Am J Physiol Heart Circ Physiol 289: H57-H65, 2005.
First published February 25, 2005; doi:10.1152/ajpheart.01001.2004
0363-6135/05 $8.00
Variants of the tissue-sensor array window chamber
Milan T. Makale,
Peter C. Chen, and
David A. Gough
Department of Bioengineering, University of California-San Diego, La Jolla, California
Submitted 1 October 2004
; accepted in final form 22 February 2005
 |
ABSTRACT
|
|---|
Sensors are being developed that can be implanted in tissues for continuous monitoring of oxygen, glucose, and other metabolites. However, there have been difficulties in inferring metabolite concentrations in blood from the signals of tissue sensors due to the properties of tissues at the implant site and local physiological phenomena that can affect sensor responses. A multisensor array has been previously developed for implantation in a hamster skinfold window chamber preparation to study these effects. The preparation allows recording of concentration-dependent signals from multiple sensors while nondestructively visualizing the adjacent tissue and microvascular function. Variants of the tissue-sensor array window chamber described here have respective advantages over the original chamber design, including improved tissue visualization and reduced surgical intervention, and allow exposure of the sensor to different tissues. Results indicate that mass transfer within tissues is heterogeneous, and sensor signals are affected by variable perfusion of local microvasculature in addition to vascular metabolite concentration. These observations suggest new strategies for sensor design and operation. Window chamber variants are important tools for validation of implanted sensors.
implanted biosensors; tissue visualization; sensor validation
THERE IS CONSIDERABLE INTEREST in developing implantable electrochemical sensors to monitor glucose, oxygen, lactate, and certain other metabolites for applications in diabetes, pulmonary insufficiency, shock, extreme exertion, and other conditions. The sensors would be implanted in a tissue site for periods of weeks to years, continuously monitor the metabolite of interest, and transmit the information externally either via a wire connection or wireless telemetry link. The goal is to infer dynamic blood concentrations of the metabolite of interest from signals of the sensors implanted in tissues. Sensors would ideally function continuously to provide an automatic and timely indication of metabolic imbalances. One example is the potentiostatic oxygen sensor based on the direct electrochemical reaction of oxygen at a membrane-covered, polarized, noble metal electrode (15). This sensor produces a signal current that is proportional to both the concentration and rate of mass transfer of oxygen to the sensor (16). Other examples include implantable enzyme electrode-type sensors for monitoring glucose, lactate, and other metabolites, in which membranes containing immobilized enzymes are coupled to electrochemical oxygen sensors (3, 4). As these sensors are based on the oxygen sensor, many of the response characteristics are similar to those of oxygen sensors, and the simpler oxygen sensor can therefore serve as a model.
In validating the function of sensors when operated in tissues, difficulties have arisen in assuring that the signals provide an accurate representation of blood concentrations (9, 11, 13, 19, 20, 23). These difficulties have been of two types. In certain cases, the sensors themselves have been the problem as a result of not being selective for the metabolite of interest, adequately sensitive at physicological concentrations, or sufficiently stable over the long term, or for various other reasons that are best addressed before implantation. In other cases where the sensor per se has been acceptable, ambiguous and confusing signals have often resulted when the sensors were operated as implants in tissues. Examples of such unpredictable responses are the following: sensor signals that vary over the short term while blood concentrations remain constant; signals that do not respond to changes in blood concentration; signals that display lags or delays; and neighboring sensors in the same tissue that produce quite different responses. Unfortunately, there have been few detailed, systematic studies to determine the causes of these phenomena. Rather, most previous studies have been attempts to establish simple correlations between signals and substrate concentrations in the blood, with little consideration of the underlying physiological events.
New biological preparations are needed to address the reasons for ambiguous sensor response. These preparations must permit a nondestructive study of the tissue environment in the vicinity of sensors. For example, it is necessary to visualize the microvascular pattern and measure the distance for mass transfer between the nearest capillaries and the sensor. Furthermore, it is important to determine whether the capillary permeability is intact and whether local blood vessels are fully perfused and functional. Blood flow must be documented in microvessels in the immediate vicinity of individual sensors and in the regional vasculature affecting all sensors in the array. It is necessary to ascertain the composition, structure, and function of tissues in the vicinity of the sensor and whether these features change with implant time over the long term.
