1. Introduction

Motile cilia are hair-like microscopic organelles that project from the surface of most eukaryotic cells, and they perform range of biological mechanosensory and chemosensory functions [1]. Synchronized beating of cilia in the respiratory tract expels the mucus out of lungs which traps foreign particles, pathogens, and bacteria inhaled during regular breathing [2]. The impairment of the mucuciliary clearance mechanism leads to chronic respiratory infections. Cilia also express sensory receptors [3], that play important roles in signal transduction pathways and in maintaining cellular homeostasis [1]. Ciliary dysfunction and its effects on developmental diseases is an active area of research, and some of the pathologies associated with ciliary dysfunction include respiratory infections, anosmia, and blindness among others [4]. Therefore, an imaging technique that can provide anatomical and functional assessment of cilia is of clinical significance, and would be helpful in better understanding of roles its dysfunction plays in various pathologies.

The mucuciliary clearance (MCC) mechanism has been investigated using a variety of imaging techniques. Gamma scintigraphy and saccharin clearance test are established procedures for diagnosis of impairment in MCC [5]. Recently, optical coherence tomography has been used for quantitative assessment of cilia driven fluid flow [6–9]. However, some of the proposed methods rely on particle tracking velocimetry which isn’t practical for in vivo imaging. Here we present a high speed spectral domain optical coherence phase microscopy (SD-OCPM) system that provides anatomical and functional assessment of tracheal epithelium with sub-cellular resolution. Micro-anatomical images of ex vivo mouse trachea and quantitative assessment of its ciliary dynamics are presented. Optical pathlength gradient maps computed from phase images are used to determine the ciliary beating frequency, and the magnitude and direction of the synchronized ciliary strokes.

2. Materials and methods

2.1. Preparation of mouse trachea

Preparation of the trachea was performed as previously described [10, 11]. Briefly, the mice were terminated via inhalation of isoflurane, the trachea was extracted and placed in an Environmental Culture Dish (model # 956, CROMAPHOR Analysen-Technik GmbH, Ascheberg), which was connected to a temperature controlled Culture Dish System (model # Delta T5, Bioptechs Inc., PA, USA). The culture dish was lined with Sylgard Polymer 184 Curing Agent Silicone Elastomer (Dow Corning GmbH, Wiesbaden) and filled with HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) buffered Ringer, whose contents are: 5.6 mM KCl; 136.4 mM NaCl; 1 mM MgCl2 × 6 H2O; 2.2 mM CaCl 2×2 H2O; 11 mM Glucose; 10 mM HEPES; pH 7.4. After washing with the HEPES Ringer, the trachea was cut longitudinally and flattened in the culture dish with insect pins (0.15 mm × 12 mm). The temperature of bathed trachea was maintained at 30–35°C during the experiment. The controlled environment of the culture dish and the medium− simulate normal physiological conditions to maintain the coordinated beatings of cilia, and preserve the morphology of mouse trachea for up to 5 hours.

Figure 1 shows H&E stained histological section of the trachea and scanning electron microscope image of ciliated cells in the epithelium, where hair-like microstructures known as cilium are clearly visualized.

 

Fig. 1 (a) H&E stained histology of trachea cross-section, (b) Scanning electron microscope image of ciliated cells.

Download Full Size | PDF

2.2. Optical coherence phase microscopy system and image acquisition

The spectral domain optical coherence phase microscopy system (SD-OCPM) used in this study combines sub-micrometer transverse spatial resolution with sub-nanometer optical pathlength sensitivity, as described previously [12]. The axial and transverse spatial resolution of the system were measured to be 0.6 μm and 2 μm respectively, with a maximum sampling speed of 217,000 A-Scans/sec. The phase sensitivity of the system was measured to be 0.0027 rad, which is equivalent to an optical pathlength sensitivity of 0.1 nanometres. Almost all SD-OCPM implementations reported in the literature so far, have used common-path interferometer for high phase sensitivity [13–20]. However, common-path interferometer topology has many drawbacks including suboptimal transverse resolution, limited depth scanning capability, and poor control on reference power which is important for high sensitivity measurements. The implemented SD-OCPM system overcomes these limitations by using a symmetric Lin-nik interferometer and by keeping all sources of noise such as galvanometer scanners, thermal deviation, etc., in common mode between sample and reference arms.

