Middle-ear development VIStructural maturation of the rat conducting apparatus.код для вставкиСкачать
THE ANATOMICAL RECORD 239:475-484 (1994) Middle-Ear Development VI: Structural Maturation of the Rat Conducting Apparatus WAYNE M ZTMMER, DEBORAH F ROSIN, AND JAMES C. SAUNDERS Department of Otorhinolarvngologv. Head and Neck Surgery 1Jni11~rsity of Punnylrmncrr, Philadelphia, Pennsylvania ABSTRACT Background: The contribution of middle-ear development to the overall development of hearing has not been explored in great detail. This presentation describes the maturation of conductive elements in the rat middle ear, and provides the basis on which future studies of middle-ear functional development will follow. Methods: The middle-ear apparatus was examined at nine different ages (between 1 and 80 days postpartum) in Long Evans rats. A t each age elements of the conducting apparatus were observed with either light or scanning electron microscopy (SEM), and quantitative measurements were made from video enhanced photomicrographs. Tympanic membrane area and cone depth, the length of the malleus and incus arms, ossicular weight, stapes foot plate and oval window areas, and bulla volume were all measured. Development of the area and lever ratios were derived from these measurements. The data were fitted to exponential equations and the time in days required to reach 90% of the adult level determined. ResuZts: The pars tensa achieved 90% of total area by 17 days. The oval window achieved the 90% criterion by 13 days, while the area ratio was within 10%of its adult size by 8 days. The ossicles took between 26 and 34 days, while bulla volume took 59 days to reach the 90% level. Conclusions: Middle-ear growth was very orderly and systematic in the data reported. When maturation of the area ratio was considered against development of the endocochlear potential or the round window compound action potential, it was clear that the growth of this important aspect of the middle ear preceded the onset of cochlear function. o 1994 WiIey-Liss, Inc. Key words: Middle ear, Auditory, Hearing development, Ossicles, Tympanic membrane, Rat (Long Evans strain) A number of animal models are currently being used to study auditory development and these include the chick, mouse, gerbil, guinea pig, and kitten. These species are interesting because their auditory systems, at birth, may be either highly developed (as in “precocial” species like the chick or guinea pig) or poorly developed (as in “altricial” species like the mouse or kitten). The rat is another altricial laboratory animal for which there is a growing literature on the development of hearing. In this species the first indication of cochlear function is seen in the endocochlear potential which emerges between the 8th to 10th day after birth (Rybak et al., 1992). This is followed shortly by the first soundevoked potentials in the cochlea between 11 and 12 days. Adult-like compound action potentials (CAP) in the cochlea were observed between the 17th and 20th days after birth (Rybak et al., 1992; Uziel et al., 1981). The development of hearing in the peripheral region of the auditory system represents a n interplay between the maturation of structures that permit the onset and development of functional capability (Burda, 1985). The middle-ear system and its capacity to conduct vi0 1994 WILEY-LISS, INC. brational energy to the cochlea is important in this process. Indeed, the serial nature of information processing in the peripheral ear suggests that the conductive apparatus may play a n essential role in the development of more central measures of auditory ability. If sound does not reach the cochlear analyzer, regardless of the inner ear’s ability to process vibrational energy, hearing capacity will be severely limited (Saunders et al., 1993). This investigation describes the morphological development of the conductive apparatus in the rat, and is a precursor to other investigations that will consider the functional development of sound transmission in the middle ear and its relation to the onset and maturation of hearing capacity in this species (Igic et al., 1994). Received January 12, 1994; accepted March 16, 1994. Address reprint requests to Dr. James C. Saunders, 5 SilversteinORL, 3400 Spruce St., Philadelphia, PA 19104. 476 W.M. ZIMMER ET AL. METHODS Animals and Groups Pregnant female rats of the Long Evans strain were obtained from commercial breeders and held in the animal quarters at the University of Pennsylvania School of Medicine. The birth of pups was determined with a n accuracy of 2 6 hours, and these were maintained and nursed by their mothers until the age of sacrifice. Animals were divided into nine groups aged 1,3,6,12,16, 22, 34, 60, and 80 days after birth. For each of the middle-ear variables examined (see below) a t least five different specimens from separate animals were sampled in each age group. Tissue Preparation The animals in each group were sacrificed with a n intraperitoneal injection of ethyl carbamate (Urethane). Following decapitation the skin, muscle, and external auditory meatus were removed taking care to avoid damage to the bulla and tympanic ring. The skull was split sagitally in animals 12 days or older and the temporal bones on both sides were