We present a multimodal nonlinear imaging approach to elucidate microstructures and spectroscopic features of oral mucosa and submucosa in vivo. The hamster buccal pouch was imaged using 3-D high resolution multiphoton and second harmonic generation microscopy. The multimodal imaging approach enables colocalization and differentiation of prominent known spectroscopic and structural features such as keratin, epithelial cells, and submucosal collagen at various depths in tissue. Visualization of cellular morphology and epithelial thickness are in excellent agreement with histological observations. These results suggest that multimodal nonlinear optical microscopy can be an effective tool for studying the physiology and pathology of mucosal tissue.
©2004 Optical Society of America
Successful development of sensitive diagnostic tools and effective therapeutic intervention for treatment of mucosal disease is highly dependent on our understanding of the molecular, morphological, and functional changes that occur during disease progression. Pathological transformation of normal mucosal tissue is accompanied by specific alterations in tissue microstructure and function that in the past have been studied mainly by in vitro methods but need to be better characterized under in vivo conditions. For instance, in vivo evaluation of drug-tissue interactions may aid in the development of new therapeutic approaches by revealing processes and microscopic alterations that are not observable under in vitro conditions. A number of morphological and function changes have been observed in the mucosa during disease progression such as an increase in nuclear size and nuclear/cytoplasmic ratio, an increase in epithelial thickness, changes in autofluorescence of biochemical markers, or alterations in extracellular matrix architecture [1–7]. A number of optical techniques have been proposed for the noninvasive assessment of morphological and functional state of epithelial and mucosal tissue, particularly related to epithelial neoplasms [8–14]. Fluorescence spectroscopy (FS) reveals intrinsic signatures that can be correlated to microstructures and biochemical composition of epithelial tissues. In the case of epithelial neoplasms, an increase in NADH and porphyrin, and a decrease in FAD have been observed in transformed tissues [7,15–17]. In addition, a decrease in collagen autofluorescence in neoplastic tissue has been attributed to architectural changes in the submucosa [18,19]. In submucous fibrosis, a disease of the oral mucosa, alterations in autofluorescence from the normal condition have also been attributed to changes in collagen content . The lack of depth-resolved imaging of these spectroscopic methods prevents the localization of autofluorescence signatures to a region of the microstructure, so direct in vivo assessment of architectural changes in submucosal collagen is not possible using fluorescence spectroscopy. Although spatial localization is possible using fluorescence microscopy, spatial overlap in fluorescence emission from collagen and other microstructures such as epithelial cells makes direct differentiation of collagen fluorescence from cellular fluorescence a difficult problem. However, the development of second harmonic generation microscopy, with inherent contrast for collagen and other structural proteins, has enabled one to specifically differentiate collagen from other microstructures in normal mucosa [20–22].
The goal of this study is to investigate the application of a multimodal nonlinear imaging approach based on the integration of multiphoton microscopy (MPM) with second harmonic generation microscopy (SHGM) for in vivo evaluation of epithelial tissue microstructure. The combination of these imaging modalities allows deep visualization of tissue with high resolution capable of resolving cellular and subcellular microstructures, which has been demonstrated in several tissue types [21–23]. MPM is ideal for functional imaging of microscopic structures by either autofluorescence (e.g., NAD(P)H etc.) or use of functional fluorescent probes [24–26]. SHGM has been used to visualize collagen in nonmucosal tumors, but was not combined with MPM . However, SHGM complements MPM by providing direct contrast for specific structural proteins such as collagen and can be coregistered with MPM. In this study, we used the combined nonlinear 3D optical imaging technique to visualize the morphological, spectroscopic, and microstructural features of mucosal tissue under in vivo conditions in hamster cheek pouch. This animal model is commonly used to study mucosal disorders and provides easy access for imaging and manipulation .
2. Materials and methods
The combined multiphoton and SHG microscope setup is shown in Fig. 1. Illumination of the sample for both multiphoton excitation and SHG is accomplished using a Ti:sapphire femtosecond laser (100 fs, 82 MHz) that is tunable from 720 to 950 nm (Tsunami, Spectra-Physic, CA). Both multiphoton autofluorescence and SHG were collected in a backscattering (epi-illumination) geometry. Excitation wavelengths centered within 730–840 nm were used, while power incident on the sample was kept less than 20 mW in all experiments. A 690 nm short pass dichroic mirror in the detection path is used to separate fluorescence emission or SHG from reflected illumination. For SHG imaging, the laser is tuned to 800 nm and a bandpass filter centered at 400 nm with a 14 nm bandwidth was inserted to reject the fluorescence emission.
