- Email: [email protected]
Buccal mucosa in vitro experiments
Journal of Controlled Release 58 (1999) 39–50
Buccal mucosa in vitro experiments I. Confocal imaging of vital staining and MTT assays for the determination of tissue viability 1, Delphine Imbert *, Christopher Cullander
Department of Biopharmaceutical Sciences, School of Pharmacy, University of California at San Francisco, 513 Parnassus Avenue, Box 0446, San Francisco, CA 94143 -0446, USA Received 7 April 1998; received in revised form 10 July 1998; accepted 29 July 1998
Abstract Delivery of drugs through the skin and the buccal mucosa has been considered as an alternative to per oral dosing for those substances that are degraded in the gastro-intestinal tract, or are subject to first-pass metabolism in the liver. In the buccal mucosa, contrary to skin, the diffusion barriers are located within living cell layers, hence the physiological state of the tissue is likely to significantly affect in vitro diffusion profiles. In this study, we were interested in assessing the viability of excised buccal mucosa and determining the limits of tissue usage under common in vitro experimental conditions. Using confocal laser scanning microscopy (CLSM), we have shown that optical sectioning of samples exposed to calcein AM and ethidium homodimer-1 (used as ‘live’ and ‘dead’ cell probes respectively) can be employed to accurately and reliably determine the viability of buccal mucosa biopsies. The results of the CLSM assay were remarkably consistent with that of an MTT assay. In both studies, viability in PBS at 348C was lost after about 8 h post-mortem, whereas it could be sustained for up to 24 h in KBR. 1999 Elsevier Science B.V. All rights reserved. Keywords: Buccal mucosa; Viability; Confocal microscopy; MTT assay; Vital staining
1. Introduction Delivery of drugs through the skin and the buccal mucosa has been considered as an alternative to per oral dosing for those substances (e.g., peptides and proteins) that are degraded in the gastro-intestinal tract, or are subject to first-pass metabolism in the *Corresponding author. Tel.: 11-650-802-1877; fax: 11-650595-2580; email: [email protected] 1 Current address: Cellegy Pharmaceuticals, Inc., 349 Oyster Point Boulevard, Suite 200, South San Francisco, CA, 94080, USA.
liver [1–3]. In transdermal delivery, the rate-limiting barrier for most compounds resides within the stratum corneum, thus, viability has been ignored in most in vitro skin studies unless effects of metabolism were specifically addressed [4]. In the buccal mucosa, however, the diffusion barriers are located within living cell layers, hence the physiological state of the tissue is likely to affect diffusion profiles obtained in vitro to a much greater extent than when working with skin or other keratinized epithelia [5]. In addition to obtaining reliable permeability coefficient, the use of viable tissue in diffusion experiments also opens up the scope of possibilities
0168-3659 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 98 )00143-6
40
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
for in vitro screening. Although more permeable than skin, buccal mucosa still represents an effective barrier to most xenobiotics and permeation enhancement strategies are required to provide therapeutic drug blood levels. One such approach—prodrug design—, for example, can only be evaluated in vitro with viable tissue. The use of viable tissue also allows a simultaneous in vitro evaluation of chemical permeation enhancers for both efficacy and toxicity [6]. Because it has not been well documented in the literature, we were interested in assessing the viability of excised pig buccal mucosa and determining the limits of tissue usage under typical in vitro experimental conditions. We examined viability using two assays which had not been developed for buccal mucosa before: (1) confocal laser scanning microscopy (CLSM) imaging of the vital stains calcein AM (CAM) and ethidium homodimer-1 (EH-1), and (2) MTT assay. The optical slicing capability of CLSM permits visualization of cells deep in thick tissue samples like buccal mucosa biopsies which are sufficiently transparent, do not scatter light strongly, and are not strongly autofluorescent. The combination of CAM (a non-fluorescent cell-permeant dye which is cleaved by intracellular esterases to fluorescent calcein) and EH-1 (which passes through damaged cell membranes to bind DNA and undergo a 40-fold enhancement in fluorescence) has proven useful for viability and cytotoxicity determination in monolayer cultures [7]. It has been used successfully in our laboratory to assess the effect of storage on the viability of both epithelium and endothelium in excised corneas [8]. Similarly, the MTT assay is a well-established viability and cytotoxicity assay in cell cultures [9], which has been adapted for the determination of skin sample viability [10,11] and, recently, in our laboratory for that of excised cornea [8].
2. Materials and methods
2,5-diphenyltetrazolium bromide) and dimethylsulfoxide (DMSO) were obtained from Sigma (Saint Louis, MO). Krebs Ringer bicarbonate (KBR) buffer was prepared within 24 h of use with 115.5 mM NaCl, 4.2 mM KCl, 2.5 mM CaCl 2 , 1.6 mM NaH 2 PO 4 , 0.8 mM MgSO 4 , 4.0 mM HEPES, 17.3 mM Na 2 CO 3 , and 12.2 mM glucose. Buffer pH was brought down to around 8.0 with dry ice and equilibrated to 7.5 with carbogen (95% O215% CO2) bubbling for about one h. After overnight storage at 48C, KBR was equilibrated to room temperature and its pH was adjusted before use (when necessary) by carbogen bubbling. All other chemicals were of analytical grade.
2.2. Animals Pig buccal mucosa was obtained from the UCSF Experimental Surgery Department (San Francisco, CA) through the UCSF Tissue Sharing Program. Pigs (females, 80–120 pounds) were sacrificed in the course of other studies. Both cheeks were isolated immediately using disposable razor blades and immersed in phosphate-buffered saline (PBS) pH 7.4 or Krebs Ringer bicarbonate buffer (KBR) pH 7.5 at room temperature, as indicated in text. In all studies, euthanasia and tissue excision occured within 5 min. Less than 15 min post-mortem, skin and connective tissues were rapidly removed with a disposable razor blade. The mucosa was then either dermatomed using a Padgett Electro Dermatome (Padgett Dermatome Division, Kansas City Assemblage Co., Kansas City, MO) set at 800 mm, or carefully dissected using surgical scissors. Sheets of mucosa were floated for up to 30 h onto various media, including PBS pH 7.4, PBS pH 7.4 with 0.02% NaN 3 , KBR pH 7.5, KBR pH 7.5 with 0.02% NaN 3 or KBR pH 7.5 without glucose. Viability experiments were conducted at room temperature, at 348C, or at 378C, as indicated in the text. In all graphs, the origin of the time axis refers to the time post-mortem (65 min), not to the time when the experiment was started.
