Biosensors & Bioelectronics 5 (1990) 351-366
Sensory Transduction and the Mammalian Epidermis
Steven B. Hoath,a c Michael M. Donnellyaf c Raymond E . Boissyb
&
a Department
of Pediatrics, bDepartment of Dermatology, ePerinatal Research Institute, University of Cincinnati, Cincinnati, Ohio, USA
(Received 24 July 1989 ; revised version received 17 November 1989 ; accepted 6 December 1989)
ABSTRACT This paper constitutes, in its main intent, an introduction to the mammalian epidermis as a surface for biosensor applications . In particular, the structure andfunction ofthe epidermis of the newborn rat are examined as a model for studies of the human state. Data are presented illustrating an an isotropic organization ofthe dorsal surface ofthe neonatal rodent with regard to lines of tension and thermal gradients The dependence of the mechanical properties of the epidermis upon calcium is examined by means of an in-vitro assay of epidermal retraction . The potential role of keratin tonofilaments as piezoelectric and pyroelectric elements in the epidermis is introduced and the spatial alignment of these macromolecular arrays is demonstrated to be a function ofphysiological tensions. Thesefindings are discussed in the context of noninvasive epidermal sensors utilized to understand mechanisms of sensory development and physiological regulation . Optoelectronic (infrared) imaging of the dorsal temperature field and the alteration in this field by treatment with epidermal growth factor are presented as examples of this methodologic approach . It is concluded that a detailed examination of the material and physical properties of mammalian epidermis is a reasonable goal ofbiosensor development and research . Hypothetically. such studies may reveal important molecular and cellular mechanisms by which sensory data are transmitted or transduced at the organism-environmental interface . Key words: biosensor, epidermal growth factor, epidermis, keratin, sensory transduction . 351
Biosensors & Bioelectronics 0956-5663/90/$03 .50 ® 1990 Elsevier Science Publishers Ltd, England. Printed in Great Britain .
3 52
Steven B. Hoath, Michael M Donnelly, Raymond E. Boissy
INTRODUCTION Recent advances in microelectronics and computerized high speed data acquisition promise a revolution in the fabrication of miniaturized sensor devices applicable to physiological systems . Parallel progress in the genetic engineering of proteins and the design of membranes with specific molecular transducer capabilities offers the real possibility of constructing 'biosensors' or `smart microstructures' which can be interfaced with the organism in a variety of medical and paramedical settings (Turner et al., 1987). The realization of this potential requires knowledge of the physics and physiology of transducible biological surfaces . The skin constitutes a major interface for noninvasive physiological data retrieval. This introductory paper focuses on potential sensor applications and the wealth of transducible information at the level of the organism's outermost boundary, the epidermis . In particular, the importance of physical `gradients' in the organization of epidermal structure and function is emphasized . A brief list is given of physiologically important epidermal gradients potentially measureable with specific sensor devices . Particular experimental examples are provided to demonstrate the anisotropic organization of tension and temperature in the skin of the neonatal rat. The similarity of human interfollicular epidermis to that of the perinatal rodent is discussed as well as the importance of an animal model for in-vivo testing of epidermal boundary functions . Finally, the idea is explored that the epidermis itself may function as a sensory membrane ; i.e. as a front-line, reactive, organic surface with signal transmission and transduction capabilities important for physiological regulation and response . In this view, parameters of epidermal organization transducible with specific sensor devices (e.g. temperature fields, tension fields, water vapor gradients) are simultaneously critical determinants of the sensory information processed by the organism . In general, it is proposed that a detailed study of epidermal structure and function using novel biosensing techniques may shed new light on factors determining autonomic and central nervous system performance . A concrete example of this approach is the use of infrared imaging to explore the effects of epidermal growth factor (EGF) (Carpenter & Cohen, 1979 ; Cohen, 1986) on the dynamics of postnatal temperature control in the rat . This noninvasive sensing technique has provided clues to an unexpected role for this important tissue growth factor in homeothermic development .