Approximately 10 minutes prior to imaging, hamsters were anesthetized with ketamine (150 mg/kg) and Xylazine (2.5 mg/kg) by intraperitoneal injection. The cheek pouch was then manually everted, rinsed with physiological saline solution, and pinned on a sample holder that was then fixed onto the microscope stage, allowing the cheek pouch to be imaged. Two objectives were used in this study: 40×, 1.2 NA, water immersion objective (C-Apochromat) and 10×, 0.3 NA, dry objective (Plan-Neofluar). The target area on the buccal pouch was centered with the aid of an aiming beam from the excitation source and then precisely located by scanning (1 s per frame with 512×512 pixels) using the low magnification objective (10×) in reflectance mode. Once the surface of the target area was located, the microscope setup was switched to multiphoton fluorescence mode with the 40× objective. Sequential depth-scans were taken using a typical z increment of 3 µm. SHGM was then performed on the same image volume using the same z increment. Thus, for each z-plane, corresponding MPM and SHGM images were acquired. To reduce noise and improve imaging quality, an averaging scanning scheme was used, resulting in a dwell time of 30–60 µs per pixel. A biopsy punch was taken at the center of the imaging area. The tissue sample was fixed in 10% neutral-buffered formalin, embedded in paraffin, and sectioned perpendicular to the sample surface with a sectioning thickness of 5 µm. Several sections near the starting and end point were omitted from further processing. Alternate sections were stained in groups of four with hematoxylin-eosin (H&E) and Massons’ trichrome; another four sections were made following a 50 µm gap in sectioning. This method allowed for sections to be sampled across a 300–400 µm distance approximately centered in the imaging area (imaging field of view was 320×320 µm).
3. Results and discussion
Four buccal pouches were imaged in this study. Typical images of tissue sections visualized at various depths with MPM and SHGM are depicted in Fig. 2. The first column consists of multiphoton fluorescence images whereas the second column shows the corresponding SHG images. All of the MPM images shown in Fig. 2 were collected using the same detector gain. For SHG images, the detector gain was increased by 50%. MPM allows imaging of cellular structures within the epithelium up to approximately 40 µm and of collagen beyond this depth. There is no detected SHG signal in the superficial depths as expected, since the source of SHG in the cheek pouch is most likely due to collagen found in the submucosa . The first sign of collagen as revealed by SHG occurs at approximately 40 µm, confirming that the fibrillar structure in the corresponding autofluorescence image is mainly collagen. Collagen is the dominant component of the connective tissue below the epithelium in oral mucosa. These images demonstrate how dual imaging with MPM and SHGM helps differentiate signals from cellular components and collagen. Although relatively small amounts of elastic fibers are also found in the submucosa , several studies have reported the absence of SHG from elastin using 800 nm excitation [21,27].
Cells identified by the first three MPM images in Fig. 2 are believed to be keratinocytes in the stratum corneum, prickle cells in the stratum spinosum, and basal cells in the stratum basale, respectively. To confirm this, we reconstructed the 3-D data of the buccal pouch shown in Fig. 2 and displayed the results as y-z cross-sections for direct comparison with histology in Fig. 3. A MPM cross-section is shown in Fig. 3(a) and the corresponding SHG cross-section is shown in Fig. 3(b). A two-color image of the MPM and SHG signals in Fig. 3(c) demonstrates strong correlation between the two types of signals at depths >45 µm. This interface is believed to be the basement membrane since 1) collagen appears to be the major signal contributor beyond this depth, and 2) comparison with histology in Fig. 3(d) indicates the distance from the surface is as expected. In the MPM image of Fig. 3(a), we see the epithelium consists of a thin bright fluorescing layer followed by layer exhibiting less fluorescence. The bright layer is the stratum corneum consisting of the bright keratinocytes observed in the most superficial MPM region of Fig. 2. The layer below the stratum corneum consists of several cell layers that make up the prickle cell region and basal cell region. The MPM x-y images at 30 µm and 42 µm revealed the presence of a higher density of cells within deeper epithelium (basal cell region), which is in agreement with previous histological and confocal microscopy studies [2,11].