We previously described a modification of the hamster skinfold window chamber in which an array of identical oxygen sensors was implanted (17). This system has a number of useful features, including 1) the tissue environment around individual sensors can be visualized in a nondestructive manner over a period of weeks to months by using optical microscopy, without the need for anesthesia and its potentially depressive effects on microcirculatory function; 2) many identical sensors can be studied in the same animal and blood concentrations can be simultaneously sampled, reducing between-animal variation; 3) studies can be carried out repeatedly to determine the effects of implant time and aging; and 4) postmortem histological analysis of tissues can be performed to analyze tissue structure and composition. These features have been advantageous in understanding the determinants of oxygen mass transfer to implanted oxygen sensors (16).
The previous tissue-sensor array window chamber is a modification of tissue-viewing chambers developed for use in the mouse, rat, hamster, and humans (6, 8, 18, 22, 25). Various types of tissue-viewing chambers have been used extensively for study of the microvasculature (12, 18), tumors (10, 22), cellular behavior (1), wound healing (7, 24), and imaging of internal organs (5). Versions of chambers have been adapted for visualization of rabbit ear tissues (2), mouse, rat, and hamster dorsal skinfold (8, 25), the body wall of the mouse (5), superficial tissues of the human arm (6), and tissues of the human leg (24).
We have found that the tissue-sensor array window chamber can be prepared in several ways with the array exposed to different tissue layers. The basic "tissuefold-sensor array" window chamber (17) has a window on one side of the debrided tissuefold and the sensor array on the opposite side, with the fold having either subcutaneous muscle or connective tissue exposed to the array. This preparation can be modified to become a "one-sided tissue-sensor array" window chamber variant, in which one side of the tissuefold is removed and the sensor array is placed in contact with inner layers of the opposing fold. The outer layers are left intact, and a window is not used. Another variant is the "single-layer tissue-sensor array" window chamber, in which all tissue layers are removed from both sides except a single layer of retractor muscle or connective tissue, with the array on one side and the window on the other. Another version is based on a novel sensor array, the "porthole sensor array," which has transparent viewing ports in the plane of the disc between the individual sensors. The "one-sided tissue/porthole sensor array" window chamber variant allows direct visualization of tissue-array interface without intervening tissue or use of a separate window. These preparations differ in the type of tissue in contact with the array, the ability to visualize the tissue-array interface, the amounts of tissue removed and associated disruption of microvascular perfusion, and the viable lifetime.
In this communication, we describe these preparations and outline their use for addressing questions about the response of implanted sensors. Detailed analyses of implanted sensor response are given elsewhere.
|
METHODS AND MATERIALS
|
|---|
Animal subjects.
Male Syrian golden hamsters (Charles River Laboratories, Cambridge, MA), weighing between 60 and 200 g, were housed in standard microisolator cages ventilated with filtered air. Male animals were used because they have less subcutaneous fat and a smaller range of normal variation in tissue composition. Animals were fed Purina Hamster Chow and water ad libitum. Animals were treated in accordance with National Institutes of Health guidelines for the care and use of laboratory animals. Protocols were approved by the University of California-San Diego Institutional Animal Care and Use Committee.
Tissue and array support apparatus.
The support apparatus, shown in Fig. 1 and described in detail elsewhere (17), consisted of two titanium alloy frames, each having a 12-mm-diameter circular opening fitted with a ring for attachment of the window or sensor array. Each ring extended 0.95 mm toward the tissue and, in conjunction with appropriate spacers to separate the frames, produced sufficient separation to allow vigorous perfusion of the microvasculature in the thin sheet of tissue. One ring had three small pins that penetrate the tissue and mate with holes of the opposing ring to restrict tissue movement. A glass microscope coverslip was secured into the window of one frame, and the sensor array disc was secured into the other frame by using slotted retaining rings. In some experiments, an additional 100-µm-thick plastic disc was glued to the internal surface of the window to minimize any fluid layer between the tissue and array surfaces. The frames were held together with four M2-4 x 10-mm bolts and nuts.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 1. Tissue support apparatus. A: window frame. B: sensor array support with sensor array. Sensor array disc is secured to the frame with connector strip. Electrical contacts at the tip are shown in black, extending vertically.
|
|
Sensor arrays.
