reflecting from the specimen and
returning through the objective lens. Linear
polarizing light filters and a single quarterwave plate are employed for this purpose. This
design reduces the number of mirrors in the
light path and the alignment of the optics is
not so critical, which makes the instrument
very robustly constructed and able to tolerate
transport and relatively rough handling (e.g.
as checked-in baggage for air travel).
The microscope configuration included
several other features critical to our research.
The authors thank Blanquer, Caja Madrid,
Cultek, March Foundations (Spain) for
support, and Ethiopia and Kenya National
Museums, South African Transvaal Museum
and Witwatersrand University for assistance.
Figure 1:
The system in use observing a
juvenile Australopithecus africanus facial skeleton (Taung
Child), ca 2.5-3.0 m.y., at the
Palaeoanthropology Research
Unit, Department of Anatomy,
University of the Witwatersrand,
South Africa.
A U T H O R D E TA I L S
Dr Timothy G. Bromage, Hard Tissue Research
Unit, Departments of Basic Science, Biomaterials and Biomimetics, New York University College of Dentistry, 345 East 24th Street,
New York, NY 10010, USA.
Tel: +1 212 998 9597
Email: [email protected]
Microscopy and Analysis, 19 (3): 5-7 (UK), 2005.
MICROSCOPY
AND
A N A LY S I S • M AY 2 0 0 5
5
Consideration was given to obtaining objective lenses with relatively long working distances (i.e. around 20 mm). Often we had little control over the geometry of broken fossil
bone surfaces examined under remote field or
museum conditions, and so we had to be prepared to image through long Z-height positions to avoid mechanical interference between the bone and the objective nosepiece.
Objectives chosen included a 10⫻ lens (19-mm
working distance; Thales-Optem Inc., Fairport,
NY, USA) and a Mitutoyo 20⫻ lens (20-mm
working distance; Mitutoyo Asia Pacific Pte
Ltd, Singapore). Flexibility in magnification
was achieved by both the introduction of a
Thales-Optem 0.5⫻ or 1.9⫻ CCD adapter or by
converting the fixed magnification optical
assembly described above into a zoom system,
which involved the introduction of a Thales
Optem 70XL zoom module (1-7⫻) between
the K2S-BIO module coupler and the manual
coarse/fine focus module. For fully automated
image acquisition, we motorized the Z focus.
Complete automation in all X, Y, and Z axes
was implemented onto the Portable Confocal
Microscope. This included a KP53 motorized
precision micro-stepping X-Y stage (Semprex
Corporation, Campbell, CA, USA) and a Vexta
2-phase Z-axis stepping motor (Oriental Motor
USA Corp., Torrance, CA, USA). Integrated XYZ
movement was performed by an Oasis 4i PCI
stepper motor controller board for XY stage
and Z focus. A three-axis trackball/mouse control of XYZ axes allowed manual stage and
focus movement to aid real-time viewing.
A JVC KY-F1030U 6-pin IEEE 1394 digital
camera economizes on weight at 470 gm and,
because it has a 6-pin IEEE 1394 connection,
does not require a separate power supply. The
camera contains a 1/2” colour progressive-scan
interline CCD containing 1360 ⫻ 1024 output
pixels, operating at 7.5 frames per second.
The 300 W Lambda LS xenon arc lamp (Sutter
Instrument Co, Novato, CA, USA) transmits a
flat and intense beam of light via a liquid light
guide, operates at wavelengths suitable for
both fluorescence and white light illumination,
is robustly constructed with a pre-aligned
lamp, and is economically packaged and light-
Figure 2:
Stereo pair of images of fractured enamel surface of a Paranthropus robustus molar (SKW 4769; Transvaal Museum) ca. 1.5-2.0 m.y. A three-dimensional
view of topographic relief in this monochrome surface reflection image may be obtained by reconstruction of left and right images into one stereoscopic
image. Syncroscopy Auto Montage image; HFW = 100 µm each frame.
weight, housing its own power supply in one
26.7 ⫻ 24.1 ⫻ 25.4 cm cabinet at 4.8 kg.
A Shuttle XPC SB52G2 with a Pentium 4 Intel
processor and Windows XP (Shuttle Computer
Group Inc., Los Angeles, CA, USA) supported
fully automated image acquisition. The
Shuttle XPC harbours a small form factor at 30
⫻ 20 ⫻ 18.5 cm and has a lightweight aluminum chassis; the entire PC weighs only 3.5
kg. The computer has PCI expansion cards
slots, a 6-pin IEEE 1394 port, a CDRW-DVD
drive, front and rear USB ports, and many
other common PC features. A reasonably lightweight and thin standard 1024 ⫻ 768 15”
monitor (Dell Inc., Round Rock, TX, USA) was
chosen for our real-time viewing.