2.1. Chemicals and reagents 2.3. Viability assays Ethidium homodimer-1 (EH-1) and calcein AM (CAM, the acetyloxymethyl ester of calcein) powder were purchased from Molecular Probes, Inc. (Eugene, OR). MTT (3-[4,5-dimethylthiazol-2-yl]-
2.3.1. MTT Assay MTT is a yellow, water-soluble compound that is enzymatically reduced to dark purple and insoluble
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
formazan by viable cells. The MTT assay protocol used in this study was adapted from earlier studies conducted in our laboratory with excised cornea [8]. The day of the assay, 2 mg / ml MTT was dissolved in freshly-prepared PBS. Any undissolved crystals were removed by filtration through a Nalgene 0.45-mm nylon syringe filter (Nalge Co., Rochester, NY). Each experiment was conducted using tissue from a single animal and three (up to six) replicates were obtained for each data points. Four-millimeter biopsy samples were obtained from the buccal tissue sheets using a disposable biopsy punch (Baker-Cummins Dermatologicals, Inc., Lakewood, NJ) at time points up to 27 h postmortem and then rinsed with the appropriate buffer. Excess solution was carefully removed and the biopsies were weighed and placed into individual wells of a six-well tissue culture plate (Becton Dickinson, Lincoln Park, NJ). Two milliliters of MTT solution were added to each well, and the plate was incubated for 2 h at 378C on a rotating platform (250 rpm). After incubation, the remaining MTT solution was removed by aspiration, and the tissue was rinsed twice with 1 ml of PBS for 1 min. Tissue was then minced with surgical scissors, and the formazan precipitate was extracted in 4 ml of DMSO, agitating with a tilted, rotating platform at 80 rpm for 80 min. Formazan absorbance was measured at 540 nm with DMSO as a blank. Formazan production has been linked to several enzymes [12]. For the purpose of this work, the group of enzymes responsible for the reduction of MTT was referred to collectively as tetrazolium reductase (TR) and data were expressed in absorbance units per mg tissue (TR index). Results were analyzed by one-way ANOVA and, when necessary, non-parametric Scheffe post-test using StatView 4.1 (Cherwell Scientific Publishing, Inc., Palo Alto, CA). Differences in values were reported as significant when P,0.05. The first assay (control) was typically begun within 30 min post-mortem. Control TR indices were dependent upon (1) the time elapsed between the tissue excision and the start of the assay and (2) the nature of the transfer buffer. Both are provided in Section 3. While concerns have been raised about the stability and sensitivity of tetrazolium salts in viability assays [12], MTT reduction has been shown to be dependent upon enzymes of the endoplasmic reticulum, unlike that of XTT (a
41
tetrazolium salt that is reduced to a soluble rather than insoluble formazan), which can be reduced under certain conditions without cells or enzymes. We also verified that MTT was chemically stable under our experimental protocol.
2.3.2. CLSM assay Four-millimeter biopsy samples were obtained from dermatomed buccal sheets floated on media at time points up to 27 h post-mortem and rinsed with the appropriate buffer. Each biopsy was immersed in 150 ml of PBS pH 7.4 or KBR pH 7.5 containing 25 mM EH-1 and 50 mM CAM prepared immediately before use from a 2 mM EH-1 stock solution in 4:1 water / DMSO and a 10 mM CAM stock solution in anhydrous DMSO. After staining (150 min), the sample was rinsed with buffer and positioned (without any processing or fixation) in a sample holder designed in our laboratory so that the anatomical surface of the tissue was orthogonal to the coverslip. To help support the biopsy, cylinder-shaped surgical sponges with cross-sections equal to that of the biopsy (4 mm) were used. The confocal microscope system used in this study was a BioRad MRC600 (BioRad, Hercules, CA) equipped with a Kr /Ar laser and mounted on a Nikon Optiphot microscope. Samples were simultaneously excited with the 488-and the 568-nm lines and imaged with Nikon dry (CF Fluor 103 / 0.5) or oil-immersion (CF Fluor DL 403 / 1.3 and CF N Plan Apochromat DM 603 / 1.40) lenses and a zoom of 1. Calcein and EH-1 fluorescence were detected with the BioRad K1 / K2 filterblock set. It should be noted that in samples stained with both probes and scanned simultaneously with the 488- and the 568nm (as in this study), a significant bleedthrough of the calcein signal was visible in the EH-1 channel. Such bleedthrough could be easily avoided by turning off the 488-nm line when acquiring EH-1 images. However, for the sake of simplicity and because the two probes had different staining patterns and each could be easily discerned, all confocal images in this study were acquired with simultaneous scanning of the 488- and the 568-nm lines. Optical sections were obtained in planes parallel to the cover slip (i.e., orthogonal to the anatomical surface of the sample) well below any cells damaged by the biopsy punch (Fig. 1). As described earlier [13,14] and as illustrated on Fig. 1, this orientation
42
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
Fig. 1. CLSM imaging of buccal mucosa biopsies. The tissue sample is positioned (without any processing or fixation) with the anatomical surface of the tissue perpendicular to the coverslip (not represented). Optical sections are obtained in planes perpendicular to the anatomical surface of the sample, well below any cells damaged by the biopsy punch. Such orientation of the tissue sample allows fluorescence imaging of all layers of the epithelium in the same optical plane.
of the tissue sample allows the imaging of all cell layers of the epithelium in the same optical plane (i.e., at the same depth below the coverslip). In preliminary experiments, the optimal imaging depth was determined using viable samples double-stained with EH-1 and CAM. In the area of the tissue immediately in contact with the cover slip (i.e., tissue that was sectioned by the biopsy punch), high levels of cell death could be observed on the EH-1 channel, and no calcein signal was detected. As the imaging plane was moved downward and viable cells were reached, the EH-1 signal decreased dramatically, and the calcein signal increased. As one proceeded further down, the calcein signal started decreasing, either because CAM might not have diffused that far, or simply because of attenuation from intervening cell layers. All our CLSM imaging was performed in a plane positioned about 10 mm below the plane of brightest calcein signal. This was
deep enough into the biopsy to ensure that no damage from the biopsy punch was visible; yet, it was close enough to the coverslip for good intensity and resolution. Typically, using the Nikon CF Fluor DL 403 / 1.3 lens, we imaged at a depth of about 40 mm, while with the Nikon CF Fluor 103 / 0.5 (which gave a much thicker optical section), imaging could performed at a depth of about 70 mm. EH-1 images were analyzed using the particle counting capability of NIH image V. 1.60 and the downloadable confocal macros. Briefly, EH-1 images were opened in NIH Image with a red density slice set so that all EH-1-stained nuclei appeared in red. Outlining, labeling and counting of the color-coded nuclei was then performed automatically using the Analyze Particles function of the Analyze menu. The area analyzed was also measured and results were expressed as number of EH-1 stained nuclei per mm 2 of tissue imaged.