Sensory transduction and the mammalian epidermis
353
MATERIALS AND METHODS Animal handling and preparation of isolated epidermis Timed gestation Sprague-Dawley rats were obtained on the fourteenth day of gestation from Zivic-Miller Laboratories Inc . (Zelienople Park, PA). All animals were housed with a 12 h light/dark cycle and given routine laboratory chow ad libitum . Mothers routinely delivered on day 22 of gestation . Newborn pups were undisturbed until the time of experimentation. Between 24 and 48 h following delivery, pups were removed from the nest and a Sarstedt highlighter was used to demarcate the dorsal skin surface such that longitudinal or transverse areas were obtained measuring exactly 20 mm across (Fig . 1). Following sacrifice by intracardiac injection of Nembutol, the dorsal skin was excised and positioned on the stage of a Mcllwain tissue chopper as described previously (Hoath et al., 1989a) . Uniform strips 1 mm in width were cut and transferred to Hank's balanced salt solution containing 20 m m dithiothreitol, buffered to pH 7 . 3 . Strips were incubated for 20 mins at 296K followed by gentle uniaxial separation of epidermal strips using fine forceps . Isolated epidermal strips were maintained beneath the surface of the culture media and allowed to assume figures of equilibrium (conformations of minimal surface energy) . The degree of tissue retraction was quantified by determining the `winding number' of
11
N
\\\\\\\1101
Transverse
Section
Longitudinal Section
Fig. 1 . Diagrammatic representation of the dorsal surface of the newborn rat The method of demarcating longitudinal versus transverse skin sections is shown . Following preparation, epidermal strips were utilized for in-vitro assay of intraepidermal tensions.
354
Steven B. Hoath, Michael M Donnelly, Raymond E . Boissy
the epidermal strips ; i .e. the number of coils (360° turns) formed between the demarcated ends of the tissue section (Hoath et al., 1989a). Electron microscopy Dorsal skin was dissected from newborn Sprague-Dawley rat pups and uniform 20 mm X 1 mm longitudinal strips of whole skin were prepared as described above . Skin strips were selected 3 mm to the left or right of the midline and manually extended to 25 mm (125% of their resting invivo length) . The ends of the strips were tethered to Whatman filter paper with collodion and the strips submerged overnight at 277K in Karnovsky's fixation (Karnovsky, 1965) prior to processing for electron scopy as described previously (Hoath et al., 1989b) . Following processing, adjacent areas of the same strip were thin-sectioned in both longitudinal (rostral-caudal) as well as transverse (right-to-left) directions, stained with uranyl acetate and lead citrate and viewed with a JEOL IOOS electron microscope . Infrared imaging Infrared imaging of the dorsal skin temperature field was performed with one-day-old rats using either a non-contact infrared thermometer (Linear Laboratories, Sunnyvale, CA) or a Model 600 Infrared Imaging Radiometer (Inframetrics, Bedford, MA) . Littermate animals were removed from the nest prior to experimentation and pre-equilibrated for 60 rains in a C-100 incubator (Air-Shields, Hatboro, PA) at 307K ambient temperature . Pups were injected subcutaneously over the low flank region with approximately 10,01 of normal saline or an equal volume containing 500 ng EGF/gram body weight . Temperatures were measured every 15 mins over the midscapular region and results in the treatment group were compared to control pups to generate thermal difference graphs . Previous experiments supported the hypothesis of a maximal EGF effect approximately 60 mins following treatment . This time point, therefore, was chosen a priori for statistical testing. In separate experiments, one-day-old rat pups were pre-equilibrated as described above at nest temperature ; i.e . 307K Forty minutes following injection, pups were removed from the incubator and infrared imaging performed as the animals cooled in an ambient environment of 296K Images were analyzed using the Model 600 scanner and high resolution photographs were obtained using an IBM PC and video image recorder .