isolated. A 1 mm hole was made in the inferior lateral aspect of the bulla and a solution of 10%phosphate buffered formalin (pH 7.0) was injected into the cavity. The bones were then immersed in the 10% formalin for at least 48 hours. After fixation the specimen was rinsed with water and a small wick was placed into the bulla hole to remove gross fluid. The specimen was then set aside and allowed to dry in air overnight. Paraffin wax was used to seal the eustachian tube orifice of the dry bones so that the bulla would be fluid tight for volume measurements. The specimen was then weighed on a mass balance to a n accuracy of 10 pg. Next a watedsurfactant (Photoflo, Kodak Co., Rochester, NY) mixture was gently injected into the bulla hole. The surfacant facilitated filling the bulla and assured that all the recesses of the cavity were filled. Water was injected into the cavity until the meniscus surrounded the injection needle. The needle was then removed and the specimen immediately reweighed. The weight of the fluid was obtained by subtracting the bulla empty from the bulla filled weight. The volume was calculated by dividing the fluid weight by fluid density (the waterlsurfacent had a density of 1.00035 glcc). This method was a n effective and accurate means of determining middle-ear volume without damaging the elements of the conductive system (Vrettakos et zl., 1988; Cohen et al., 1992a). After the bulla volume was determined, the specimen was further dissected. The temporal bone was viewed with a stereomicroscope at magnifications between 20 and 50 x . The bone forming the terminal zone of the external auditory canal (DeMaio and Tonndorf, 1978) was removed with a high speed drill using a small diamond burr (3.0 mm). This allowed us to clearly expose the tympanic ring. The specimen was placed on the stage of a Nikon Multiphot micro/macro photography system equipped with a TV camera and image enhancing electronics. A calibrated millimeter rule was placed alongside the specimen and this was used in the subsequent measurements of membrane area. A 10.3 x 7.5 cm video micrograph was obtained using a Mitsubishi (Model P-4Ou) video printer. Re- flected light was used to photograph the tympanic membrane and the magnification was adjusted until the membrane just filled the area of the video micrograph. These same video microscopic procedures were used to obtain video pictures of all the middle-ear components. The tympanic membrane and most of the lateral extent of the bulla was then removed to reveal the ossicles contained in the bulla cavity. Pictures of the ossicles were obtained either in situ under reflected-light illumination, or after removal using transmitted light. The latter approach proved useful for pups 12 days old or younger, where the ossicles were transparent and there was little contrast between the developing bony structure and the surrounding mesenchyma. These video prints were again used to measure the lengths of the ossicles and their lever arms. I n some specimens the tissue was prepared for scanning electron microscopy (SEM) using procedures described elsewhere (Huangfu and Saunders, 1983). Polaroid micrographs taken with the SEM were used to make comparisons with the video prints, and to examine middle-ear structures in greater detail when needed. Length or area measurements from SEM or video photos were the same. Fleischer (1978) described the morphology of many rodent middle ears as exhibiting the so-called "microtype" organization. This design has two characteristic features: the malleus is fused to the tympanic ring a t the gonial, and there is a large mass a t the head of the malleus called the orbicular apophysis (Cocherell e t al., 1914; Huangfu and Saunders, 1983; Saunders and Garfinkle, 1983). Fleischer (1978) suggested t h a t the ossicles in the microtype middle ear exhibit two modes of vibration. Two modes were indeed identified in a n examination of sound driven tympanic membrane and ossicular motion in the mouse middle ear (Saunders and Summers, 1982). The r a t also exhibits a microtype organization, and for the purpose of examining the lever arms in this species, two axes of rotation were defined. The first axis is indicated by line AB in Figure 1 and extends from the gonial through the incudomalleal joint to the posterior ligament of the incus. The corresponding malleus lever (Ml) extends from the tip of the umbo to form a 90" angle with the AB intercept. The incus lever (11)extends from the distal tip of the incus at the lenticular process and also forms a 90" angle with the AB axis. The second presumed axis, as described for the mouse (Saunders and Summers, 1982; Fleischer, 1978), extends from the gonial to the center of the orbicular apophysis (line CD in Fig. 1). The malleus lever (M2) for this axis extends from the tip of the umbo to form a 90" angle with CD. The incus lever (12) extends from the distal tip and forms a 90" angle with CD. The functional significance of these axes in the rat remains to be determined, but both were traced during development for this presentation. Prior to removal of the incus and malleus, the incudostapedial joint was gently separated, and the tensor tympani tendon carefully cut. This provided excellent exposure to the stapes and the stapedial artery coursing