The fabrication of oxygen sensor arrays has been described in detail elsewhere (16). The standard sensor array, shown in Fig. 2A, is a 12-mm-diameter ceramic disc with thick-film electrodes in specified patterns. The electrodes are composed of disc platinum working electrodes of 125 or 300 µm in diameter, four common Ag/AgCl disc reference electrodes 
875 µm in diameter, and a ribbon-like common platinum counter electrode. There are 18 working electrodes separated from each other by distances of 1–2 mm on each array. A 25-µm layer of conductive electrolyte (1.0 N NaCl, pH 7.3, in cross-linked polyhydroxyethylmethacrylate gel) is deposited by spin casting on the ceramic substrate, followed by spin casting a 25-µm layer of polydimethylsiloxane for specificity to oxygen and electrochemical isolation from the tissue.
View larger version (84K):
[in this window]
[in a new window]
|
Fig. 2. Sensor Arrays. A: basic sensor array with 18 small, disc-shaped working electrodes, 4 large disc-shaped Ag/AgCl potential reference electrodes, and a ribbon-shaped common counterelectrode. Polymer membrane is not visible. B: porthole sensor array. Several elongated windows near working electrodes allow direct visualization of the tissue-array interface.
|
|
The porthole sensor array, shown in Fig. 2B, has the same electrode array but includes five variously shaped holes in the ceramic plate near the edge between certain working electrodes. The holes are filled with optically clear silicone rubber (Nusil MED-6015, Carpenteria, CA) to allow tissue imaging through the sensor array. It was observed that extracellular material adsorbed on the silicone rubber material after 2 days, limiting the period of visibility, suggesting that the use of other transparent materials may lead to improvement.
An individual wire connection was made to each sensor from the back of the array disc. A rectangular ceramic plate having patterned conductive traces mated with a multipin fan connector, which was connected via ribbon cables to a custom, multichannel potentiostat instrument for polarization of the individual working electrodes and conveyance of the signal currents. The data were acquired, displayed, and stored using software written in Labview.
Sensor calibration.
Although individual sensors were similar in their manufacture and in vitro response to oxygen concentration, there was no intention in this study to achieve identical sensitivity and performance. Rather, the sensitivity of each sensor was determined individually by using a method described previously (16), in which measurements are made in the gas phase where boundary layers are absent. This method allows repeatable measurement of sensor signals without error due to the effects of variable mass transfer boundary layers found in the stirred liquid phase. The sensitivity of each sensor was archived respectively and used in calculation of oxygen concentrations. The measurement of oxygen sensitivity was repeated at the end of the study to verify individual sensor stability.
Sterilization.
The window chamber plates, bolts, glass coverslips, and sensor arrays were sterilized by soaking for 24 h in a solution of 6% glutaraldehyde in phosphate-buffered isotonic saline (PBS), pH 7.4. The sensors were removed from the solution, rinsed with sterile PBS, and then soaked in sterile PBS with several changes over 48 h. All surgical instruments and gauze were sterilized before use, and a bead sterilizer was used to resterilize instruments during surgery.
Animal care.
The animals were anesthetized by intraperitoneal injection of 100 mg/kg of ketamine and 250 µg/kg of medetomidine using a 1-ml syringe fitted with a 30-gauge needle. In our experience, this anesthetic combination offers a higher margin of safety than pentobarbital sodium, with a smoother induction, longer duration, and more rapid recovery, especially when atipamezole, the antidote to medetomidine, is injected upon completion of surgery. When the animals attained a surgical plane of anesthesia, as determined by toe pinch, the fur was shaved from the dorsum. Fine residual hair was removed with a depilatory, and both skin areas were swabbed with Betadine and washed with sterile PBS.
The anesthetized animal was placed prone on a heated pad, and the dorsal skin was gently suspended as a longitudinal fold using 4-0 silk secured to vertical metal rods. After tissues were dissected and the respective tissue-sensor chamber was installed as described below, the animal was injected subcutaneously with 25–75 µg of atipmezole and allowed to recover in a heated enclosure. After a 24-h recovery period, the animal was ready for study. At the conclusion of the experiment, the animal was euthanized with an overdose of pentobarbital (300 mg/kg) according to National Institutes of Health guidelines.
In some cases, it was necessary to anesthetize the animal with 75 mg ketamine and 2.5 mg xylazine/100 mg body wt during imaging so that acceptable images could be obtained without movement artifact. This anesthetic regimen (14) resulted in a 20% reduction in mean arterial blood pressure and 35% reduction in heart rate in other species. These perturbations were considered tolerable to produce acceptable images.
Tissuefold window chamber preparation.