Quantitative analysis of MPM images enabled measurement of epithelial thickness and cellular nucleus diameter (Table 1). As an example, the y-z cross-section image in Fig. 3(a) has been used to determine the epithelial thickness by measuring the thickness of the darker layer located between the two brighter layers in the sandwich structure. The measurement was repeated at 20 positions evenly apart along the y direction, resulting in an average thickness of 30.0±4.8 µm. The epithelial thickness was also measured with the histology image shown in Fig. 3(d), resulting in an average value of 40.3±7.2 µm. For measuring nuclear diameter, we took into account that the nuclei in histological sections are in a plane (y-z) perpendicular to the tissue surface rather than in a plane parallel to the tissue surface as in individual x-y scan taken with multiphoton fluorescence. A fair comparison between histology and multiphoton imaging involved measuring cells along the same y direction (parallel to tissue surface in y direction). The superficial epithelium was defined as the first cell layer below the keratin layer and the deep epithelium as the cell layer above the basement membrane for both histology and MPM. In MPM measurements, images were first thresholded following histogram analysis to reveal nuclei of individual cells. The results of the above quantitative analysis on MPM images are listed in Table 1, showing good agreement with histology. Some discrepancy in epithelial thickness was noted between the two methods. This error is attributed to artifacts such as delamination of the keratinized layer during histological processing.
Figure 4 demonstrates the emission properties of normal buccal pouch upon excitation of three distinct wavelengths (730, 780, and 800 nm). In the case shown, the sample was imaged at a single depth of 35 µm where both epithelial cells and collagen are present near the basement membrane. Figure 4(a)–(d) are taken upon illumination with 730 nm radiation, whereas Fig. 4(e)–(h) were acquired using 780 nm, and Fig. 4(i)–(m) using 800 nm. Emission band pass ranges are listed in the left column in Fig. 4. There are a number of observations that can be made based on assessment of these images. First, cellular structures are most
evident using 730 nm excitation (Fig. 4(a)–(c)). Although individual nuclei can be identified using 800 nm excitation (Fig. 4(i)), the cell boundaries are not as distinct as in the case of 730 nm using the same incident power. Autofluorescence originating from epithelial cells due to 730 nm excitation is primarily observed in the 400–500 nm region (Fig. 4(b) and (c)). Recent studies on mucosal tissues using fluorescence spectroscopy have identified the major endogenous fluorophores in epithelial tissues [7,10,15–19]. These include reduced nicotinamide adenine dinucleotide (NAD(P)H), flavin adenine dinucleotide (FAD), and tryptophan in the epithelial cells, and collagen in the underlying submucosa. In these studies, NAD(P)H fluorescence was found to be strongest in the 400–500 nm region, corresponding to the two emission ranges shown in Fig. 4(b) and (c), while FAD autofluorescence peak falls within 500–550 nm corresponding to Fig. 4(d) .
In the image series using 780 nm illumination, fluorescence from cells is lower (Fig. 4(e)–(h)). However one observes collagen autofluorescence in the 450–550 nm range (collagen appears on right side of the field), which was not observed using 730 nm excitation. This collagenous region appears bright and diffuse (Fig. 4(g) and (h)). At 800 nm excitation, however, the same region appears highly structured in the 400–450 nm range (Fig. 4(j)), but diffuse at emission wavelengths >450 nm. The corresponding SHGM image is shown in Fig. 4(m), demonstrating a highly structured fibrillar pattern for collagen similar to the pattern of Fig. 4(j). These results indicate that at 800 nm, the main sources of signal are autofluorescence and SHG from collagen with a weaker optical signal originating from cellular structures. The cellular microstructure and the extracellular matrix can be differentiated by multimodal MPM and SHGM imaging to discriminate SHG signals of collagen from the background autofluorescence signal as shown in Fig. 4(m). At excitation wavelengths greater than 800 nm we observed a strong signal from collagen (both autofluorescence and SHG), but virtually no signal from cellular structures. The above results suggest that 800 nm excitation is an optimal wavelength for simultaneous visualization of cellular microstructures and extracellular matrix in the oral mucosa and submucosa. Shorter excitation wavelengths are optimal for imaging of epithelial cells alone.