between the crura. Fine microscissors were used to cut the artery as i t passed in a n antero-inferior direction between the stapes crura. A fine needle was used to mobilize and pull the artery through the crura pos- 477 MIDDLE-EAR DEVELOPMENT B +lrnrnd / C, / \ // \ / / / / / A' Fig. I . This line drawing of the adult rat ossicular system is viewed from above with the tympanic membrane removed. The two axes of rotation (AB, CD) are indicated as are the respective lengths of the lever arms for axis AB (Ml, 11) and for axis CD (M2, 12). We have chosen to measure the lever arms as a line segment perpendicular to the axes of rotation. The long arm of the malleus is defined by M2 while the long arm of the incus is defined by 11. Ma1 = malleus; In = incus; St = stapes; OA = Orbicular apophysis; Gon = gonial. The tympanic ring is seen as the outer perimeter of the drawing. tero-superiorly. The stapedius tendon was cut, and a pick was used to "rock" the stapes back and forth in order to rupture the annular ligament. This greatly facilitated removal of the intact stapes bone. Because of the extremely soft, nearly gelatinous texture of the stapes in 12-day-old and younger pups, this maneuver was vitally important in removing the ossicle without crushing it. Video micrographs of the stapes and oval window were then obtained. The remaining ossicles were disarticulated and allowed to dry for a t least 24 hours. The malleus and incus were then weighed on a mass balance to a n accuracy of 10 pg. Rat pups younger than 12 days were prepared differently to facilitate dissection of the tympanic structures in these very immature, technically challenging age groups. The animals were decapitated and the skulls were maintained intact. The external auditory canal and the epidermis in the immediate vicinity were removed leaving most of the head covered with skin. This provided substance to the otherwise soft skull and prevented damage due to crushing during the dissection. Minimal soft tissue was removed in exposing the tympanic ring and the bony structures that defined the tympanic cavity. Dissection was performed with forceps and blunt instruments as all structures were soft and cartilaginous. It was impossible to assess bulla volume in these pups because of the immaturity of this cavity and the fact that it was filled with mesenchyme. Photomicrographs of the ossicles were obtained immediately after their removal to avoid desiccation. Structural Analysis The purpose of the dissections described above was to expose the important components of the middle-ear apparatus so that they could be accurately photographed and measured. We wanted to determine the development of the tympanic membrane area, including that of its two divisions, the pars tensa and pars flaccida. The 478 W.M. ZIMMER ET AL. area of the stapes foot plate and the oval window was also measured. In addition, the depth of the tympanic membrane cone, the lengths of the ossicular lever arms (the long arm of the malleus and incus), and the volume of the middle-ear cavity (the bulla) were also quantified with increasing postnatal age. The identified structure was visualized under the Multiphot system and carefully oriented perpendicular to the optical axis of the microscope. Video image analysis hardware and software (Model OC-200, Correco, Inc., Ville St.-Laurent, Quebec, Canada) was used to enhance contrast and capture the image, and a video photo of the image was then produced as described earlier. The video picture was then placed on a digitizing tablet and the area of interest circumscribed with a sensing pointer. Computer software (Sigma Scan by Jandel Scientific, Inc., Sausalito, CA) integrated with the digitizing tablet was then used to calculate either areas, lengths, or angles. The lengths or area measurements were corrected for their magnification with respect to the millimeter rule which accompanied each video print and each of these measurements was expressed a s millimeters or square millimeters. RESULTS The mean and standard deviation for each middleear parameter within all age groups were calculated. A regression line of the form: y = A (1 - e( - t h ) ) + B was calculated for each set of data. The estimated size of the structure a t birth is given by B, t is the number of days after birth, A is the asymptotic or adult size of the structure, and T is the time constant of the regression line. This regression analysis permitted quantitative comparisons among the development of each middle-ear structure examined by determining the number of days needed to achieve 90% of the adult size (the approximate asymptote in each equation). General Observations The skull of the rat was not fused until about 12-16 days after birth. The tympanic cavity was virtually non-existent through 6 days of age since the medial wall (which consisted of the cochlea and vestibule) was in contact with the tympanic ring and membrane. At this stage, mesenchyme filled the tympanum and surrounded and adhered to the middle-ear elements. The homogeneous appearance of all the conductive structures made visual identification extremely difficult in these early age groups, and the ossicles could only be identified after carving them out of the “mesenchymal gravy.” Indeed, identification was only possible