Sensory transduction and the mammalian epidermis
355
RESULTS Epidermal tension bioassay
Incubation of isolated epidermal strips from newborn rats in tissue culture media resulted in rapid in-vitro morphogenesis . Figure 2 shows the different geometric forms obtained using transverse versus longitudinal epidermal strips . These different morphologies were interpreted to result from the abrupt change in boundary conditions imposed by excision and sample preparation . Morphogenetic differences obtained in littermate animals solely on the basis of biopsy orientation pre-
(a)
(b) Fig. 2 . Spontaneous coiling patterns formed in-vitro by 2-cm strips of neonatal rat epidermis. Epidermal isolates were prepared as described in the methods section; i.e. dorsal skin biopsies were cut in either : (a) a parallel (longitudinal) or (b) perpendicular (transverse) direction relative to the sagittal midline . Following submersion in tissue culture media, longitudinal sections rapidly retracted into tight, helical forms whereas transverse sections remained extended with little overall retraction and poor coil formation .
356
Steven B. Hoath, Michael M. Donnelly, Raymond E. Boissy
sumably reflect differences in in-vivo tissue organization and lines of tension (Hoath et al., 1989a). In general, the transversely-cut strips remained extended or exhibited slow, sparse coiling at their extremities . Longitudinal strips obtained from rostral-caudal epidermal sections, however, rapidly retracted into tightly coiled, helical forms . Measurement of coil formation (the number of 360° turns/2 cm strip) showed a significant increase in the longitudinal versus the transverse group (5 . 7± 0 .2 turns/strip versus 2 . 7 ± 0. 3 turns/strip, p < 0. 001, N = 24, Mean ± SE). Epidermal separation elicited by ethylenediamine-tetraacetic acid (EDTA), heat or collagenase yielded similar results ; i.e. coiling of longitudinal epidermal strips was greater than for transverse strips (data not shown) . Incubation of whole skin strips in 0, 0 . 1,1 .0, and 10. 0 mm CaC12 for 60 mins prior to separation of the epidermis with dithiothreitol yielded mean coil numbers of 5 . 7, 5 . 9, 6 . 6,* and 8 . 0* (all standard errors
Sensory Transduction and the Mammalian Epidermis
Steven B. Hoath,a c Michael M. Donnellyaf c Raymond E . Boissyb
&
a Department
of Pediatrics, bDepartment of Dermatology, ePerinatal Research Institute, University of Cincinnati, Cincinnati, Ohio, USA
(Received 24 July 1989 ; revised version received 17 November 1989 ; accepted 6 December 1989)
ABSTRACT This paper constitutes, in its main intent, an introduction to the mammalian epidermis as a surface for biosensor applications . In particular, the structure andfunction ofthe epidermis of the newborn rat are examined as a model for studies of the human state. Data are presented illustrating an an isotropic organization ofthe dorsal surface ofthe neonatal rodent with regard to lines of tension and thermal gradients The dependence of the mechanical properties of the epidermis upon calcium is examined by means of an in-vitro assay of epidermal retraction . The potential role of keratin tonofilaments as piezoelectric and pyroelectric elements in the epidermis is introduced and the spatial alignment of these macromolecular arrays is demonstrated to be a function ofphysiological tensions. Thesefindings are discussed in the context of noninvasive epidermal sensors utilized to understand mechanisms of sensory development and physiological regulation . Optoelectronic (infrared) imaging of the dorsal temperature field and the alteration in this field by treatment with epidermal growth factor are presented as examples of this methodologic approach . It is concluded that a detailed examination of the material and physical properties of mammalian epidermis is a reasonable goal ofbiosensor development and research . Hypothetically. such studies may reveal important molecular and cellular mechanisms by which sensory data are transmitted or transduced at the organism-environmental interface . Key words: biosensor, epidermal growth factor, epidermis, keratin, sensory transduction . 351
Biosensors & Bioelectronics 0956-5663/90/$03 .50 ® 1990 Elsevier Science Publishers Ltd, England. Printed in Great Britain .