The microscope returns image detail from a
very thin optical plane at and immediately
below the object surface (1-50 µm, depending
upon specimen characteristics). To obtain twoor three-dimensional projections from a surface which is anything but perfectly flat,
potential fields of view must be compiled from
a through-series of captured images at all optical planes represented in the Z-axis. Computerized control over image acquisition using
Syncroscopy Auto Montage software (Syncroscopy Inc., Frederick, MD, USA) permitted an
even and fully representative image of either
a pseudo-planar field of view or a threedimensional reconstruction of surface or sub-
Figure 3:
Same field of view as Figure 2, imaging deep to the surface and revealing incremental enamel microanatomy.
Overlain on this image is a color relief map of Figure 2 (dark blue through orange represents low though
elevated surfaces respectively). Syncroscopy Auto Montage image; HFW = 100 µm.
6
MICROSCOPY
AND
A N A LY S I S • M AY 2 0 0 5
surface details. For more extensive automated
XY image montaging, Syncroscopy Montage
Explorer (Syncroscopy Inc., Frederick, MD,
USA) software was employed, which can operate in ‘3D mode’ to acquire all pertinent Z
focal planes.
The stand and system integration was provided by GT Vision (Frederick, MD, USA). The
stand is simple and lightweight. Composed of
aluminum, it includes an upright cylinder, containing within a lead screw operable from
above, which drives the Nipkow disk module
platform up or down and serves as a coarse
focus adjustment. The cylinder inserts into a
sleeve at the base from which two hollow rectangular feet slide forward and rotate out at
any angle appropriate for the balance of
weight and required workspace. The platform
rides on a bearing, conveying the Nipkow disk
module in any rotational position.
The entire microscope assembly fits into two
suitcases (Pelican Products, Inc., Torrance, CA,
USA), automatically switches between 110V
and 220V electrical supplies (only the Nipkow
disk motor requires optional 110/220V adapters), and may be set up and tested within one
hour of arrival at museum locations.
R E S U LT S
Several applications of the Portable Confocal
Microscope have been tested at African repo-
Figure 4:
Natural enamel surface of an early Homo molar (KNM-ER 2600; Kenya National Museum) ca. 1.5 m.y. Enamel
prisms at depth are visible as small specks, while large surface incremental structures (perikymata) are visible in
this false color image as arching structures top-to-bottom. Syncroscopy Auto Montage image; HFW = 100 µm.
CONFOCAL MICROSCOPY
Figure 5:
Fractured surface across an Australopithecus
afarensis femur (A.L. 288-1ap, ‘Lucy’;
Ethiopia National Museum) ca. 3.0 m.y.
Image brightness relates to collagen fiber orientation that existed in life. Syncroscopy
Montage Explorer image; HFW = 450 µm
book-based PC monochrome image acquisition
system (Figures 2 and 3 were obtained with this
system). This microscope is dedicated to specific long-term projects (e.g. dissertations). The
microscope described here (PCM-2K2; Figure 1)
is fully automated in X, Y and Z, and offers the
maximum potential of this technology for the
imaging and analysis of unique early hominid
bone and tooth specimens at remote locations.
With development of the Portable Confocal
Microscope the potential for non-destructive
mineralized tissue research on rare and unique
early hominid remains is great.
CONCLUSIONS
sitories of early hominid skeletal remains.
There is much interest in obtaining details of
hominid enamel microanatomy from fractured
surfaces, but such surfaces are rarely forgiving
and the resolving power has been wanting.
Figure 2 illustrates a 3D image of a complex
naturally fractured surface from a Paranthropus robustus molar. Figure 3 presents the same
field of view after the adherence of a glass coverslip to the surface with glycerin. Throughfocus imaging of this topographically complex
surface revealed a plane view of enamel increments (e.g. striae of Retzius coursing from
upper left to lower right). Overlain onto this
image is an Auto Montage colour-coded relief
map of the original surface for comparison.
Examination of intact enamel surfaces provides excellent reflection images, though it is
also useful to combine this with sub-surface
image content. Figure 4 represents such an
image of an early Homo molar in which
through-focus images include both surface
incremental features (perikymata) as well as
subsurface enamel prism details.
Because reflected light conditions are intriguingly different from the properties of
transmitted light, the polarizing and quarter
wave filter arrangement employed for reducing unwanted internal reflections results in
the conditions necessary to generate a circularly polarized light (CPL) image [7]. For this
reason birefringence associated with collagen
fibre orientation in bone may be examined [8].