because the dissections were performed sequentially beginning with the oldest and then progressing to younger groups. This allowed us to acquire a sense of how the various middle-ear structures were oriented and where they were located before tackling the more challenging, young animals. The tympanic cavity underwent impressive qualitative and quantitative changes during development, and these are not presented. Tympanic Membrane Morphologic development proceeded rapidly through the first 20 days. Between 1 and 3 days of age the tympanic membrane (TM) was transparent, thin, and slightly convex in all places except where fused to the malleus. The malleus itself was not clearly discernible except for the influence it had on the morphology of the membrane. This is seen in the top panel of Figure 2 where the long process of the malleus a t 1 day of age produces a depression on a n otherwise flat surface. There are no other discernible features of the ossicles visible. The TM served to contain the gelatinous mesenchyme, which blanketed the floor of the cavity and surrounded the ossicles. As time progressed, the TM became more fibrous and concave, and by 12 days this membrane formed the lateral wall of the air-filled bulla. By 22 days the TM achieved its adult appearance which is presented in the lower panel of Figure 2. The transparency of the membrane is now apparent with the ossicles below clearly visible. The average areas of the total tympanic membrane, as well as the components of pars tensa and pars flaccida, were calculated for each of the age groups and are presented in Figure 3. The vertical bars give a n indication of one standard deviation above or below the mean. The total tympanic membrane area, as well as that of the pars tensa were between 48% and 49% of adult size at 1 day. These expanded to 90% of adult size by 23 and 17 days of age, respectively. As Figure 3 shows, the pars tensa matured earliest and was responsible for most of the expansion of the tympanic membrane. The pars flaccida was 37% (0.94 mm2) of asymptotic size at 1 day and exhibited a slower rate of maturation, reaching 90% of adult size (2.25 mm2) by 49 days postpartum. The expansion of pars flaccida was responsible for most of the overall tympanic membrane area expansion after 22 days of age. The development of tympanic membrane cone depth is shown in Figure 4 and the insert indicates how this variable was measured. The cone depth was determined from photomicrographs of the umbo taken in a plane parallel to the tympanic ring. Although the umbo was clearly below the plane of the tympanic ring a t 6 days of age, it did not form the apex of a cone since the tympanic membrane was, for the most part, completely flat. The malleus was connected to the tympanic membrane by a thin fold which appeared to expand progressively starting a t approximately 6 days. During the next 6-10 days this process resulted in formation of the conically shaped adult tympanic membrane. When the cone was first identified clearly, it was 67% of adult depth and reached 90% when the pup was 15 days old. The exponential time constants of growth for pars tensa area (see Fig. 3) and cone depth were very similar, suggesting the uniform maturation of the pars tensa in three dimensions. It is worth noting that previously we have corrected for the conical shape of the chick tympanic membrane in calculating its surface area (Cohen et al., 1992a). In the case of the gerbil (Cohen et al., 1992b), and now with the present data for the rat, this correction has proven very frustrating. The complex shape of the pars tensa cone, due to the long arm of the malleus being attached to the membrane along its length, and the fact that the umbo is offset from the exact center of pars tensa, made a n accurate calculation of the conical area very difficult. One should realize that the tympanic membrane areas reported in Figure 3 modeled the 479 MIDDLE-EAR DEVELOPMENT Total TM Area Pars Flaccida Pars Tensa 1 l.Omm H 0 40 20 80 60 AGE IN DAYS Fig. 3. The surface area of pars tensa (PT), pars flaccida (PF), and the total tympanic membrane (TM) area, measured in square millimeters, is presented for animals between 1 and 80 days of age. These areas are modeled as flat plates and the depth of the pars tensa cone was not factored into these measures. Each data point is the average of 5 animals and all three measures were obtained in the same animal. The solid line is the exponential best fit to the data. The vertical bars show one standard deviation above, below, or about the mean. Fig. 2. Two photomicrographs of the tympanic membrane taken in reflected light are presented. The top panel illustrates a 1-day-old pup while the bottom panel shows a n 80-day-old animal. The change in size is apparent and in the older animal the transparent tympanic membrane reveals the ossicular structures below. 