3 52
Steven B. Hoath, Michael M Donnelly, Raymond E. Boissy
INTRODUCTION Recent advances in microelectronics and computerized high speed data acquisition promise a revolution in the fabrication of miniaturized sensor devices applicable to physiological systems . Parallel progress in the genetic engineering of proteins and the design of membranes with specific molecular transducer capabilities offers the real possibility of constructing 'biosensors' or `smart microstructures' which can be interfaced with the organism in a variety of medical and paramedical settings (Turner et al., 1987). The realization of this potential requires knowledge of the physics and physiology of transducible biological surfaces . The skin constitutes a major interface for noninvasive physiological data retrieval. This introductory paper focuses on potential sensor applications and the wealth of transducible information at the level of the organism's outermost boundary, the epidermis . In particular, the importance of physical `gradients' in the organization of epidermal structure and function is emphasized . A brief list is given of physiologically important epidermal gradients potentially measureable with specific sensor devices . Particular experimental examples are provided to demonstrate the anisotropic organization of tension and temperature in the skin of the neonatal rat. The similarity of human interfollicular epidermis to that of the perinatal rodent is discussed as well as the importance of an animal model for in-vivo testing of epidermal boundary functions . Finally, the idea is explored that the epidermis itself may function as a sensory membrane ; i.e. as a front-line, reactive, organic surface with signal transmission and transduction capabilities important for physiological regulation and response . In this view, parameters of epidermal organization transducible with specific sensor devices (e.g. temperature fields, tension fields, water vapor gradients) are simultaneously critical determinants of the sensory information processed by the organism . In general, it is proposed that a detailed study of epidermal structure and function using novel biosensing techniques may shed new light on factors determining autonomic and central nervous system performance . A concrete example of this approach is the use of infrared imaging to explore the effects of epidermal growth factor (EGF) (Carpenter & Cohen, 1979 ; Cohen, 1986) on the dynamics of postnatal temperature control in the rat . This noninvasive sensing technique has provided clues to an unexpected role for this important tissue growth factor in homeothermic development .
Sensory transduction and the mammalian epidermis
353
MATERIALS AND METHODS Animal handling and preparation of isolated epidermis Timed gestation Sprague-Dawley rats were obtained on the fourteenth day of gestation from Zivic-Miller Laboratories Inc . (Zelienople Park, PA). All animals were housed with a 12 h light/dark cycle and given routine laboratory chow ad libitum . Mothers routinely delivered on day 22 of gestation . Newborn pups were undisturbed until the time of experimentation. Between 24 and 48 h following delivery, pups were removed from the nest and a Sarstedt highlighter was used to demarcate the dorsal skin surface such that longitudinal or transverse areas were obtained measuring exactly 20 mm across (Fig . 1). Following sacrifice by intracardiac injection of Nembutol, the dorsal skin was excised and positioned on the stage of a Mcllwain tissue chopper as described previously (Hoath et al., 1989a) . Uniform strips 1 mm in width were cut and transferred to Hank's balanced salt solution containing 20 m m dithiothreitol, buffered to pH 7 . 3 . Strips were incubated for 20 mins at 296K followed by gentle uniaxial separation of epidermal strips using fine forceps . Isolated epidermal strips were maintained beneath the surface of the culture media and allowed to assume figures of equilibrium (conformations of minimal surface energy) . The degree of tissue retraction was quantified by determining the `winding number' of
11
N
\\\\\\\1101
Transverse
Section
Longitudinal Section
Fig. 1 . Diagrammatic representation of the dorsal surface of the newborn rat The method of demarcating longitudinal versus transverse skin sections is shown . Following preparation, epidermal strips were utilized for in-vitro assay of intraepidermal tensions.