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
43
3. Results and discussion
3.1. MTT assay A significant decrease in TR index was observed with dissected buccal mucosa stored in PBS at room temperature as early as 3 h post-mortem (Fig. 2, open symbols). The influence of temperature on viability was studied in a separate experiment, using tissue from a second animal (solid symbols). As shown on Fig. 2, TR indices at 378C were significantly lower than those at room temperature (218C628C) at all time points. Additionally, the close similarity between the two sets of data obtained at room temperature from two different animals illustrate the robustness and good reproducibility of the assay. Although PBS is still used by some groups as a flow-through buffer [15], most now use KBR-type buffers in an effort to sustain the viability of the membrane. We wanted to evaluate whether the viability of dermatomed mucosa was indeed improved by such buffer systems. Moreover, because KBR buffers contain glucose, and due to the high bacterial load of the buccal mucosa, the influence of sodium azide was also investigated. Fig. 3 shows that, at room temperature, the difference in viability
Fig. 2. MTT assay of dissected buccal mucosa: Influence of temperature. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in PBS at room temperature. Buccal mucosa was carefully dissected using surgical scissors, floated on PBS at room temperature and assayed at various time points post-mortem (Mean6s.e., n53–5).
Fig. 3. MTT assay of dermatomed buccal mucosa: Influence of medium composition at room temperature. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in KBR at room temperature. Buccal mucosa was dermatomed to 800 mm, stored in various media at room temperature, and assayed 3 h and 20 h post-mortem (Mean6s.e., n53).
after 2.5 h in PBS and in KBR is significant, although not as important as we expected. After 20 h, however, PBS yielded TR index 50% lower than those obtained from KBR. Additionally, no significant differences were found between PBS and KBR lacking glucose, or when comparing media with or without sodium azide. In their MTT assay with skin, Hood et al. used a 50% decrease from control as a limit for viable samples [11]. Control samples in KBR assayed within 1 h post-mortem reproducibly yielded TR indices around 0.05 (Figs. 3 and 4). Thus, by Hood’s criteria (TR indices above 0.025 are from viable samples), after 20 h in KBR (with or without sodium azide) at room temperature, buccal samples were still viable (TR indices of 0.035 and 0.038, respectively) whereas those in PBS or in KBR without glucose were not (TR indices of 0.020 and 0.021 respectively). Because most in vitro diffusion studies are conducted at 348C, a study similar to that in Fig. 3 was conducted at that temperature. TR indices for KBR samples were maintained at 85% of the control value after 6.3 h and at 72% after 9.3 h (Fig. 4a). Contrary to Fig. 3, however, striking differences in TR indices between KBR and PBS samples were visible at all time points. As early as 3.3 h post-mortem, TR
44
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
(Fig. 6). In PBS, there was a significant increase in nuclear staining with time for about 9 h post-mortem. In some cases, this increase was highly linear (Fig. 6, closed circles, R 2 50.90); in others, however, the trend was not as clearly marked, particularly at higher temperatures (Fig. 6, open squares). Nevertheless, in all cases, about 9 h post-mortem, the number of EH-1 nuclei in the KBR samples was significantly lower (|50%) than that in the PBS samples, which is remarkably consistent with the MTT assay. After 10 h in PBS, loss of tissue structure could be observed by CLSM, and translated into an apparent decrease of the number of EH-1 nuclei. Despite these encouraging results, however, the EH-1 signal acquisition and quantification was tedious and often provided noisy results at early time points. Therefore, we concentrated our efforts on the calcein results. CLSM images of the calcein fluorescence detected at the basal level of buccal mucosa samples stored at 348C in PBS or in KBR are shown in, respectively, Figs. 7 and 8. After about 9 h post-mortem, no calcein fluorescence could be elicited from the PBS samples as illustrated on Fig. 7d. In KBR samples, however, calcein fluorescence was detected more than 24 h post-mortem (Fig. 8d). Fig. 4. MTT assay of dermatomed buccal mucosa: Influence of medium composition at 348C. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in KBR (a) or PBS (b) at room temperature. Buccal mucosa was dermatomed to 800 mm (a) or dissected (b), stored in various media at 348C, and assayed at various time points post-mortem (Mean6s.e., n53).
indices from PBS samples were down to 0.03 (i.e., 37% decrease from the KBR control). A 50% decrease was reached at 6.3 h. These results were confirmed in a separate experiment (Fig. 4b).
3.2. CLSM assay In the CLSM assay, EH-1 images of buccal mucosa kept at room temperature in PBS showed an increase in nuclear staining with time (Fig. 5). Using standard image analysis techniques, the number of EH-1 stained nuclei per unit surface area imaged was quantified and plotted as a function of time postmortem for tissue samples stored in PBS or KBR
3.3. Discussion Viability is a parameter that is gaining increased attention in studies where freshly excised membranes are used to mimic in vivo conditions, as in in vitro diffusion experiments. With skin, viability may not critically affect in vitro diffusion behavior, provided that the stratum corneum is the main barrier to diffusion, an hypothesis which holds true in many circumstances. However, in instances where metabolism is of interest, where the compound’s diffusion is epidermis-limited, and where non-keratinized epithelia (like buccal mucosa and cornea) are studied, viability should be carefully evaluated as it has the potential to directly affect results obtained in vitro. Numerous viability assays have been developed for a large array of cell lines and cytotoxic events, including 51 Cr release, 3 H-thymidine uptake [16], trypan blue exclusion, ATP and glucose utilization, tetrazolium salts colorimetric assays [16,17] and fluorescent vital staining. Only a limited number of these assays, however,
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
45
Fig. 5. CLSM imaging of EH-1 fluorescence in the upper layers of the buccal mucosa at 2.5 h (left) and 7.0 h (right) post-mortem. Scale bar represents 100 mm.
have been or can be adapted to the study of thick tissue sample viability. The tetrazolium salts assays [10,11,18–21] and the glucose utilization assays
Fig. 6. Quantitative assessment of viability by CLSM. Surface density of nuclei stained by EH-1 as a function of time postmortem at the end of staining. Buccal mucosa was dermatomed to 800 mm and stored in PBS at room temperature (d), PBS at 348C (h), or KBR at 348C (s). Each experiment was conducted separately using tissue from a different animal. One biopsy was imaged throughout the thickness of the epithelium at each time points, at 2–4 different locations. Straight lines indicate regression lines for PBS at room temperature (d, 2–9 h, R 2 50.90) and KBR at 348C (s, 2–11 h, R2 50.27).
