Collagen fiber is an important component of the extracellular matrices in the human body (Matsuda et al., 1987; Malkusch et al., 1995; Mercer and Crapo, 1990). The collagen fiber bundles in the matrix play an important role for the mechanical functions of connective tissues such as skin (Osaki, 2001) and bones (Osaki et al., 2002). The lung is also considered to be a type of connective tissue with mechanical functions because the periodic mechanical stress ascribed to respiration is applied to the lung consisting of collagen fibers, elastin, and proteoglycan. Toshima et al. (2005) reported that collagen fibers in aggregated state were observed at the alveolar orifices of the human lung using scanning electron microscope, while sac like collagen networks was observed at the alveolar septa, forming basket like networks. Collagen fibers in the collapsed lung of rat showed a wavelike configuration at the alveolar orifices and septa, while they became straight in the inflated lung (Toshima et al., 2004). Kononov et al. (2001) observed changes in morphology related to collagen fibers in the alveolar walls of rat by mechanical stretching using optical microscope and evaluated the role of collagen fiber on the mechanical property of alveolar walls. The two reports were restricted on qualitative observation in the small regions of the lung. Thus there have been few reports evaluating the collagen fiber orientation in the lung quantitatively.
It is important to evaluate the orientation of collagen fiber as related to the mechanical properties of the lung because the lung needs to maintain its structure for mechanical force accompanied with respiratory movement. However there are few reports on evaluating the relationship between the mechanical properties of the whole lung and the orientation of collagen fibers because staining method using specific antibody for some types of collagen is generally used not for determining the orientation of collagen fibers, but for detecting the existence of a part of collagen fibers. The numerical analysis for the orientation of collagen fibers will be required for large sections of tissues for studying the role of collagen fibers regarding the mechanical properties. Previously, one of authors (S.O.) established microwave method which is able to determine the orientation of collagen fibers in the sheet from the angular dependence of the transmitted microwave intensity (Osaki, 1987a, 1987b, 1989, 1990a, 1997). The method was applied to human tissues such as skin (Osaki, 1990b, 1999; Osaki and Ohashi, 2004) and bones (Osaki et al., 2002; Ohuchi et al., 2003). However, this method has not been applied to the human lung, because it has been very difficult to cut human lung into several plates, which is a spheroid organ containing inspired air. And it has been also very difficult to prepare the sliced lung plate into sheet samples of about 1 mm thickness without curling ascribed to drying. Nevertheless, we finally succeeded in preparing lung sheet samples appropriate for microwave measurements.
The present study describes the preparation of sheet samples from a lower lobe of the human lung, the determination of orientation of collagen fibers, and provides the distribution of collagen fiber orientation, suggesting that the fiber orientation is closely related to the respiratory movement.
MATERIALS AND METHODS
Human lung removed on autopsy of a 60-year-old man who had died of hepatocellular carcinoma was used in the present study after his family provided written informed consent. He had no histological abnormality in the lung. Several lungs were also used in preliminary experiments. The Institutional Review Board at Nara Medical University approved our study. Samples for measurements were prepared from the lung tissue as follows: After upper and lower lobes of the left lung have been fixed in 10% formalin solution at a room temperature for a week, it was sliced in the direction of coronal plane into several lobe plates 10-mm thick by a trimming knife (FEATHER, Tokyo, Japan) used for autopsy. Middle plate from the lower lobe, which was the largest in size among the plates, was used. The plate was cut into two parts, upper (U) and lower (L) part (Fig. 1A) because it was too large to cut the whole plate into thin slices with a trimming knife (MicroGlass, WA). The lower part was used for determining collagen fiber orientation because it contained no large artery and few bronchi compared with the upper part.
After being embedded in paraffin at 40°C for fixation, the lower part was further cut into 2-mm thin slices with a trimming knife. The slices were then dipped into a 70% ethanol solution to remove the paraffin. To use sheets without curling is inevitable for the microwave measurements. Each slice was sandwiched between two pieces of wood cut such as to avoid curling. After drying at a room temperature for 24 h the resulting sheet lung about 1 mm thickness was used as a sample for determining the fiber orientation using the microwave method.
Measurements of Collagen Fiber Orientation
Collagen fiber orientation was determined using the Osaki microwave method (Osaki, 1990b). The lung sheet sample was inserted into the narrow gap between a pair of waveguides constituting the cavity resonator system and was rotated around the central axis normal to the sample plane. Microwaves were irradiated to the samples and the transmitted microwave intensity of the sheet samples was measured at different rotation angles (Osaki, 1987a, 1987b, 1989). Effective sample size to which microwaves were irradiated was 25 mm × 25 mm. The angular dependence of transmitted microwave intensity, called the orientation pattern, was measured at 12 GHz. The orientation pattern gives the orientation angle (β) and the degree of fiber orientation (DFO) (Osaki, 1987a, 1987b, 1989). The direction at which the transmitted microwave intensity is minimal is designated as the orientation angle corresponding to the angle between the main axis of the collagen-fiber chains and the standard direction (SD). The SD corresponds to the direction of the longitudinal axis of the spine. The maximal-to- minimal ratio of transmitted microwave intensity is defined as the DFO reflecting mechanical anisotropy. The collagen fiber orientation can be explained by two factors of DFO and β.
Observation of Morphology of the Human Lung
The lung contains air spaces for gas exchange. The lung which receives the inspired air changes the void spaces during respiration, expanding on inspiration and shrinking on expiration. The morphology of the lung sample was observed by optical microscope. The direction of long axis of air spaces was determined. From the shapes of air spaces in the sample we then compared relationship between the anisotropic shapes of air spaces and the collagen fiber orientation.