2 -I -I 0.4 - I z 0.2 I l- membrane a s a flat plate. The increasing depth of the cone indicates t h a t the actual area of the membrane would be slightly larger than that presented. Oval Window, and Stapes Foot Plate Figure 5 illustrates the changes in oval window and stapes foot plate areas. The oval window area increased by 50% from 0.24 mm2 a t 1 day to its adult size (0.36 mm2) at 60 days after birth. The stapes foot plate underwent a n expansion of 76% from 0.21 mm2 a t 1day to 0.37 mm2 at 60 days. The oval window and foot plate areas reached 90% of their asymptotic sizes by 13 and 17 days, respectively. The development of the oval window and stapes foot plate area plotted in Figure 5 demonstrated a n interesting phenomenon. As might be expected, the surface area of the oval window, prior to 15 days, exceeded that of the stapes foot plate, with the space between them occupied by the annular ligament. However, after 16 days a reversal occurred, with the foot plate area exceeding the oval window area by approximately 5%.An explanation for this latter observation lies in the defi- , Tympanic Ring c Ii Uhbo t a W 0.0 1 f W - ? I I 0 n = 5 J I 20 I I 40 I I 60 I I 1 I 80 AGE IN DAYS Fig. 4. Changes in the conical depth of pars tensa is presented as a function of age. The insert in the figure illustrates how the depth was estimated. The solid line represents the exponential fit to the data. nition we used for oval window and foot plate areas. The oval window was defined by the opening into the vestibule. The stapes foot plate area was defined by its outer perimeter when viewed from below. In 1-day-old specimens the sides of the foot plate were a t nearly right angles to the under surface of the plate (Fig. 6A). In the adult specimen the lateral walls of the foot plate formed a n acute angle of nearly 30" to the lower surface, and this relationship is seen in the adult specimen 480 W.M. ZIMMER ET AL. 80 DAYS I w 5 1 DAY 0.35 1 A w E 0.30 U 0 3 Oval Window v Foot Plate a v, Z - 0.25 a I TI [ i w w 0.20 0 1.0 mm n = 5 I 20 I 40 8 60 I 1 80 AGE IN DAYS Fig. 5. The changing area of the stapes foot plate and the oval window area are shown during an 80-day interval from birth. The solid lines are the exponential fits to the data. The fact that the oval window appears to be smaller than the foot plate beyond about 15 days of age is explained in the text. The vertical bars (one standard deviation) are shown in only one direction and for only the oval window to keep from cluttering the figure. Foot plate variability was nearly the same as that of the oval window. of Figure 6A. Thus, the perimeter of the oval window opening in adults is smaller than the outermost dimension of the foot plate. This aspect of foot plate development is clearly seen in the micrograph of Figure 6B, where the right edge of the foot plate forms a n acute angle with the actual oval window opening. This observation is consistent with earlier reports on annular ligament development in the chick and gerbil. Cohen e t al. (1992a,b) have suggested that the ligament progressively decreases in size to accommodate the foot plate, which, due to its slower maturational rate, continues to expand after the oval window reaches its adult size. Bulla Volume The changes in bulla volume were the most dramatic of all the middle-ear aspects examined. At 3 days after birth the middle-ear cavity consisted of a tiny mesenchyme filled space enveloping the immature ossicles. Indeed, much of the medial wall was in contact with the tympanic membrane and tympanic ring. Over the next 9 days these structures separated rapidly and by 12 days all mesenchyme was absorbed allowing reproducible measurements of bulla volume. Our measures showed that bulla volume expanded by 235% from 0.018 cc at 12 days of age to 0.061 cc in the adult specimens (Fig. 7). The exponential fit to these data showed it to be the slowest maturing aspect of the middle-ear and after 59 days bulla volume was 90% of its asymptotic level. Malleus and incus The ossicles underwent marked qualitative changes from a soft and gelatinous composition with very high water content a t 3 days to a firm and bony appearance I 1.0mm I Fig. 6. A The two line drawings depict the shape of the adult and 1-day-old stapes in the rat. The 1-day-old specimen only retained its shape in water since it was formed entirely of cartilage. The scanning electron micrograph (B) shows the adult stapes in situ. ST = stapedius muscle; FP = foot plate of the stapes; SA = stapedial artery; OA = orbicular apophysis; TT = tensor tympani; C = the incudostapedial joint has separated. by 22 days. By 6 days after birth the first signs of ossification were identified a s a diffuse speckling uniformly distributed throughout the ossicles. These “speckles” represented the centers of ossification. The ossification became concentrated in the cortical margins of the ossicles by 12 days, and by 22 days they were fully mineralized. Prior to 12 days after birth the ossicles had a n extremely high water content and were observed to shrival into formless masses less than 50% of their original size upon even modest dehydration. The severe loss of mass after drying prevented us from accurately measuring the weight of the malleus prior to 6 days old and that of the incus prior to 12 days. The upper panel of Figure 8 shows the length of the long arm of the malleus and the long arm of the incus measured from video micrographs taken immediately after removal of these structures. These measures extend 481 MIDDLE-EAR DEVELOPMENT 3.5 v, CY 3.0 - W I- I - W I- 0.04 rn 3 0.02 0 w w E -I 2.5 - =Z- 2.0 - I 1.5 - 2 0 0 - Z Q T v, 0.06 5 I f k 0 Malleus Incus Bulla Volume I I 0.00 a n = 5 L 0 20 i 40 L