[4,22,23] are two popular skin viability assays. Thymidine uptake [24] and trypan blue exclusion [25,26] were used recently with cornea. Ferrera et al. determined heart biopsies viability using the MTT bioassay [20]. Recently, Potter and coworkers [27] used nuclear fluorescence staining to estimate in vivo tissue injury following laser surgery. However, fluorescent viability probes, although widely used in cell cultures, have only found a limited number of application with thick samples [28]. Typically, buccal tissue viability has been assessed through a combination of ATP measurements [29] and histopathological examinations [30], complemented (or not) by linearity of transport data [30,31]. One could argue that the latter is a test of integrity rather than viability, as are electrophysiological measurements (transepithelial potential and resistance), which are also commonly used not only for buccal mucosa [32–34], but for other epithelia as well [33,35]. We have addressed this point in detail elsewhere [36,37]. Nevertheless, despite this awareness, only limited viability data are published in these studies. This is of concern with respect to the interpretation of flux data when most laboratories using porcine buccal mucosa get their tissue from slaughterhouses, and the time of death (often unknown or rarely reported) seldom coincides with the
46
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
Fig. 7. CLSM imaging of calcein fluorescence in the basal layers of dermatomed buccal mucosa stored in PBS at 348C, 2.8 h (a), 4.8 h (b), 6.8 h (c), and 8.8 h post-mortem. Scale bar represents 100 mm.
time of tissue excision. Furthermore, once excised, although it is a common practice to keep buccal tissue in KBR or PBS (ice-cold during transfer from the slaughterhouses, and at 348C during diffusion experiments), limited data are available to document the extent to which viability is actually maintained with these media [38]. In our laboratory, the MTT assay and CLSM vital staining with the EH-1 / CAM probe pair were used earlier to assess the viability of excised rabbit corneas. In this study, we adapted these two assays for the study of buccal mucosa viability. Contrary to the common belief that buccal mucosa viability is relatively stable for about 6 h [32], we have shown that TR activity in dissected mucosa decreases sharply in the first 10 h post-mortem at room temperature (Fig. 2). These experimental conditions are close to those encountered in the slaughterhouse between the time the animals are sacrificed
and the time the tissue is actually excised by the researchers. The decrease in viability is sharper as the temperature is increased to 348C (Figs. 3 and 4) and 378C (Fig. 2). We have not investigated the rate of TR decrease at 48C, however it is likely to be slower than at the above temperatures. This would constitute relevant information, however, for laboratories obtaining tissue from carcasses stored overnight in cold rooms [39]. Having established that the MTT assay can successfully discriminate between various experimental parameters (time and temperature) and shows a good inter-animal reproducibility, buccal mucosa viability was then determined under conditions closer to those used in in vitro experimental setups. With a few exceptions [15,40], most in vitro buccal studies are conducted using dermatomed mucosa with KBR rather than PBS as diffusion cell medium [41,33]. At room temperature, the difference between KBR and
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
47
Fig. 8. CLSM imaging of calcein fluorescence in the basal layers of dermatomed buccal mucosa stored in KBR at 348C, 2.7 h (a), 8.7 h (b), 10.7 h (c), and 25 h post-mortem. Scale bar represents 100 mm.
PBS was statistically significant but practically minimal (Fig. 3). At 348C, however, the temperature of choice for in vitro diffusion studies, differences between the two media were dramatically accentuated. As early as 3.3 h post-mortem, PBS samples yielded TR indices 60% lower than those from KBR samples (Fig. 4b), whereas buccal mucosa viability in KBR was maintained above 72% of its original value for more than 9 h post-mortem (Fig. 4a). The influence of sodium azide (NaN 3 ), an antimicrobial commonly used in in vitro skin studies, was also investigated because (1) NaN 3 is known to have adverse effects on skin viability and may therefore affect buccal tissue viability, (2) buccal mucosa has a high bacterial load, and (3) media for buccal work usually contain high levels of glucose, which promotes bacterial growth. Although it appeared to have a significant effect at some time points, the addition of 0.02% sodium azide to PBS or KBR did not
consistently affect viability compared to the original media. Glucose however, was shown to be indispensable to the maintenance of TR activity in the tissue (Fig. 3). These MTT results were remarkably consistent with both the EH-1 and calcein results obtained by CLSM. Obviously, as opposed to the MTT assay, CLSM imaging cannot be used as a routine laboratory assay for viability determination and screening. Furthermore, as we showed in this study, CLSM does not readily lend itself to obtaining quantitative data. Nevertheless, CAM staining reliably confirmed the results obtained through the MTT assay. Additionally, contrary to SEM / TEM, it allows the direct visualization of unfixed and viable samples, which, while not directly valuable in this particular study, could potentially be very informative in other instances [6]. For example, one can see how CAM staining could be conducted in parallel with the use
48
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
of a second probe (an enzyme maker or a diffusion probe excitable by the 568-nm or 647-nm line of the KrAr laser) to ensure tissue viability during confocal imaging of some other event of interest. Although it is beyond the scope of this paper to address this issue in details, it should be noted that in viable CAM-stained samples, we consistently observed that calcein fluorescence was much brighter at the basal level than anywhere else in the tissue. Consequently, our CLSM assay was based primarily on fluorescence signal from the basal level of the epithelium, which we found to be reliably consistent with the EH-1 signal and the results from the MTT assay. Although the localization of calcein fluorescence did not prevent us from drawing conclusions about the viability of our biopsy samples, it was sufficiently intriguing that we are currently trying to understand it in more details. Three hypotheses are currently being investigated in our laboratory to explain this phenomenon: (1) there is a preferential diffusion pathway for CAM from the connective tissue towards the epithelium or along the basal membrane, and the area of bright calcein fluorescence actually represents CAM’s diffusion front towards the surface of the epithelium; (2) CAM is able to diffuse throughout the epithelium, but is only cleaved to calcein in the lower cell layers of the epithelium, indicating higher esterase activity in this part of the tissue, or finally, (3) CAM is able to diffuse throughout the whole epithelium, it is cleaved by esterases inside all viable cells, but calcein fluorescence is quenched (pH or concentration effect) in the upper part of the tissue. A concentration-induced quenching effect is unlikely considering that staining is performed using a 50 mM CAM solution and that calcein self-quenching was reported to occur at concentrations above 100 mM. The existence of a pH gradient within the tissue would certainly be of interest to drug delivery laboratories like ours, however, it is currently difficult to demonstrate due to the lack of pH probes excitable by a Kr /Ar laser. We are currently focusing on quantifying the esterase activity in various levels of the tissue (hypothesis 2). Hot spots of enzyme activities have been observed earlier in skin [13]. Moreover, recently, Gherzi et al. reported high expression levels of a glucose transporter in the basal cells of oral mucosa [42].