For the original tissuefold window chamber, a 12-mm-diameter disc of skin, subcutaneous fat and fascia, and subcutaneous muscle was removed from each side of the skinfold to expose both faces of the folded retractor muscle. Topical broad-spectrum antibiotic cream (polysporin) was applied around the open sites. Approximately 0.4 ml of sterile buffer containing 0.05 ml of a 40 mg/ml solution of a broad-spectrum antibiotic (tobramycin) was deposited dropwise onto the exposed side of the retractor muscle, the titanium frame with the sensor array attached was fitted to one side, and the frame with the glass window was fitted to the opposite side. The plates were then secured in place with bolts and nuts, and a topical antibiotic was applied along the top margin of the skinfold.
One-sided variant.
A disc of skin, subcutaneous fat and fascia, and two layers of retractor muscle were removed from one side of the skinfold to expose the subcutaneous fat and skin. In some cases, a single layer of the retractor muscle, subcutaneous tissue, and skin were left entirely intact, whereas in other preparations all the muscle and fascia layers were removed in the vicinity of the A-0 vessels and only the skin was left intact. Antibiotic cream was applied around the site, sterile buffer containing broad-spectrum antibiotic (tobramycin) was deposited onto the exposed tissue, and the sensor frame was secured in place with the sensor in direct contact with the feeding tissue layer sheet or the skin. A frame was secured on the intact skin side, and the frames were fixed in place as before.
Single-layer variant.
The skin and subcutaneous fat and fascia were removed as a disc on one side, and the sensor array was placed against an unfolded, single layer of retractor muscle or fascia, with the retractor muscle, subcutaneous tissue and skin removed from the opposite side. The result was a single thin layer of retractor muscle or fascia sandwiched between the sensor array and the coverslip window. This preparation differed from the skin variant in that more tissue was removed at implantation reducing the viable lifetime of the preparation, but visualization through the single muscle layer was improved. Otherwise, the preparation involved procedures described above.
Porthole variant.
For the tissue-porthole sensor array preparation, the disc of skin, subcutaneous fat, and fascia was removed from the anterior region of one side of the skinfold opposite to the frame. Both layers of the retractor muscle were also removed. Several drops of sterile PBS containing broad-spectrum antibiotic were applied to subcutaneous fat, and topical antibiotic was spread on the skin surrounding the surgical site. The array frame was secured in place so that the array surface contacted either the vascularized fascia or the skin.
Catheterization.
Catheters were implanted in the jugular vein and carotid artery of certain animals to facilitate blood sampling and infusion of solutions, fluorescent dye, and pharmacological agents. Catheterization was performed at the time of chamber implantation in some animals and 24–48 h after chamber implantation in others. The animal was anesthetized with 100 mg/kg ketamine and 250 µg/kg medetomidine injected intraperitoneally. The ventral aspect of the neck was shaved, swabbed with Betadine, and washed with saline. A 2-cm midline incision was made, and the right external jugular was exposed using blunt dissection. A sterile, 2-Fr cannula made of polyurethane tubing with a polished tip containing 100% glycerol USP with 50 U/ml of heparin was introduced into the vein and drawn under the skin to exit dorsally between the shoulder blades. The ventral and dorsal incisions were closed with tissue cement, and the catheter was coiled and secured to the dorsal skinfold chamber frames. The left carotid artery was similarly cannulated in certain animals. A schematic of the complete tissue-sensor array preparation and catheters mounted on a hamster is shown in Fig. 3.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 3. Complete tissue-sensor array window chamber preparation. Potentiostat cables connect to a multichannel potentiostat. Arterial and venous catheters allow blood sampling and fluid infusion.
|
|
Microscopy.
The unanesthetized animal was introduced into a clear, perforated Plexiglas tube, and the chamber plates were secured to a block of Plexiglas to minimize movement of the chamber. The apparatus was placed on the microscope stage, and the chamber vasculature was imaged at x4 and x10. Approximately 12.5 mg FITC-dextran (2,000,000 mol wt) was injected via the jugular vein catheter. A CCD video camera (model 8215-1000, Cohu) attached to the microscope with a x10 objective was used to visualize the vasculature in the vicinity of a working oxygen sensor electrode. Still images were obtained with a digital camera attached to the microscope. The video captured blood flow through the vasculature in the vicinity of the working electrode. Both still images and videos were obtained and transferred to a computer for storage and analysis.
Oxygen challenges.
Step challenges in inspired oxygen were created by allowing the animal to breathe room air (20.9% oxygen) for typically 30 min, followed by 15% oxygen for 10 min, and a return to room air. Challenges produced no apparent discomfort.
|
RESULTS AND DISCUSSION
|
|---|
Tissue structure.