l 60 I I 80 AGE IN DAYS Fig. 7. The bulla volume averaged over 5 samples, is plotted against age in days. The fitted exponential line shows that after 80 days the volume was still increasing slightly. The vertical bars indicate one standard deviation about the mean. from the tip of the arms to the approximate center of the malleus and incus a s defined by axes AB and CD in Figure 1. These arms increased by nearly the same magnitude, 67 and 65%, respectively, from 1 day to their adult size. The arm of the incus, however, achieved 90% of its adult length in 26 days while the malleus took 34 days. The lower panel of Figure 8 depicts the increase in the malleus and incus dry weight. The malleus dry mass grew by 800%from 0.1 pg, when it could first be measured at 6 days, to its adult value of 0.9 pg at 35 days. Nearly half of that increase was realized between 6 and 12 days, when it more than tripled. The weight of the incus could not be determined prior to 12 days, because of its high water content. Interestingly, the exponential functions revealed that the time constants for increasing mass were nearly identical for the malleus and incus, with both structures attaining 90% of their asymptotic size by 36 and 34 days of age, respectively. The relationship between mass and length changes, for the malleus and incus, are shown in Figure 9. We elected to plot these variables as a percentage of adult size so that they could be compared against each other more easily. The time constant describing the maturation of length exceeded that of weight, for each of these ossicles, by a narrow margin, and the incus developed slightly faster than the malleus with regard to both parameters. Maturation of the Elements of Sound Transfer While the process of pressure amplification from the tympanic membrane to the stapes foot plate is complex, most of the amplification can be accounted for quantitatively by the ratio between the tympanic membrane (pars tensa) and oval window areas, and the ratio between the length of the long arms of the malleus and the incus along the AB axis. These are the so-called 0.0 ‘ 0 40 20 AGE IN 60 80 DAYS Fig. 8. The changing length (in millimeters) of the long arm of the malleus and incus (top panel) are plotted against age. These are the M1 and I1 lines in Figure 1.The lower panel shows the development of dry weight for the malleus and incus (in micrograms) over the same age interval. The solid lines show the best fit exponential functions. “areal” and “lever” ratios. The development of the area ratio is plotted in Figure 10, and shows a n increase from 17.0 a t 1 day to 24.8 by 80 days. An area ratio of 22.3 was 90% of the adult value when the pup reached 8 days of age. The lever ratios along the AB and CD axes were calculated by MU11 and M2/12 (see Fig. l), respectively, and are presented a s a function of age in Figure 11. These ratios show a relatively constant value during maturation, and essentially fluctuate about the adult values. This reflects the fact that the relative lengths of the malleus and incus lever arms increased simultaneously (see Fig. 8). The mean lever ratios for the AB and CD axes were 2.25:l and 2.21:1, respectively. As can be appreciated from Figure 11,the variation in the mean values for different ages was quite small. The maximum difference between the high and low values for MU11 was only 0.36 while for M2iI2 it was 0.86. DISCUSSION The middle-ear conductive apparatus is responsible for the transfer of sound vibrations from the tympanic W.M. ZIMMER ET AL. 0 0 0 Lever Ratio (A - B) A Lever Ratio ( C - D) Weight of Malleus Weight of Incus n = 5 n=5 I 0 20 40 60 a0 0 AGE IN DAYS 4 I 0 c Q 21 fx 1 li n = 5 17t 0 40 60 a0 AGE IN DAYS Fig. 9. The results in Figure 8 are replotted as a percent of the maximum value. In this way the relative rate of development for weight and arm length of the malleus and incus can be compared. As seen, the ossicular arms gain length before their mass increases up to about 20 days of age. ‘ 20 Pars Tensa/Oval Window Fig. 11. The changes in the lever ratio, for the levers in axis AB (MU111and axis CD (MWIZ), are compared over age. Little systematic change is seen over the 80-day interval. the rate of hearing development measured from more central locations in the auditory system (Saunders et al., 1993). While these studies provided intriguing evidence for the hypothesis that middle-ear maturation is a n important contributor to the rate of auditory functional development, contrary observations exist. Studies in the cat (Thomas and Walsh, 1990) demonstrated that a mature area ratio between the tympanic membrane and stapes foot plate was present at the end of the first week, a t a time when action potential thresholds recorded from central auditory nuclei were very insensitive. This observation suggested that the kitten cochlea was quite immature at a time of mature middle-ear structural development. In addition, the development of umbo velocity in the chick, a n important measure of middle-ear function, was found to be independent of the maturation of auditory thresholds measured from brainstem nuclei (Cohen et al., 1992a). 