Collagen Fiber Orientation in Human Lung
Figure 2 shows the angular dependence of transmitted microwave intensity at four different positions (sample number 4–8, sample number 4–6, sample number 4–4, sample number 4–2) for the human lung sheet sample prepared by slicing the lower part of left lower lobe (see Fig. 1B). The angular dependences at about 12 GHz were ellipsoidal. The degree of fiber orientation (DFO) was determined to be 1.665 for sample number 4–8, 1.618 for sample number 4–6, 1.575 for sample number 4–4 and 1.813 for sample number 4–2. Here, the large value of DFO showed marked anisotropy. The orientation angle β was determined to be 25°for sample number 4–8, 20° for sample number 4–6, 2° for sample number 4–4, and −4° for sample number 4–2. The value of DFO changed slightly with changing position while the value of β changed markedly with changing position. The results show that collagen fibers are oriented anisotropically, depending on the position of lung.
Distribution of Collagen Fiber Orientation
Figure 3 shows the distribution of collagen fiber orientation at 61 different positions of the human lung sample prepared from the lower part of the coronal plates of the left lower lobe. Collagen fiber orientation is represented as a bar. Inclination of a bar gives β the deviation of fibers from SD, while the length of a bar gives DFO the degree of fiber orientation.
DFO changed from 1.120 to 2.657 while β changed from −18° to 73°. Both DFO and β change with changing position. Here the lung sample is divided into three different lesions: inner, middle, and outer lesions. The DFO is relatively high in the outer part, while it is low in the middle and inner parts. β for the inner part was about −30° to SD while β for the middle and outer parts was small. That is, the collagen fibers in the middle and outer parts were almost aligned in SD. The collagen fiber orientation in the lung sheet sample changed with changing position, as shown at Fig. 3.
Fine Structure of Air Spaces in Human Lung
Fine structures at two different positions (sample numbers 5–3 and 5–9) of the lung sample were observed using an optical microscope. Shapes of airspaces were on average ellipsoidal for the sample (sample number 5–3) with large DFO (see Fig. 4A). The direction of the longitudinal axis of the ellipsoidal air spaces was parallel to that of the spine. Similarly, the bundles in the circumference of air spaces were also parallel to that of the spine (see Fig. 4A). On the other hand, air spaces were almost circular for the sample (sample number 5–9) with low DFO, while the bundles were not oriented in a fixed direction (see Fig. 4B).
The results suggest that alignment of bundles is related to the collagen fiber orientation in the human lung.
Collagen fiber, a major component of the extracellular matrix of the human lung, is crucial in keeping the mechanical function of the lung accompanying with inspiration. However, it has not been elucidated how collagen fiber should contribute to maintain the lung structure against mechanical load based on the respiratory movements. It is very important to determine the orientation of collagen fibers quantitatively for studying the mechanical function of the lung. In the present study, we succeeded in preparing sheet samples for measurements of human lung and in determining the collagen fiber orientation using Osaki's microwave method (Osaki, 1987a). It was very difficult to prepare samples as described in the Introduction because of factors such as cutting, drying, and so forth. However we measured the orientation of collagen fibers for several sheet samples from other subjects. We showed the orientational distribution for one sample since the orientational distribution of collagen fiber was roughly similar.
The results indicate that collagen fibers in the human lung are mainly orientated in the direction parallel to the spine. The angle of collagen fiber orientation (β) changed between −18° and 73°. The variation in β was very large, while the degree of orientation was relatively small. It is important to determine the distribution of collagen fiber orientation at about 60 points to make clear the orientational distribution of collagen fiber in the lung. The result demonstrated that collagen fibers were, on average, aligned in parallel to the longitudinal axis of the spine and that the degree of orientation is relatively high in the outer part of lung.
Previously one of the authors (S.O.) applied the microwave method to human bone, calf skin and cobra skin (Osaki, 1999; Osaki et al., 2002, Niitsuma et al., 2005). He demonstrated that collagen fibers were anisotropically orientated and that the anisotropy was closely related to the mechanical anisotropy. In a similar way, the present result indicates that the anisotropic orientation of collagen fiber in the human lung may be related to the mechanical anisotropy. Human lung changes its shape during respiration. On inspiration the lung dose not expand uniformly. As is well known, the shapes change remarkably in the direction longitudinal to the spine compared with the other directions while those change remarkably in the outer area compared with inner area. The anisotropy in respiratory movement may be related with those of collagen fiber orientation.
In the present study, we also observed the void structure containing air spaces in the lung sample and investigated the relationship between the void structure and the collagen fiber orientation. The void structure contains inspired air and changes during respiration. The inspiration makes the structure expand while the expiration makes it contract. Because the mechanical stress based on respiration contributes to change in the void structure in the lung, the structures should reflect the mechanical properties of the lung. Many airspace shapes were not circular but ellipsoidal in highly oriented regions. Moreover, the long axes of the air spaces were proved to be mainly parallel to that of the spine. One of the authors (S.O.) has reported that hair pores in the skin were ellipsoidal and that the longitudinal axes of the hair pores postulated the direction of collagen fibers orientation of the skin (Osaki, 2001). The direction of the longitudinal axes of the air spaces agreed with that of collagen fiber orientation of the lung sample determined by the microwave method. This result suggests that collagen fiber orientation in the human lung is markedly related to the mechanical properties.
In conclusion, we succeeded in preparing the human lung sample for Osaki's microwave method and in determining the collagen fibers orientation in human lung tissue by Osaki's microwave method, and found that the collagen fibers in the sample from the coronal plate are generally oriented parallel to the spine. In the near future, we will present the mechanical properties of the human lung for asserting the present using microwave method.
The authors are grateful to Professor and Chairman J.Patrick Barron of Department of the International Medical Communication Center of Tokyo Medical University for his review of the manuscript.