The hamster intact dermal tissue is composed of several tissue layers with respective vascular supplies (21). A single tissue layer is shown in histological cross section in Fig. 4A, and a tissuefold is shown schematically in Fig. 4B. The components are epidermis or skin, which is typically 15–50 µm thick and minimally vascularized; the dermis and hypodermis, referred to here as fat and fascia, contains adipose cells and is typically 350–750 µm thick; and the retractor muscle, which is typically 25–200 µm thick and well vascularized.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 4. Structure of hamster dorsal tissue. A: histological section. A single layer of hamster skin stained with hematoxylin and eosin is shown. B: schematic of folded tissue. Dimensions (in µm) depend on the region of the dorsum and hamster age. C: pattern of major feeding vessels in dorsal skin. A, head; R, right; L, left. Arrows show the points at which vessels communicate to the skin from the body wall.
|
|
Underlying the muscle is loose connective tissue that can be readily separated by dissection. The dimensions above are specific to the hamster and depend on body mass and the region from which the dorsum is sampled. Typical patterns of large feeder vessels in the plane of the skin are shown in Fig. 4C.
Summary of window chamber variants.
Considerations in design of window chamber variants are the ability to 1) provide direct contact or exposure of the sensor array to different types of tissue, such as muscle, fascia, or dermis; 2) visualize the tissue-sensor interface with minimal interference from tissue layers; 3) employ nonsensor materials implants for biocompatibility studies; 4) minimize surgical intervention to the tissue, including interruption and redirection of the vascular structure and perfusion patterns; and 5) apply the preparation to other rodent species. The different preparations address these considerations to various degrees. The design variants are represented schematically in Fig. 5, A–C, and are summarized in Table 1.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 5. Schematic of variant designs. A: tissuefold-sensor array (17). B: one-sided tissue-sensor array. This preparation can accommodate a window, a sensor array, or a porthole sensor. C: single-layer tissue-sensor array.
|
|
Features of the variants.
The tissuefold-sensor array window chamber (17) consists of a fold of dorsal skin retractor muscle with the skin layer removed, with the sensor array held adjacent to one side of the tissue and the window adjacent to the other side, as shown schematically in Fig. 5A. Visualization of the tissue-sensor array interface is through two layers of folded retractor muscle using either reflected light or illumination through the opaque ceramic disc. An advantage of the preparation is that there is relatively little modification of natural perfusion patterns in the muscle tissue, because only the dermal layer or the dermal layer and subcutaneous muscle layer are removed. Depending on the membrane material and its surface roughness, the tissue either becomes attached to the membrane surface or approximates the surface with a very small fluid-filled gap (1–5 µm). The impermeable glass window can serve as a control for the effects of tissue contact with the membrane and for potential effects on the tissue of oxygen consumption by the sensors.
The one-sided tissue-sensor array preparation (Fig. 5B) has intact, depilated skin on one side (a window is not used), and dermal tissue in contact with the sensor array or biomaterial on the opposite side. Visualization is through the single-thickness skin layer, and clarity can be comparable to the above tissuefold-sensor array preparation. This preparation provides direct contact between the tissue and sensor array surface, without stretching the tissue to adhere to a glass window on the opposing side. The preparation can be maintained viable for several weeks, because there is less damage due to tissue removal and rearrangement of perfusion. This preparation is particularly useful to display the effects of blood flow and vascular density on sensor function. A different type of sensor array used with this tissue configuration is the one-sided tissue-porthole sensor array, which has transparent holes in the plane of the ceramic disc between certain working electrodes and is implanted against the muscle, fascia, or subcutaneous fat. This variant allows direct imaging of the vascularity at the tissue-array interface adjacent to the working electrodes.
In the single-layer tissue-sensor array preparation (Fig. 5C), the tissue on both sides of the fold is removed leaving a single sheet of retractor muscle sandwiched between the array and window glass. Although visualization of the tissue is improved, the viable lifetime of the preparation is reduced due to the substantial amount of tissue removed.
Examples of tissue imaging.