15 0 20 40 60 a0 AGE IN DAYS Fig. 10. The area ratio (pars tensaioval window area) is averaged over 5 specimens between 1 and 80 days of age. The solid line is the exponential fit to the data. The area ratio reaches asymptotic conditions within 10 days. Thereafter the expansion of both structures is proportionally the same. membrane to the fluid-filled spaces of the cochlea via the stapes foot plate. It stands to reason that maturation of the conductive system is intimately tied to measurement of hearing development recorded in the cochlea, the eighth nerve, or other higher auditory processing centers. In fact, studies on the mouse (Huangfu and Saunders, 1983; Doan et al., 1993), the gerbil (Cohen et al., 1992a,b), and hamster (Relkin and Saunders, 1980) suggested that both the structural and functional maturation of the middle ear sets a limit on Comparison With Other Data To our knowIedge this is the first report describing in detail the structural changes in the rat middle-ear during early postnatal life. A summary of developmental times for various parameters to reach 90% of adult size in the rat and two other mammals (mouse and gerbil) is given in Table 1. In general, the time required for various rat middle-ear structures to reach 90% of their adult size were significantly longer than those of the mouse (Huangfu and Saunders, 1983) and gerbil (Cohen e t al., 199213). Perhaps the most interesting difference between these species was the early maturation of the area ratio in the rat. This reflects the fact that pars tensa and oval window growth combined in such a way to achieve large area ratios very early. Structural Development of the Pressure Amplifier The long arms of the malleus and incus also expanded a t similar rates with both achieving adult 483 MIDDLE-EAR DEVELOPMENT TABLE 1. Summary of developmental time to achieve 90%of adult value' W Y Y LL I T r I Age in days Structure Pars tensaiOW ratio TM cone depth Pars tensa area Oval window area Stapes foot plate area Tympanic membrane area Incus length Malleus length Incus mass Malleus mass Pars flaccida area Bulla volume Rat Mouse2 8 15 17 13 17 23 26 34 34 36 49 59 17.5 10 18.5 6 19.5 15 16 Gerbil3 12 11.5 10 4 7 11.5 4 6 - 20.5 18 1 1 i ~ LL ~ 13 36 'The 90% values were determined from the corresponding exponential regression line for each of the middle ear structures measured. 'From Huangfu and Saunders (1983). 'From Cohen et al. (1992a). O w a I I I I f I /I 2ot 0 W 20 CAP at 8.0 kHz 40 60 1 80 AGE IN DAYS length prior to adult mass. This, in part, accounts for the minimal variation in lever ratios seen during the course of development (see Fig. 11).The amplification effect of the M2/I2 lever as defined here, was very close to that of the more traditional MU11 lever. However, we wonder whether this would actually be the case in the middle ear response to sound, since rotation on the CD axis requires both pivoting around the gonial and stretching at the posterior ligament of the incus. Stretching at the ligament seems likely since the ligament and incus were mobile at that location during the dissections. Pivoting may be less likely since the malleus was solidly attached to the tympanic ring a t the gonial. Indeed, separating the malleus from the tympanic ring during dissection required that the tympanic ring be either broken or cut at its fusion point. In light of these observations, i t is likely that considerable stiffness will be associated with pivotal motion a t the gonial. It seems unreasonable to expect that sound transmission occurred through the middle ear earlier than 6 days postpartum. The ossicles were soft and completely flaccid, and the middle-ear cavity was filled with gelatinous mesenchyme at that time. During the next 6 days the bulla gradually aerated, the ossicles became much more rigid (although not yet fully ossified), and the area ratio was near its adult value. On the basis of structural maturity alone i t is possible that sophisticated sound transmission through the middle ear was well established by 12 days of age. It has been shown (Uziel et al., 1981) that the cochlear microphonic and whole-nerve AP response, evoked by tone bursts or filtered clicks of various frequencies, were within the adult range when 15 days old, and this, of course, would indicate mature sound conduction by this time. Sound Conduction and Hearing Development The time course for the development of cochlear function has been well established in the rat (Uziel et al., 1981; Rybak e t al., 1992). The endocochlear potential (EP) was first reported a t 7 days, increased gradually until 11 days, rapidly escalated through 13 days, and then increased more slowly to achieve a n adult level 17 Fig. 12. The growth of the area ratio, the CAP threshold (at 8.0 kHz), and the endochochlear potential (EP) is plotted as a percentage of its maximum value. The relative rate of growth can be compared among the conditions. Rapid maturation of the areal ratio appears to occur prior to the onset of electrophysiologic events in the cochlea. The lines in this figure are smoothed between data points. days after birth (Rybak et al., 1992). Eighth nerve compound action potentials (CAP) elicited by click stimuli were first detected between 11 and 13 days after birth. However, very loud clicks, up to 100 dB SPL, were required to produce a response a t this time. Adult-like waveforms in the CAP response were identified between 16 and 