Our above preliminary results on in vivo imaging of hamster cheek pouch demonstrate that a high-resolution multimodal imaging modality combining multiphoton fluorescence with second harmonic generation microscopy can clearly elucidate the microstructure of both the epithelium and deeper submucosa in mucosal tissues. Cellular and epithelial morphology measurements were in excellent agreement with histology, demonstrating the potential of this technique for the study of disease of epithelial tissues. In particular, we demonstrated the ability to visualize microstructure and measure morphological parameters that are known to change with disease progression. For example, during neoplastic transformation cellular morphology is altered, epithelial thickness increases, and collagen found in the submucosa undergoes architectural changes [1–2]. Submucous fibrosis is marked by the accumulation of subepithelial (submucosal) collagen  and could in particular benefit from SHGM for direct study of collagen. Diabetes mellitus is also associated with cellular morphological changes in the epithelium and alterations in epithelial thickness . Finally, it has been suggested the susceptibility of vaginal mucosal tissue to HIV transmission may be correlated to epithelial thickness and may be reduced by estrogen treatment . The method presented in this paper will not only allow for these pathological changes to be studied in vivo, but also will allow for the assessment of drug interactions with mucosal tissue. Although in this study MPM was used to measure microstructure autofluorescence, there is the possibility to further study tissue function by MPM using functional biomarkers or fluorescent probes. Finally, this technique shows potential to be developed into a diagnostic tool. While the expense of ultrafast lasers could be a hindrance to translation into the clinic, such sources will likely become more user-friendly, less expensive, and more available in the future.
In this study, we have demonstrated the ability of MPM/SHGM for high resolution imaging of the oral mucosa and submucosa in vivo. The technique allows the identification of specific microstructures that may alter in morphology and function with disease progression. These include cellular/nuclear morphology, thickness of the keratinized stratum corneum, thickness of the epithelium, and direct identification of collagen in the submucosa. Imaging results of normal oral epithelial tissues are in excellent agreement with histology. An illumination wavelength of 800 nm was found to be optimal for imaging of both cellular morphology and submucosal collagen using the combined MPM/SHGM approach, however 730 nm is preferred for imaging only cellular structures.
Financial support from the John Sealy Memorial Endowment Fund for Faculty Recruitment (#6074-03), National Institutes of Health (# R21-CA89266), and NASA (# NAS2-0205) are gratefully acknowledged.
References and links
1. William O. Dobbins, “Diagnostic Pathology of the Intestinal Mucosa: an Atlas and Review of Biopsy Interpretation,” (Springer-Verlag, New York, NY, 1990).
2. “Histological Typing of Cancer and Precancer of the Oral Mucosa,”J.J. Pindborg, ed. (Springer-Verlag, New York, NY, 1997).
5. S. Alberti, C.T. Spadella, T.R.C.G. Francischone, G.F. Assis, T.M. Cestari, and L.A.A. Taveira, “Exfoliative cytology of the oral mucosa in type II diabetic patients: morphology and cytomorphometry,” J. Oral Pathol. Med. 32, 538–543 (2003). [CrossRef] [PubMed]
6. H.M. Chen, C.Y. Wang, C.T. Chen, H. Yang, Y.S. Kuo, W.H. Lan, M.Y.P. Kuo, and C.P. Chiang, “Auto-fluorescence spectra of oral submucous fibrosis,” J. Oral. Pathol. Med. 32: 337–343 (2003). [CrossRef] [PubMed]
7. M.G. Muller, T.A. Valdez, I. Georgakoudi, V. Backman, C. Fuentes, S. Kabani, N. Laver, Z.M. Wang, C.W. Boone, R.R. Dasari, S.M. Shapshay, and M.S. Feld, “Spectroscopic detection and evaluation of morphologic and biochemical changes in early human oral carcinoma,” Cancer 97, 1681–1692 (2003). [CrossRef] [PubMed]
8. J.A. Izatt, M.D. Kulkarni, H.W. Wang, K. Kobayashi, and M.V. Sivak, “Optical coherence tomography and microscopy in gastrointestinal tissues,” IEEE J. Sel. Top. Quant. Electron. 2, 1017–1028 (1996). [CrossRef]