Figure 5 presents a CPL image of the cross section of an Australopithecus afarensis femur
(‘Lucy’). The laminated concentric structures
are vascular canal spaces surrounded by lamellar increments of bone formed during the
incorporation of a new blood vessel. This
process of internal remodeling is a natural
process occurring with age. As bone lamellae
are laid down, the orientation of their contained collagen is seen to be either relatively
plane with the cross-section of bone (appearing bright) or perpendicular to the cross-section of bone (appearing dark). Images such as
these may be used to interpret the manner in
which the bone was used in life. This image
derived from the anterior cortex is relatively
‘dark’, indicating that this cortex was better
adapted to resist tensile forces in life.
It is useful to know something of the
distributions of cell spaces in bone (osteocyte
lacunae), particularly when reconstructing the
developmental history of a bone whose surface has been damaged during fossilization
and cannot be evaluated. Bone cells align with
the prevailing collagen fibre orientation,
which varies as a function of bone growth pattern, thus 3D images of cell spaces in fossil
bone illustrates something of its developmental history. Figure 6 presents a 3D image of
Australopithecus afarensis facial bone lacunae, whose regular orientation is an indication
that the bone surface above was originally
forming (not resorbing, as it would be in this
location in modern humans) during growth.
DISCUSSION
While the improvement over conventional
light microscopy in imaging thin sections may
not be too important, the improvement made
by the Portable Confocal Microscope for the
examination of the surface layers of bulk
samples is nothing short of revolutionary.
Even if images cannot be obtained through a
great depth, the convenience factor of not
having to produce a thin section as a prerequisite for excellent optical microscopy is a
very great advantage in our research.
Two microscopes are in service to date. The
first (PCM-1K2) was described by Bromage et
al. [6]; it is automated in Z and operates a note-
The Portable Confocal Microscope is the first
instrument to offer superb analytical light
microscopy of early hominid skeletal material.
Limitations over the handling and transport of
rare fossils have motivated the development
of this technology, but it is as well suited to
applications as diverse as art conservation,
forensics and the space sciences, i.e. wherever
and whenever the microscope must go to the
place and the subject.
REFERENCES
1. Boyde, A. Confocal optical microscopy. In: Image Analysis in
Histology: Conventional and Confocal Microscopy, Eds:
Wootton, R., Springall, D.R. and Polak, J.M. Cambridge
University Press, Cambridge, UK, pp 151-196, 1995.
2. Nipkow, P. Elektrisches Teleskop. Patentschrift 30105
(Kaiserliches Patentamt, Berlin), patented 06.01.1884, 1884.
3. Petran, M. and Hadravsky, M. Method and arrangement for
improving the resolving power and contrast. US Patent No.
3,517,980, priority 05.12.1966, patented 30.06.1970 US, 1966.
4. Boyde, A. et al. Tandem scanning reflected light microscopy of
internal features in whole bone and tooth samples. Journal of
Microscopy 132:1-7, 1983.
5. Kino, G. S. Intermediate optics in Nipkow disk microscopes. In:
Handbook of Biological Confocal Microscopy, Ed: Pawley, J.B.
Plenum Press: New York. pp155-165, 1995.
6. Bromage, T. G. et al. The portable confocal microscope:
Scanning optical microscopy anywhere. In: Science,
Technology and Education of Microscopy: An Overview, Ed:
Méndez-Vilas, A. Formatex, Badajoz, Spain, pp 742-752, 2003.
7. Bromage, T. G. et al. Circularly polarized light standards for
investigations of collagen fibre orientation in bone.
Anatomical Record: The New Anatomist 274B:157–168, 2003.
8. Goldman, H. M. et al. (2003) Preferred collagen fiber
orientation at the human mid-shaft femur. Anatomical Record
272A:434-445, 2003
©2005 John Wiley & Sons, Ltd
Figure 6:
Stereo pair of sub-surface real-color images from alveolar clivus (below the nose) of juvenile Australopithecus afarensis upper jaw (A.L. 333-105; Ethiopia
National Museum) ca. 3.0 m.y. A three-dimensional view of bone producing cell spaces (osteocyte lacunae) below the surface to a depth of ca. 50 µm
may be obtained by reconstruction of left and right images into one stereoscopic image. Syncroscopy Auto Montage image; HFW = 450 µm each frame.
MICROSCOPY
AND
A N A LY S I S • M AY 2 0 0 5
7