Based on the results of this study, we are currently conducting our in vitro diffusion experiments at 348C with KBR for no longer than 8 h post-mortem. As mentioned briefly earlier, the question remains whether viability is directly linked to tissue integrity, i.e. whether diffusion pathways are altered as tissue viability is gradually lost. In experiments designed to measure permeability coefficients, this may be the only relevant issue, and already some groups have focused their efforts on demonstrating linearity of transport during the time frame of the experiment [30,43]. Studies conducted in our laboratory have shown that care should be taken in choosing the model permeant for such investigation [36,37]. Moreover, as detailed earlier, for experiments with broader goals [6], this information is clearly not sufficient.
4. Conclusions CLSM used in combination with CAM and EH-1 (used as ‘live’ and ‘dead’ probes respectively) is a powerful tool to assay the viability of unfixed, thick buccal mucosa samples submitted to various storage conditions. Although CLSM cannot realistically be the method of choice when viability is the only variable of interest, it becomes a powerful tool, however, when viability probes are used in combination with other reporting probes to monitor two events at once. The MTT assay, on the other hand, proved to be an easy, reliable, sensitive, and reproducible assay that can be routinely performed in most laboratories. To our knowledge, it is the first time that this assay is used for the determination of buccal tissue viability. It is a more convenient alternative than histology, electrophysiology, or ATP level determination, all of which were used earlier with this tissue. Recently, we have also shown it can be reliably used for toxicity screening purposes [6]. Using these two assays, we showed that buccal mucosa viability is significantly altered in PBS in the first 8 h following tissue collection. In KBR, viability was maintained well above 75% of the original value for about 6 h. Work is now under way in our laboratory to determine whether the loss of viability of buccal tissue in PBS is associated with alterations in in vitro permeation behavior, and vice-versa,
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
whether maintenance of viability in KBR is synonym with preservation of in vivo diffusion pathways [36,37].
[9] [10]
Acknowledgements The authors wish to thank Harolyn Hood (FDA, Washington DC) for sharing her MTT protocol, Ron Baireuther and Gabriela Fuentes from UCSF Experimental Surgery for providing pig tissue on a regular basis, and Willy Fong and Joanne Lee (UC Berkeley) for their help with some of the MTT assays. NIH Image is a public domain program developed at the U.S. National Institutes of Health and available on the Internet at http: / / rsb.info.nih.gov / nih-image /. This study was supported by NIH-DE11275 (CC and DI), Fondation Singer Polignac (DI), and an AACP New Investigator Grant (CC).
[11]
[12]
[13]
[14]
[15]
[16]
References ´ A.G. de Boer, J.A. Bouwstra, H.E. [1] J.C. Verhoef, H.E. Bodde, Junginger, F.W. Merkus, D.D. Breimer, Transport of peptide and protein drugs across biological membranes, Eur. J. Drug Metab. Pharmacokinet. 15(2) (1990) 83–93. [2] A.J. Hoogstraate, J.C. Verhoef, A. Pijpers, L.A.M.G. Van ´ In Leengoed, J.H.M. Verhijden, H.E. Junginger, H.E. Bodde, vivo buccal delivery of the peptide drug buserelin with glycodeoxycholate as an absorption enhancer in pigs, Pharm. Res. 13(8) (1996) 1233–1237. [3] M.J. Rathbone, B.K. Drummond, I.G. Tucker, The oral cavity as a site for systemic drug delivery, Adv. Drug Deliv. Rev. 13 (1994) 1–22. [4] S.W. Collier, N.M. Sheikh, A. Sakr, J.L. Lichtin, R.F. Stewart, R.L. Bronaugh, Maintenance of skin viability during in vitro percutaneous absorption / metabolism studies, Toxicol. Appl. Pharmacol. 99(3) (1989) 522–533. [5] P.W. Wertz, C.A. Squier, Cellular and molecular basis of barrier function in oral epithelium, Crit. Rev. Ther. Drug Carrier Syst. 8(3) (1991) 237–269. [6] D. Imbert, C. Cullander, Assessing penetration enhancers for buccal drug delivery: tools for simultaneous efficacy and toxicity screening using viable tissue samples. Proceed. Int’l. Symp. Control. Rel. Bioact. Mater., 25 (1998), 26–27. [7] P.L. Moore, I.C. Maccoubrey, R.P. Haugland, A rapid pH insensitive two color fluorescence viability cytotoxicity assay, J. Cell Biol. 111(5 part 2) (1990) 58A. [8] D. Imbert, C. Cullander, Assessment of cornea viability by
[17]
[18]
[19]
[20]
[21] [22]
[23]
49
confocal laser scanning microscopy and MTT assay, Cornea 16(6) (1997) 666–674. R. Supino, MTT assays, Methods Mol. Biol. 43 (1995) 137–149. F.B. Hershey, C.N.D. Cruickshank, L.I. Mullins, The quantitative reduction of 2,3,5-triphenyl tetrazolium chloride by skin in vitro, J. Histochem. Cytochem. 6 (1958) 191–196. H.L. Hood, M.G. Robl, R.L. Bronaugh, Skin absorption methodology: Improving the lipophilicity of the receptor fluid and the accuracy of the skin viability assay, Fundam. Appl. Toxicol. 30(1) (1996) 168. M.V. Berridge, A.S. Tan, K.D. McCoy, R. Wang, The biochemical and cellular basis of cell proliferation assays that use tetrazolium salts, Biochemica 4 (1996) 15–20. P. Boderke, H.P. Merkle, C. Cullander, M. Ponec, H.E. ´ Localization of aminopeptidase activity in freshly Bodde, excised human skin: direct visualization by confocal laser scanning microscopy, J. Invest. Dermatol. 