The objective of the study was to characterize both morphological structures and functional dynamics of the tracheal epithelium using SD-OCPM. While the acquisition speed is not critical for imaging morphological structures, accurate recording of the ciliary dynamics requires faster image acquisition speeds. Therefore volumes of two sizes 128 μm × 128 μm × 30 μm and 30 μm × 30 μm × 30 μm were acquired at the speed of 1 and 108 volumes/second, respectively.

3. Results and discussion

3.1. Micro-anatomical images of mouse trachea

The en face micro-anatomical images of mouse trachea at different depth positions are as shown in Fig. 2. The images cover an area of 128 μm × 128 μm with 0.25 μm/pixel display resolution, and were acquired at the acquisition rate of nearly 1 frame/sec. During the image acquisition, cilia of the tracheal epithelium were beating at the mean frequency of 16 Hz, and as such they appear slightly blurred and their movement trails can be visualized in Fig. 2(a). The secretory and basal cells can be seen in Figs. 2(b) and 2(c), respectively. Below the epithelium, a thin layer of loose connective tissue consisting of elastic fibres and fibroblast cells, also known as lamina propria mucosae, can be seen in Figs. 2(d) and 2(e). Below the lamina propria mucosae is the loose connective tissue of the tela submucosa, which contains blood capillaries. Longitudinal cross-section of a blood capillary along with single red blood cells (marked with red arrow) can be seen in Fig. 2(f). Further below, collagen fibres of the connective tissue that provide scaffolding support to the trachea, get relatively larger.

 

Fig. 2 En face images (128 μm × 128 μm) of ex-vivo mouse trachea at different depth positions, with 0.25 μm/pixel display resolution. (a) Ciliated epithelium with individual cilia slightly blurred because of their coordinated beating. (b) and (c) Different cell layers beneath the ciliated cells. (d) and (e) Connective tissue layer. (f) Longitudinal cross-section of blood capillary where single red blood cells can be visualized (marked with red arrow). (g) and (h) Collagen fibres of the connective tissue. (i) Reconstructed three dimensional structure of tracheal epithelium, also shown in Visualization 1. Scale bar represents 20 μm.

Download Full Size | PDF

The three dimensional structure of the trachea was reconstructed by stacking the en-face SD-OCM images acquired at 0.75 μm step along the depth, as shown in Fig. 2(i) and Visualization 1. The dimensions of the reconstructed volume are 128 μm × 128 μm × 40 μm. The Z dimension of the volume was enlarged by linear interpolation, for better visualization of morphological structures.

It can be noted, that the optical sectioning capability of SD-OCM rejects out of focus light much more effectively than the confocal microscopy alone [21], and therefore it provides sharp images of different tissue layers in depth.

3.2. High-speed volumetric imaging of ciliary beatings

Coordinated beatings of the cilia, imaged using the implemented SD-OCPM system are shown in Fig. 3 (a) and Visualization 2, which covers 128 μm × 128 μm × 30 μm volume and imaged at 1 volumes/sec. Although, image acquisition speed was insufficient to fully capture the complete cycle of motile cilia movement, nevertheless shown images demonstrate that single cilia and their movement can be clearly visualized with SD-OCM. In order to capture the complete beating cycle of the cilia, the imaging volume was reduced to 30 μm × 30 μm × 30 μm and volumes were acquired at 108 volumes/sec, as shown in Fig. 3(b) and Visualization 3.

 

Fig. 3 High-speed volumetric imaging of mouse tracheal epithelium with SD-OCM. (a) Still image from Visualization 2 which shows volume of size 128 μm × 128 μm × 30 μm acquired at 1 volumes/sec. (b) Still image from Visualization 3 which shows volume of size 30 μm × 30 μm × 30 μm acquired at 108 volumes/sec.