9. C. Pitris, C. Jesser, S.A. Boppart, D. Stamper, M.E. Brezinski, and J.G. Fujimoto, “Feasibility of optical coherence tomography for high-resolution imaging of human gastrointestinal tract malignancies,” J. Gastroenterol. 35, 87–92 (2000). [CrossRef] [PubMed]
10. A. Gillenwater, R. Jacob, R. Ganeshappa, B. Kemp, A.K. El-Naggar, J.L. Palmer, G. Clayman, M.M. Follen, and R.R. Richards-Kortum, “Noninvasive diagnosis of oral neoplasia based on fluorescence spectroscopy and native tissue autofluorescence,” Arch. Otolaryngol. Head Neck Surg. 124, 1251–1258 (1998). [PubMed]
11. W.M. White, M. Rajadhyaksha, S. Gonzalez, R.L. Fabian, and R.R. Anderson, “Noninvasive imaging of human oral mucosa in vivo by confocal reflectance microscopy,” Laryngoscope 109, 1709–1717 (1999). [CrossRef] [PubMed]
12. A.L. Clark, A.M. Gillenwater, T.G. Collier, R. Alizadeh-Naderi, A.K. El-Naggar, and R.R. Richards- Kortum, “Confocal microscopy for real-time detection of oral cavity neoplasia,” Clin. Cancer Res. 9, 4714–4721 (2003). [PubMed]
13. A.A. Elfert, P. J. Pasricha, B. Bell, R. Johnigan, S.Y. Xiao, K.H. Calhoun, and M. Motamedi, “High resolution optical coherence tomography for early detection of epithelial neoplastic transformation,” Gastrointest. Endosc. 53, AB117 (2001). [CrossRef]
14. K. Sokolov, R. Drezek, K. Gossage, and R.R. Richards-Kortum, “Reflectance spectroscopy with polarized light: is it sensitive to cellular and nuclear morphology,” Opt. Express 5, 302–317 (1999), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-5-13-302. [CrossRef] [PubMed]
18. R. Drezek, C. Brookner, I. Pavlova, I. Boiko, A. Malpica, R. Lotan, M. Follen, and R. Richards-Kortum, “Autofluorescence microscopy of fresh cervical-tissue sections reveals alterations in tissue biochemistry with dysplasia,” Photochem. Photobiol. 73, 636–641 (2001). [CrossRef] [PubMed]
19. I. Georgakoudi, B.C. Jacobson, M.G. Muller, E.E. Sheets, K. Badizadegan, D.L. Carr-Locke, C.P. Crum, C.W. Boone, R.R. Dasari, J. Van Dam, and M.S. Feld, “NADH and collagen as in vivo quantitatitive fluorescent biomarkers of epithelial precancerous changes,” Cancer Res. 62, 682–687 (2002). [PubMed]
21. W.R. Zipfel, R.M. Williams, R. Christie, A.Y. Nikitin, B.T. Hyman, and W.W. Webb, “Live tissue intrinsic emission microscopy using multiphoton-excited native fluorescence and second harmonic generation,” P. Natl. Acad. Sci. USA 100, 7075–7080 (2003). [CrossRef]
22. A. Zoumi, A. Yeh, and B.J. Tromberg, “Imaging cells and extracellular matrix in vivo by using second-harmonic generation and two-photon excited fluorescence,” P. Natl. Acad. Sci. USA 99, 11014–11019 (2002). [CrossRef]
23. A.T. Yeh, N. Nassif, A. Zoumi, and B.J. Tromberg, “Selective corneal imaging using combined second-harmonic generation and two-photon excited fluorescence,” Opt. Lett. 27, 2082–2084 (2002). [CrossRef]
24. E.B. Brown, R.B. Campbell, Y. Tsuzuki, L. Xu, P. Carmeliet, D. Fukumura, and R.K. Jain, “In vivo measurement of gene expression, angiogenesis and physiological function in tumors using multiphoton laser scanning microscopy,” Nat. Med. 7, 864–868 (2001). [CrossRef] [PubMed]
25. S. Huang, A.A. Heikal, and W.W. Webb, “Two-photon fluorescence spectroscopy and microscopy of NAD(P)H and flavoprotein,” Biophy. J. 82, 2811–2825 (2002). [CrossRef]
26. W.R. Zipfel, R.M. Williams, and W.W. Webb, “Nonlinear magic: multiphoton microscopy in the biosciences,” Nat. Biotechnol. 21, 1368–1376 (2003). [CrossRef]
27. E. Brown, T. Mckee, E. Ditomaso, A. Pluen, B. Seed, Y. Boucher, and R. K. Jain, “Dynamic imaging of collagen and its modulation in tumors in vivo using second-harmonic generation,” Nat. Med. 9, 796–800 (2003). [CrossRef] [PubMed]
28. L. Coghlan, U. Utzinger, R.A. Drezek, D. Heintzelmann, A.F. Zuluaga, C. Brookner, R.R. Richards-Kortum, I. Gimenez-Conti, and M. Follen, “Optimal fluorescence excitation wavelengths for detection of squamous intra-epithelial neoplasia: results from an animal model,” Opt. Express 7, 436–446 (2000), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-7-12-436. [CrossRef] [PubMed]
29. L. Moss-Salentijn and M. Hendricks-Klyvert. Dental and Oral Tissues (Lea & Febiger, Philadelphia, 1985), Chap.4.