108(1) (1997) 83–86. M.E.M.J. Meuwissen, J. Janssen, C. Cullander, H.E. Junginger, J. Bouwstra, A cross-section device to improve visualization of fluorescent probe penetration into the skin by confocal laser scanning microscopy, Pharm. Res. 15(2) (1998) 352–356. C.A. Squier, M. Kremer, P.W. Wertz, Continuous flow mucosal cells for measuring the in-vitro permeability of small tissue samples, J. Pharm. Sci. 86(1) (1997) 82–84. R.S. Gieni, Y. Li, K.T. HayGlass, Comparison of [ 3 H]thymidine incorporation with MTT- and MTS-based bioassays for human and murine IL-2 and IL-4 analysis. Tetrazolium assays provide markedly enhanced sensitivity, J. Immunol. Methods 187(1) (1995) 85–93. T. Mosmann, Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays, J. Immunol. Methods 65 (1983) 55–63. J.N. Kearney, L.A. Wheldon, G. Gowland, Cryopreservation of skin using a murine model: validation of a prognostic viability assay, Cryobiology 27(1) (1990) 24–30. R.F. Brown, A.R. Groves, Succinate dehydrogenase as an index of thermal injury to skin, Br. J. Exp. Pathol. 54(2) (1973) 117–122. R. Ferrera, A. Larese, F. Berthod, J. Guidollet, C. Rodriguez, G. Dureau, A. Dittmar, Quantitative reduction of MTT by hearts biopsies in vitro is an index of viability, J. Mol. Cell. Cardiol. 25(9) (1993) 1091–1099. R.F.R. Brown, J.V. Harvey Kemble, Tetrazolium reductase as an index of viability of stored skin, Burns 1 (1975) 179–185. M. Kietzmann, W. Loscher, D. Arens, P. Maass, D. Lubach, The isolated perfused bovine udder as an in vitro model of percutaneous drug absorption. Skin viability and percutaneous absorption of dexamethasone, benzoyl peroxide, and etofenamate, J. Pharmacol. Toxicol. Methods 30(2) (1993) 75–84. R.C. Wester, J. Christoffel, T. Hartway, N. Poblete, H.I. Maibach, J. Forsell, Human cadaver skin viability for in vitro percutaneous aborption: Storage and detrimental effects of heat-separation and freezing, Pharm. Res. 15(1) (1998) 82– 84.
50
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
[24] J.H. Lass, W.J. Reinhart, D.L. Skelnik, W.E. Bruner, R.P. Shockley, J.Y. Park, D.L. Hom, R.L. Lindstrom, An in vitro and clinical comparison of corneal storage with chondroitin sulfate corneal storage medium with and without dextran, Ophthalmology 97(1) (1990) 96–103. [25] M. Morton, Corneal storage at 4 degrees Celsius in a chondroitin sulphate and dextran containing medium, Aust. N.Z. J. Ophthalmol. 20(3) (1992) 211–213. [26] M. Canals-Imohr, J. Costa-Vila, D. Ruano-Gil, D. PitaSalorio, Comparative study of current corneal short-term, medium-, and long-term preservation methods, Transplant Proc. 27(4) (1995) 2418. [27] R.F. Potter, G. Peters, M. Carson, T. Forbes, C.G. Ellis, K.A. Harris, G. DeRose, W.G. Jamieson, Measurement of tissue viability using intravital microscopy and fluorescent nuclear dyes, J. Surg. Res. 59(5) (1995) 521–526. [28] C.A. Poole, N.H. Brookes, G.M. Clover, Keratocyte networks visualised in the living cornea using vital dyes, J. Cell Sci. 106(2) (1993) 685–691. [29] D. Harris, J.R. Robinson, Drug delivery via the mucous membranes of the oral cavity, J. Pharm. Sci 81(1) (1992) 1–10. [30] M.E. Dowty, K.E. Knuth, B.K. Irons, J.R. Robinson, Transport of thyrotropin releasing hormone in rabbit buccal mucosa in vitro, Pharm. Res. 9(9) (1992) 1113–1122. [31] M. Nair, Y.W. Chien, I. Buccal Delivery of Progestational Steroids, Characterization of barrier properties and effect of penetrant hydrophilicity, Int. J. Pharm. 89(1) (1993) 41–49. [32] C.R. Bland, S.S. Davis, D.A. Rawlins, Evaluation of the viability of an in vitro porcine buccal mucosa preparation, Proceed. Intern. Symp. Control. Rel. Bioact. Mater. 18 (1991) 499–500. [33] Y. Rojanasakul, L.Y. Wang, M. Bhat, D.D. Glover, C.J. Malanga, J.K. Ma, The transport barrier of epithelia: a comparative study on membrane permeability and charge selectivity in the rabbit, Pharm. Res. 9(8) (1992) 1029– 1034. ´ J.C. Verhoef, M. Ponec, W. [34] M.E. de Vries, H.E. Bodde,
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
Craane, H.E. Junginger, Localization of the permeability barrier inside porcine buccal mucosa: a combined in vitro study of drug permeability, electrical resistance and tissue morphology, Int. J. Pharm. 76(1-2) (1991) 25–35. C. Bjerre, B. Sjostrom, E. Bjork, O. Camber, Pig nasal tissue: a characterization and in vitro permeation study, Pharm. Res. 13(9) (1996) S–171. D. Imbert, C. Cullander, Permeability of buccal mucosa in vitro: viability versus integrity, J. Dent. Res. 77(Suppl. A) (1998) 290. D. Imbert, C. Cullander, Buccal mucosa in vitro experiments: II. Influence of viability on tissue permeability, J. Control. Rel., (1998) Manuscript in preparation, to be submitted. ´ Methods for assessing the A.J. Hoogstraate, H.E. Bodde, buccal mucosa as a route of drug delivery, Adv. Drug Del. Rev. 12 (1993) 99–125. V.A. Moore, I.W. Kellaway, P. Timmins, A. Dennis, Development and chraracterization of an ex vivo buccal mucosal model to assess solute permeability, Pharm. Res. 14(11) (1997) S–660. C.A. Lesch, C.A. Squier, A. Cruchley, D.M. Williams, P. Speight, The permeability of human oral mucosa and skin to water, J. Dent. Res. 68(9) (1989) 1345–1349. A.J. Hoogstraate, S. Senel, C. Cullander, J. Verhoef, H.E. ´ Effects of bile salts on transport rates Junginger, H.E. Bodde, and routes of FITC-labelled compounds across porcine buccal epithelium in vitro, J. Control. Rel. 40(3) (1996) 211–221. R. Gherzi, G. Melioli, M. De Luca, A. D’Agostino, M. Guastella, C.E. Traverso, F. D’Anna, A.T. Franzi, R. Cancedda, High expression levels of the ‘erythroid / brain’ type glucose transporter (GLUT1) in the basal cells of human eye conjunctiva and oral mucosa reconstituted in culture, Exp. Cell Res. 195(1) (1991) 230–236. B. Polentarutti, A. Sjoberg, A. Peterson, E.-K. Anderberg, A.-L. Ungell, Viability of rat intestinal segments in vitro, Proceed. Int. Symp. Control. Rel. Bioact. Mater. 24 (1997) 333–334.