Download Full Size | PDF

3.3. Quantitative measurement of ciliary dynamics

The phase images obtained with high-speed volumetric recording are used to quantitatively characterize the ciliary dynamics such as strength, measured as induced change of the optical pathlength, and direction of cilia beating and its frequency. High-speed sampling of the cilia movement ensures that there are no temporal discontinuities in the recorded phase images, which were unwrapped and scaled into optical pathlength (OPL) values. The implemented SD-OCPM system has an OPL sensitivity of 0.9 nanometers for volumetric measurements. Therefore, spatio-temporal changes in OPL caused by the coordinated beating of cilia, were measured with very high phase sensitivity, and is utilized in quantitative assessment of ciliary dynamics.

3.3.1. Optical pathlength gradient mapping

Since the phase images quantitatively measure the changes in OPL over time with high sensitivity, calculating its spatial gradient yields a vector field whose amplitude and direction indicates the strength and spatial orientation of cilia beatings, respectively.

The OPL gradient of a phase image can be represented as:

Where OPL(x,y) is the two-dimensional OPL image, $\overrightarrow{u}$ and $\overrightarrow{v}$ are the gradient vectors along $\widehat{\mathbf{x}}$ and ŷ, which are unit vectors in X and Y spatial dimensions.

Optical pathlength images were obtained from the high-speed video shown in Visualization 3. Figure 4(a) shows OPL values over time, along a line in the spatial y-direction in the volume. Figure 4(b) shows part of the OPL values over time image, enlarged and overlaid with vector field that represents the strength and orientation of changes in OPL. It can be seen that, the direction of changes in OPL are coordinated and uniform across the lateral X-dimension, and reverse periodically over time. Temporal profile of changes in OPL, along the dotted line in (a), are shown in Fig. 4(c). The time-series was high pass filtered with a cut-off frequency of 4 Hz, to remove dc component and slow varying frequencies. The frequency spectrum of the time-series is shown in Fig. 4(d). Because of the complex beating cycle of motile cilia, The Fourier transform of changes in OPL over time yields a range of frequency peaks, and the one with highest amplitude is 16 Hz, which is expected at 32^◦C temperature [22].

 

Fig. 4 (a) OPL values over time along a line in the y-direction in Fig. 3(b). (b) Gradient vectors representing the directional changes in OPL values. (c) OPL values over time at position marked by the dashed lines in (a). (d) Frequency spectrum of the OPL values over time.

Download Full Size | PDF

4. Conclusion

Micro-anatomical images of ex vivo mouse trachea obtained with SD-OCM resolve different cell layers of epithelium with sub-cellular resolution. Fine morphological structures of the trachea such as cilia, collagen fibres and individual red blood cells in the capillaries are clearly visualized. Generally, these kind of fine morphological structures can only be visualized with nonlinear optical microscopy or by performing a biopsy on the trachea and then processing it into histological sections. SD-OCPM is particularly well suited for imaging of tracheal epithelium as it combines morphology with a functional contrast that quantitatively measures the strength and direction of cilia strokes, and its frequency.

Ciliary beating is clearly visualized with high-speed recording, and phase images were used for quantitative assessment of ciliary dynamics such as beating frequency, magnitude and direction of the cilia strokes. Coordinated beating of cilia and reversal of cilia strokes over time are visualized through optical pathlength gradient maps. Currently, there are no diagnostic tools or imaging technique that can provide in vivo functional assessment of tracheal epithelium without using extraneous agents. SD-OCPM can be miniaturized using a fibre-optic probe, and could potentially be used for in vivo diagnosis of primary ciliary dyskinesia, where cilia beating are not coordinated.

In conclusion, SD-OCPM has been used for morphological and functional imaging of mouse tracheal epithelium with sub-cellular resolution. Quantitative phase contrast has been utilized to determine the strength and direction of cilia strokes. An endoscopic implementation of SD-OCPM could potentially be used for in vivo diagnosis and monitoring of various pathologies associated with mucuciliary clearance.