17 days, while the CAP thresholds showed substantial improvement in sensitivity from 13 to 20 days of age. This was followed by continued gradual improvement until adult thresholds were reached between 25 and 28 days. The relationship between middle-ear structural maturation, given by the area ratio, and cochlear auditory development represented by the E P and CAP threshold (at 8.0 kHz) is summarized in Figure 12. Although the EP is not a sound driven potential, its presence is necessary for the normal transduction of sound in the cochlea (Rybak et al., 1992). Thus, its inclusion as another indicator of cochlear functional development is justified. We calculated the percent difference in the measured unit from the youngest age a t which data could be obtained to the adult level. While this distorts the results somewhat by forcing upper and lower age boundaries, it produces a plot in which the value a t the youngest age was 0% while the adult value was loo%, for each variable examined. By normalizing all curves against a common percent axis, a direct comparison of the relative rate of development across ages can be made. An examination of Figure 12 reveals that the area ratio matured earlier than the EP, and the CAP thresholds. Indeed the 90% of the maximum response occurred a t about 10 days for the area ratio, 16 days for the EP and after 28 days for the CAP. Electron micrographic studies of the developing rat cochlea (Burda, 1985; Roth and Bruns, 1992) also revealed striking 484 W.M. ZIMMER ET AL. changes between 4 and 12 days after birth. Thus, at approximately the tirne of hearing onsetat 12 days (as was found defined by the CAP responses), the to be structurally mature. The time course of area ratio and CAP maturation suggests that the middle ear and the cochlea develop independently of each other with sound conduction preceeding cochlear function. This study has provided information regarding the static development of the middle ear. The results presented here provide a strong foundation for further examination of functional middle-ear development in the rat (Igic et al., 1994). ACKNOWLEDGMENTS This research was supported by a n award from the NIDCD (ROl-DC00510) to J.c.s. The authors greatly appreciate the assistance of Daryl Doan, Petar kit, and Ms. Linda Markoff. Portions of this paper were presented at the 1994 Mid-Winter Meeting of the AsSt. peterssociation for ~~~~~~~hin Oto~aryngo~ogy, berg Beach, FL. LITERATURE CITED Burda, H. 1985 Quantitative assessment of postnatal maturation of the organ of Corti in two rat strains. Hear. Res., 17~201-208. Cocherell, T., I. Miller, and M. Printz 1914 The auditory ossicles of American rodents. Bull. Am. Mus. Natl. Hist., 33t347-380. Cohen, Y.E., H.N. Hernandez, and J.C. Saunders 1992a Middle ear development: 11. Morphometric changes in the conducting apparatus of the chick. J. Morphol., 212r257-267. Cohen, Y.E., C.K. Bacon, and J.C. Saunders 1992b Middle ear development 111. Morphometric changes in the conducting apparatus of the mongolian gerbil. Hear. Res., 62t187-193. DeMaio, F.H.P., and J . Tonndorf 1978 The terminal zone of the external auditory meatus in a variety of mammals. Arch. Otolaryngol., 104.5570-577. Doan, D.E., Y.E. Cohen, and J.C. Saunders 1994 Middle ear development: Iv. Umbo motion in neonatal mice. J. Comp. Physiol. A., 174t103-110. Fleischer, G. 1978 Evolutionary principles of the mammalian middle ear. Adv. Anat. Embryol. Cell Biol., 55t1-70. Huangfu, M., and J.C. Saunders 1983 Auditory development in the mouse: Structural maturation of the middle ear. J . Morphol., 176: 249-259. Igic, P.G., D.E. Doan, and J.C. Saunders 1994 The development of the tympanic membrane velocity in the rat. Abstr. Assoc. Res. Otolaryngol., 17t243. Relkin, E.M., and J.C. Saunders 1980 The development of tympanic membrane displacements in the hamster. Acta Otolaryngol., 90: 6-15. Roth, B., and V. Bruns 1992 Postnatal development of the rat organ of Corti: 11. Hair cell receptors and their supporting elements. Anat. Embryol., 185571-581. Rybak, L.P., C. Whitworth, and V. Scott 1992 Development of endocochlear potential and compound action potential in the rat. Hear. Res., 59t189-194. Saunders, J.C., and T.J. Garfinkle 1983 Peripheral physiology I. In: The Auditory Psychobiology of the Mouse. J.F. Willott, ed. Charles c. Thomas, Springfield, IL., PP. 131-168. Saunders, J.C., and R.M. Summers 1982 Auditory structure and function in the mouse middle ear: An evaluation by SEM and capacitive probe. J. Comp. Physiol. A., 46t517-525. Saunders, J.C., D.E. Doan, and Y.E. Cohen 1993 The contribution of middle ear sound conduction to auditory development. Comp. BioPhYSiol., 106*:7-13. Thomas, J.P., and E.J. Walsh 1990 Postnatal development of the middle ear: A real ratios in kittens. OtolaWWJl. Head Neck Surg., 103:427-435. Uziel, A., R. Roman4 and M. Marot 1981 Development of cochlear potentials in rats. Audiology, 2Ot89-100. van den Berge, H., A. van Geest, J.W. Rensema, and J. Drukker 1990 Three dimensional graphic reconstruction of the tympanic bulla of the rat with special reference to the middle ear muscles. Acta Otolaryngol., 1IOr253-261. Vrettakos, P.A., S.P. Dear, and J.C. Saunders 1988 Middle ear structure in the chinchilla: A quantitative study. Am. J. Otolaryngol., 9t58-67.