Buccal mucosa in vitro experiments I. Confocal imaging of vital staining and MTT assays for the determination of tissue viability 1, Delphine Imbert *, Christopher Cullander
Department of Biopharmaceutical Sciences, School of Pharmacy, University of California at San Francisco, 513 Parnassus Avenue, Box 0446, San Francisco, CA 94143 -0446, USA Received 7 April 1998; received in revised form 10 July 1998; accepted 29 July 1998
Abstract Delivery of drugs through the skin and the buccal mucosa has been considered as an alternative to per oral dosing for those substances that are degraded in the gastro-intestinal tract, or are subject to first-pass metabolism in the liver. In the buccal mucosa, contrary to skin, the diffusion barriers are located within living cell layers, hence the physiological state of the tissue is likely to significantly affect in vitro diffusion profiles. In this study, we were interested in assessing the viability of excised buccal mucosa and determining the limits of tissue usage under common in vitro experimental conditions. Using confocal laser scanning microscopy (CLSM), we have shown that optical sectioning of samples exposed to calcein AM and ethidium homodimer-1 (used as ‘live’ and ‘dead’ cell probes respectively) can be employed to accurately and reliably determine the viability of buccal mucosa biopsies. The results of the CLSM assay were remarkably consistent with that of an MTT assay. In both studies, viability in PBS at 348C was lost after about 8 h post-mortem, whereas it could be sustained for up to 24 h in KBR. 1999 Elsevier Science B.V. All rights reserved. Keywords: Buccal mucosa; Viability; Confocal microscopy; MTT assay; Vital staining
1. Introduction Delivery of drugs through the skin and the buccal mucosa has been considered as an alternative to per oral dosing for those substances (e.g., peptides and proteins) that are degraded in the gastro-intestinal tract, or are subject to first-pass metabolism in the *Corresponding author. Tel.: 11-650-802-1877; fax: 11-650595-2580; email: [email protected] 1 Current address: Cellegy Pharmaceuticals, Inc., 349 Oyster Point Boulevard, Suite 200, South San Francisco, CA, 94080, USA.
liver [1–3]. In transdermal delivery, the rate-limiting barrier for most compounds resides within the stratum corneum, thus, viability has been ignored in most in vitro skin studies unless effects of metabolism were specifically addressed [4]. In the buccal mucosa, however, the diffusion barriers are located within living cell layers, hence the physiological state of the tissue is likely to affect diffusion profiles obtained in vitro to a much greater extent than when working with skin or other keratinized epithelia [5]. In addition to obtaining reliable permeability coefficient, the use of viable tissue in diffusion experiments also opens up the scope of possibilities
0168-3659 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 98 )00143-6
40
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
for in vitro screening. Although more permeable than skin, buccal mucosa still represents an effective barrier to most xenobiotics and permeation enhancement strategies are required to provide therapeutic drug blood levels. One such approach—prodrug design—, for example, can only be evaluated in vitro with viable tissue. The use of viable tissue also allows a simultaneous in vitro evaluation of chemical permeation enhancers for both efficacy and toxicity [6]. Because it has not been well documented in the literature, we were interested in assessing the viability of excised pig buccal mucosa and determining the limits of tissue usage under typical in vitro experimental conditions. We examined viability using two assays which had not been developed for buccal mucosa before: (1) confocal laser scanning microscopy (CLSM) imaging of the vital stains calcein AM (CAM) and ethidium homodimer-1 (EH-1), and (2) MTT assay. The optical slicing capability of CLSM permits visualization of cells deep in thick tissue samples like buccal mucosa biopsies which are sufficiently transparent, do not scatter light strongly, and are not strongly autofluorescent. The combination of CAM (a non-fluorescent cell-permeant dye which is cleaved by intracellular esterases to fluorescent calcein) and EH-1 (which passes through damaged cell membranes to bind DNA and undergo a 40-fold enhancement in fluorescence) has proven useful for viability and cytotoxicity determination in monolayer cultures [7]. It has been used successfully in our laboratory to assess the effect of storage on the viability of both epithelium and endothelium in excised corneas [8]. Similarly, the MTT assay is a well-established viability and cytotoxicity assay in cell cultures [9], which has been adapted for the determination of skin sample viability [10,11] and, recently, in our laboratory for that of excised cornea [8].
2. Materials and methods
2,5-diphenyltetrazolium bromide) and dimethylsulfoxide (DMSO) were obtained from Sigma (Saint Louis, MO). Krebs Ringer bicarbonate (KBR) buffer was prepared within 24 h of use with 115.5 mM NaCl, 4.2 mM KCl, 2.5 mM CaCl 2 , 1.6 mM NaH 2 PO 4 , 0.8 mM MgSO 4 , 4.0 mM HEPES, 17.3 mM Na 2 CO 3 , and 12.2 mM glucose. Buffer pH was brought down to around 8.0 with dry ice and equilibrated to 7.5 with carbogen (95% O215% CO2) bubbling for about one h. After overnight storage at 48C, KBR was equilibrated to room temperature and its pH was adjusted before use (when necessary) by carbogen bubbling. All other chemicals were of analytical grade.