Figure 6, A–D, shows images taken from each of the preparations. The microvasculature was readily visualized by trans- or epi-illumination in all preparations. In some cases, the presence of intervening tissue precluded optical imaging of discrete capillaries, but the general level of perfusion on various regions of the chamber could nevertheless be readily identified using intravenously injected fluorescent dye. The tissue-sensor array preparations showing the retractor muscle and the feeding layer allowed still- and video-based imaging of the microvasculature, as performed with epi-illumination at x1.25, x4, and x10 objective magnification. The best visualization of the microvasculature adjacent to the working electrodes was achieved with the single-layer tissue-sensor array (Fig. 6C) and the tissue/porthole sensor array preparations (Fig. 6D) at x10. This created a limited field of view but provided a high degree of detail in terms of capillaries and the flow of blood through the microvessels. The porthole array preparation is advantageous in that it was relatively less traumatic, and blood flow appears relatively undisturbed, even though windows were not directly over individual sensors.
View larger version (69K):
[in this window]
[in a new window]
|
Fig. 6. Examples of tissue images. A: tissuefold-sensor array preparation, x1.25. B: one-sided tissue-sensor array preparation, x4. Direct and fluorescent images are of the same tissue. C: single-layer tissue-sensor array preparation, x4. D: one-sided tissue-porthole sensor array, x10.
|
|
Optical parameters obtained with the preparations are summarized in Table 2. Estimates of working distance, depth of field, and resolution were made with three optical objectives used here.
Advantages of direct imaging.
There are several advantages to direct tissue visualization provided by the preparations. First, the ability to see the vascular pattern adjacent to individual sensors can help explain different steady-state signals of neighboring sensors. Sensors in regions of higher capillary density can be expected to sustain higher substrate fluxes and produce higher currents. However, an important additional variable is the distance orthogonal to the tissue plane from the microvascular plexus and the sensor surface, which can be estimated by focal plane measurements in the living tissue or by postmortem histological measurements. Second, visualization of the living tissue is advantageous to determine the degree of perfusion of the tissue as a whole as well as of the local microvasculature adjacent to individual sensors. Finally, visualization is advantageous to document any changes in tissue or microvasculature structure and function with implant time. Without the ability to visualize the tissue, a full explanation of sensor signals would not be possible.
Examples of sensor recordings.
Typical simultaneous recordings from implanted sensors arrays are shown in Fig. 7, A–D. The results are shown for oxygen because the interpretation is relatively straightforward, but results having similar characteristics have been obtained for sensors for glucose. Individual oxygen sensor signals are reported in nanoamperes as a function of time in minutes during challenges in inspired oxygen.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 7. A–D: examples of sensor recordings from each preparation. Simultaneous signal recordings (in nA) as a function of time in minutes during an inspired oxygen challenge are shown. Recordings were low-pass filtered, and some tracings have been removed for clarity. Inspired oxygen challenges (15% oxygen) caused a linear drop in all signals to a minimum, followed by a return to atmospheric inspired oxygen (21%) and a rise in signals. A detailed analysis of signals is given elsewhere (16).
|
|
There are several interesting observations. After an initial decline in sensitivity immediately after implantation, the sensitivity to substrate concentration typically remains stable for the lifetime of the window chamber. After this stabilization period, the steady-state signal magnitude of the implanted sensor is substantially smaller than when tested in vitro for the same oxygen exposure. This difference is attributed to differences in mass transfer resistance, which have been quantitatively analyzed (16). There is also a certain variance in steady-state signal magnitude among sensors within a given array when tested in vitro in homogeneous mass transfer conditions, reflecting small irregularities in fabrication. The effect of this variance is compensated by individual sensor calibration and confirmation of the sensitivity to oxygen by the same method (16) after the study. However, when sensors are used in the window chamber, there is an additional, relatively large steady-state variance among sensors due to tissue heterogeneity, which is prominent in Fig. 7, A–D. Specific values of mass transfer parameters corresponding to individual implanted sensors have been reported (16).
Signals also show two types of transients during steady-state recording sessions. Rapid, small-amplitude fluctuations in signals may represent electronic noise or the effects spatially heterogeneous vasomotion in the vicinity of individual sensors. Slower, large-amplitude variations in signal that are often correlated among multiple sensors are also observed, which may reflect variations in regional perfusion. Interpretation of these phenomena is made feasible by the ability to visualize tissues afforded by the window chamber preparations. Extensive analysis of the relationship between signals and variations in perfusion local microvasculature and regional blood vessels is described in detail elsewhere. Furthermore, sensors respond rapidly and substantially to changes in inspired oxygen, indicating a relatively rapid overall mass transfer from the airways to local tissues. However, this response is not necessarily proportional to the inspired oxygen concentration, suggesting that local physiological mechanisms protective against hypoxia may be important.