354
Steven B. Hoath, Michael M Donnelly, Raymond E . Boissy
the epidermal strips ; i .e. the number of coils (360° turns) formed between the demarcated ends of the tissue section (Hoath et al., 1989a). Electron microscopy Dorsal skin was dissected from newborn Sprague-Dawley rat pups and uniform 20 mm X 1 mm longitudinal strips of whole skin were prepared as described above . Skin strips were selected 3 mm to the left or right of the midline and manually extended to 25 mm (125% of their resting invivo length) . The ends of the strips were tethered to Whatman filter paper with collodion and the strips submerged overnight at 277K in Karnovsky's fixation (Karnovsky, 1965) prior to processing for electron scopy as described previously (Hoath et al., 1989b) . Following processing, adjacent areas of the same strip were thin-sectioned in both longitudinal (rostral-caudal) as well as transverse (right-to-left) directions, stained with uranyl acetate and lead citrate and viewed with a JEOL IOOS electron microscope . Infrared imaging Infrared imaging of the dorsal skin temperature field was performed with one-day-old rats using either a non-contact infrared thermometer (Linear Laboratories, Sunnyvale, CA) or a Model 600 Infrared Imaging Radiometer (Inframetrics, Bedford, MA) . Littermate animals were removed from the nest prior to experimentation and pre-equilibrated for 60 rains in a C-100 incubator (Air-Shields, Hatboro, PA) at 307K ambient temperature . Pups were injected subcutaneously over the low flank region with approximately 10,01 of normal saline or an equal volume containing 500 ng EGF/gram body weight . Temperatures were measured every 15 mins over the midscapular region and results in the treatment group were compared to control pups to generate thermal difference graphs . Previous experiments supported the hypothesis of a maximal EGF effect approximately 60 mins following treatment . This time point, therefore, was chosen a priori for statistical testing. In separate experiments, one-day-old rat pups were pre-equilibrated as described above at nest temperature ; i.e . 307K Forty minutes following injection, pups were removed from the incubator and infrared imaging performed as the animals cooled in an ambient environment of 296K Images were analyzed using the Model 600 scanner and high resolution photographs were obtained using an IBM PC and video image recorder .
Sensory transduction and the mammalian epidermis
355
RESULTS Epidermal tension bioassay
Incubation of isolated epidermal strips from newborn rats in tissue culture media resulted in rapid in-vitro morphogenesis . Figure 2 shows the different geometric forms obtained using transverse versus longitudinal epidermal strips . These different morphologies were interpreted to result from the abrupt change in boundary conditions imposed by excision and sample preparation . Morphogenetic differences obtained in littermate animals solely on the basis of biopsy orientation pre-
(a)
(b) Fig. 2 . Spontaneous coiling patterns formed in-vitro by 2-cm strips of neonatal rat epidermis. Epidermal isolates were prepared as described in the methods section; i.e. dorsal skin biopsies were cut in either : (a) a parallel (longitudinal) or (b) perpendicular (transverse) direction relative to the sagittal midline . Following submersion in tissue culture media, longitudinal sections rapidly retracted into tight, helical forms whereas transverse sections remained extended with little overall retraction and poor coil formation .
356
Steven B. Hoath, Michael M. Donnelly, Raymond E. Boissy
sumably reflect differences in in-vivo tissue organization and lines of tension (Hoath et al., 1989a). In general, the transversely-cut strips remained extended or exhibited slow, sparse coiling at their extremities . Longitudinal strips obtained from rostral-caudal epidermal sections, however, rapidly retracted into tightly coiled, helical forms . Measurement of coil formation (the number of 360° turns/2 cm strip) showed a significant increase in the longitudinal versus the transverse group (5 . 7± 0 .2 turns/strip versus 2 . 7 ± 0. 3 turns/strip, p < 0. 001, N = 24, Mean ± SE). Epidermal separation elicited by ethylenediamine-tetraacetic acid (EDTA), heat or collagenase yielded similar results ; i.e. coiling of longitudinal epidermal strips was greater than for transverse strips (data not shown) . Incubation of whole skin strips in 0, 0 . 1,1 .0, and 10. 0 mm CaC12 for 60 mins prior to separation of the epidermis with dithiothreitol yielded mean coil numbers of 5 . 7, 5 . 9, 6 . 6,* and 8 . 0* (all standard errors