These features were observed in all preparations. These observations suggest that new sensor designs or new operating procedures may be needed to reduce or adequately account for the effects of tissue phenomena that modify the response to intravascular substrate concentration. Detailed analysis of sensor design is presented elsewhere.
Limitations of the window chamber methodology.
Although the window chamber preparations have major advantages, several limitations are recognized. It is often necessary to perform extensive blood sampling to record blood substrate dynamics in certain experiments, which is not feasible in the hamster due to the limited blood volume. The respective window preparations have a maximal lifetime of several days to months, which can limit long-term testing. The sustained immobilization of the tissue may lead to unnatural perfusion and tissue growth. The surgical preparation represents a substantial tissue modification and prolonged tissue exposure, which likely leads to an unnatural wound healing response. With these reservations in mind, the disadvantages are mitigated in certain experimental situations by the unusual advantage of being able to observe tissues in a relatively natural, unanesthetized state while sensor recordings are being made.
In conclusion, sensors implanted in tissues for continuous monitoring of metabolites have promise for clinical applications, but there is a need to understand the effects of the tissue on the concentration-dependent signals. Variants of the tissue-sensor array window chamber described here are powerful tools for visualization of tissues and recording sensor signals while administering physiological challenges. The experimental system permits real-time, nondestructive observation of tissues, eliminates between-subject variance in the signals, and allows repeated studies with implant time. The respective window chamber variants have advantages for improved visualization, minimization of perturbations of normal vascular function, and extension of window chamber lifetime. The observations made possible here may suggest improved designs of implantable sensors to compensate for tissue phenomena.
|
GRANTS
|
|---|
This work was supported by grants from the National Institutes of Health and the Technology for Metabolic Monitoring Initiative of the Department of Defense.
|
DISCLOSURES
|
|---|
D. Gough holds equity interest in and is a scientific advisor to GlySens, Inc., from which sensor arrays and potentiostats were purchased. This arrangement has been approved by the University of California-San Diego in accordance with its conflict of interest policies.
|
ACKNOWLEDGMENTS
|
|---|
We thank Jared Goor for the schematic of the hamster vasculature, Rick Calou for preparation of the skin section and the chamber plate renderings, and Isgard Hueck for preparation of the skin histology cross section.
|
FOOTNOTES
|
|---|
Address for reprint requests and other correspondence: D. A. Gough, Dept. of Bioengineering, Univ. of California-San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0412 (E-mail: dgough{at}ucsd.edu)
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES
|
|---|
- Algire GH. Visualizing cellular behavior in vivo. J Int Chirg 13: 381–384, 1953.
- Arfors KE, Jonsson JA, and McKenzie FN. A titanium rabbit ear chamber: assembly, insertion and results. Microvasc Res 2: 516–518, 1970.[CrossRef][Medline]
- Armour JC, Lucisano JY, McKean BD, and Gough DA. Application of a chronic intravascular blood glucose sensor in dogs. Diabetes 39: 1519–1526, 1990.[Abstract]
- Baker DA and Gough DA. A continuous, implantable lactate sensor. Anal Chem 67: 1536–1540, 1995.[CrossRef]
- Bertera S, Geng X, Tawdrous Z, Bottino R, Balamurugan AN, Rudert WA, Drain P, Watkins SC, and Trucco M. Body window-enabled in vivo multicolor imaging of transplanted mouse islets expressing an insulin-timer fusion protein. Biotechniques 35: 718–722, 2003.[ISI][Medline]
- Branemark-I. Intravascular Anatomy of Blood Cells in Man. Basel: Karger, 1971, p. 4–10.
- Devoisselle JM, Begu S, Tourne-Peteilh C, Desmettre T, and Mordon S. In vivo behaviours of long circulating liposomes in blood vessles in hamster inflammation and septic shock models - use of intravital fluorescence microscopy. Luminescence 16: 73–78, 2001.[CrossRef][ISI][Medline]
- Endrich B, Asaishi K, Gotz A, and Messmer K. Technical report–A new chamber technique for microvascular studies in unanesthetized hamsters. Res Exp Med (Berl) 177: 125–134, 1980.
- Gough DA and Armour JC. Development of the implantable glucose sensor: What are the prospects? and, Why is it taking so long? Diabetes 44: 1005–1009, 1995.[ISI][Medline]
- Jain RK, Munn LL, and Fukumura D. Dissecting tumor pathophysiology using intravital microscopy. Nat Rev Cancer 2: 266–276, 2002.[CrossRef][ISI][Medline]
- Johnson KW, Mastrototaro JJ, Howey DC, Brunelle RL, Burden-Brady PL, Bryan NA, Andrew CC, Rowe HM, Allen DJ, Noffke BW, MacMaham WC, Morff RJ, Lipson D, and Nevin RS. In vivo evaluation of an electroenzymatic glucose sensor implanted in subcutaneous tissue. Biosens Bioelectron 7: 709–714, 1992.[CrossRef][ISI][Medline]
- Kerger H, Ivo P, Torres F, Rivas M, Winslow RM, and Intaglietta M. Systemic and subcutaneous microvascular oxygen tension on conscious Syrian golden hamsters. Am J Physiol Heart Circ Physiol 268: H802–H810, 1995.[Abstract/Free Full Text]
- Kerner W, Kiwit M, Linke B, Keck FS, Zier H, and Pfeiffer EF. The function of hydrogen-peroxide-detecting electroenzymatic glucose electrode is markedly impaired in human subcutaneous tissue and plasma. Biosens Bioelectron 8: 473–482, 1993.[CrossRef][ISI][Medline]
- Leunig M, Yuan F, Menger MD, Boucher Y, Goetz AE, Messmer K, and Jain RK. Angiogenesis, microvascular architecture, microhemodynamics, and interstitial fluid pressure during early growth of human adenocarcinoma LS174T in SCID Mice. Cancer Res 52: 6553–65560, 1992.[Abstract/Free Full Text]
- Lucisano JY, Armour JC, and Gough DA. In vitro stability of an oxygen sensor. Anal Chem 59: 736–739, 1987.[Medline]
- Makale MT, Jablecki MC, and Gough DA. Mass transfer and gas-phase calibration of implanted oxygen sensors. Anal Chem 76: 1773–1776, 2004.[Medline]
- Makale M, Lin J, Calou R, Tsai A, Chen P, and Gough DA. Tissue window chamber system for validation of implanted oxygen sensors. Am J Physiol Heart Circ Physiol 284: H2288–H2294, 2003.[Abstract/Free Full Text]
- Menger MD, Laschke MW, and Vollmer B. Viewing the microcirculation through the window: Some twenty years experience with the hamster dorsal skinfold chamber. Eur Surg Res 34: 83–91, 2002.[CrossRef][ISI][Medline]
- Metzger M, Leibowitz G, Wainstein J, Glasser B, and Raz I. Reproducibility of glucose measurements using the glucose sensor. Diabetes Care 25: 1185–1192, 2002.[Abstract/Free Full Text]
- Moatti-Sirat D, Capron F, Poitout V, Reach G, Bindra DS, Zhang Y, Wilson GS, and Thevenot DR. Towards continuous glucose monitoring: In vivo evaluation of a miniaturized glucose sensor implanted for several days in rat subcutaneous tissue. Diabetologia 35: 224–230, 1992.[CrossRef][ISI][Medline]
- Mordon S, Desmettre T, Devoisselle JM, and Mitchell V. Selective laser photocoagulation of blood vessels in a hamster skin flap model using a specific ICG formulation. Lasers Surg Med 21: 365–373, 1997.[CrossRef][ISI][Medline]
- Pahernik S, Harris AG, Schmitt-Sody M, Kransnici S, Goetz AE, Dellian M, and Messner K. Orthogonal polarization spectral imaging as a new tool for the assessment of antivascular tumor treatment in vivo: A validation study. Br J Cancer 86: 1622–1627, 2002.[CrossRef][ISI][Medline]
- Thome-Duret V, Gangerau MN, Zhang Y, Wilson GS, and Reach G. Modification of the sensitivity of glucose sensors implanted into subcutaneous tissue. Diabetes Metab 22: 174–178, 1996.[ISI][Medline]
- Vranckx JJ, Slama J, Preuss S, Perrez N, Svensjo T, S, Breuing K, Bartlett R, Pribaz J, Weiss D, Eriksson E. Wet wound healing. Plast Reconstr Surg 110: 1680–1687, 2002.[CrossRef][ISI][Medline]
- Yuan F, Leunig M, Berk DA, and Jain RK. Microvascular permeability of albumin, vascular surface area, and vascular volume measured in human adenocarcinoma LS174T using dorsal chamber in SCID mice. Microvasc Res 45: 269–289, 1993.[CrossRef][ISI][Medline]
Copyright © 2005 by the American Physiological Society.
TOP
ABSTRACT
METHODS AND MATERIALS