2.2. Animals Pig buccal mucosa was obtained from the UCSF Experimental Surgery Department (San Francisco, CA) through the UCSF Tissue Sharing Program. Pigs (females, 80–120 pounds) were sacrificed in the course of other studies. Both cheeks were isolated immediately using disposable razor blades and immersed in phosphate-buffered saline (PBS) pH 7.4 or Krebs Ringer bicarbonate buffer (KBR) pH 7.5 at room temperature, as indicated in text. In all studies, euthanasia and tissue excision occured within 5 min. Less than 15 min post-mortem, skin and connective tissues were rapidly removed with a disposable razor blade. The mucosa was then either dermatomed using a Padgett Electro Dermatome (Padgett Dermatome Division, Kansas City Assemblage Co., Kansas City, MO) set at 800 mm, or carefully dissected using surgical scissors. Sheets of mucosa were floated for up to 30 h onto various media, including PBS pH 7.4, PBS pH 7.4 with 0.02% NaN 3 , KBR pH 7.5, KBR pH 7.5 with 0.02% NaN 3 or KBR pH 7.5 without glucose. Viability experiments were conducted at room temperature, at 348C, or at 378C, as indicated in the text. In all graphs, the origin of the time axis refers to the time post-mortem (65 min), not to the time when the experiment was started.
2.1. Chemicals and reagents 2.3. Viability assays Ethidium homodimer-1 (EH-1) and calcein AM (CAM, the acetyloxymethyl ester of calcein) powder were purchased from Molecular Probes, Inc. (Eugene, OR). MTT (3-[4,5-dimethylthiazol-2-yl]-
2.3.1. MTT Assay MTT is a yellow, water-soluble compound that is enzymatically reduced to dark purple and insoluble
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
formazan by viable cells. The MTT assay protocol used in this study was adapted from earlier studies conducted in our laboratory with excised cornea [8]. The day of the assay, 2 mg / ml MTT was dissolved in freshly-prepared PBS. Any undissolved crystals were removed by filtration through a Nalgene 0.45-mm nylon syringe filter (Nalge Co., Rochester, NY). Each experiment was conducted using tissue from a single animal and three (up to six) replicates were obtained for each data points. Four-millimeter biopsy samples were obtained from the buccal tissue sheets using a disposable biopsy punch (Baker-Cummins Dermatologicals, Inc., Lakewood, NJ) at time points up to 27 h postmortem and then rinsed with the appropriate buffer. Excess solution was carefully removed and the biopsies were weighed and placed into individual wells of a six-well tissue culture plate (Becton Dickinson, Lincoln Park, NJ). Two milliliters of MTT solution were added to each well, and the plate was incubated for 2 h at 378C on a rotating platform (250 rpm). After incubation, the remaining MTT solution was removed by aspiration, and the tissue was rinsed twice with 1 ml of PBS for 1 min. Tissue was then minced with surgical scissors, and the formazan precipitate was extracted in 4 ml of DMSO, agitating with a tilted, rotating platform at 80 rpm for 80 min. Formazan absorbance was measured at 540 nm with DMSO as a blank. Formazan production has been linked to several enzymes [12]. For the purpose of this work, the group of enzymes responsible for the reduction of MTT was referred to collectively as tetrazolium reductase (TR) and data were expressed in absorbance units per mg tissue (TR index). Results were analyzed by one-way ANOVA and, when necessary, non-parametric Scheffe post-test using StatView 4.1 (Cherwell Scientific Publishing, Inc., Palo Alto, CA). Differences in values were reported as significant when P,0.05. The first assay (control) was typically begun within 30 min post-mortem. Control TR indices were dependent upon (1) the time elapsed between the tissue excision and the start of the assay and (2) the nature of the transfer buffer. Both are provided in Section 3. While concerns have been raised about the stability and sensitivity of tetrazolium salts in viability assays [12], MTT reduction has been shown to be dependent upon enzymes of the endoplasmic reticulum, unlike that of XTT (a
41
tetrazolium salt that is reduced to a soluble rather than insoluble formazan), which can be reduced under certain conditions without cells or enzymes. We also verified that MTT was chemically stable under our experimental protocol.
2.3.2. CLSM assay Four-millimeter biopsy samples were obtained from dermatomed buccal sheets floated on media at time points up to 27 h post-mortem and rinsed with the appropriate buffer. Each biopsy was immersed in 150 ml of PBS pH 7.4 or KBR pH 7.5 containing 25 mM EH-1 and 50 mM CAM prepared immediately before use from a 2 mM EH-1 stock solution in 4:1 water / DMSO and a 10 mM CAM stock solution in anhydrous DMSO. After staining (150 min), the sample was rinsed with buffer and positioned (without any processing or fixation) in a sample holder designed in our laboratory so that the anatomical surface of the tissue was orthogonal to the coverslip. To help support the biopsy, cylinder-shaped surgical sponges with cross-sections equal to that of the biopsy (4 mm) were used. The confocal microscope system used in this study was a BioRad MRC600 (BioRad, Hercules, CA) equipped with a Kr /Ar laser and mounted on a Nikon Optiphot microscope. Samples were simultaneously excited with the 488-and the 568-nm lines and imaged with Nikon dry (CF Fluor 103 / 0.5) or oil-immersion (CF Fluor DL 403 / 1.3 and CF N Plan Apochromat DM 603 / 1.40) lenses and a zoom of 1. Calcein and EH-1 fluorescence were detected with the BioRad K1 / K2 filterblock set. It should be noted that in samples stained with both probes and scanned simultaneously with the 488- and the 568nm (as in this study), a significant bleedthrough of the calcein signal was visible in the EH-1 channel. Such bleedthrough could be easily avoided by turning off the 488-nm line when acquiring EH-1 images. However, for the sake of simplicity and because the two probes had different staining patterns and each could be easily discerned, all confocal images in this study were acquired with simultaneous scanning of the 488- and the 568-nm lines. Optical sections were obtained in planes parallel to the cover slip (i.e., orthogonal to the anatomical surface of the sample) well below any cells damaged by the biopsy punch (Fig. 1). As described earlier [13,14] and as illustrated on Fig. 1, this orientation
42
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
Fig. 1. CLSM imaging of buccal mucosa biopsies. The tissue sample is positioned (without any processing or fixation) with the anatomical surface of the tissue perpendicular to the coverslip (not represented). Optical sections are obtained in planes perpendicular to the anatomical surface of the sample, well below any cells damaged by the biopsy punch. Such orientation of the tissue sample allows fluorescence imaging of all layers of the epithelium in the same optical plane.
of the tissue sample allows the imaging of all cell layers of the epithelium in the same optical plane (i.e., at the same depth below the coverslip). In preliminary experiments, the optimal imaging depth was determined using viable samples double-stained with EH-1 and CAM. In the area of the tissue immediately in contact with the cover slip (i.e., tissue that was sectioned by the biopsy punch), high levels of cell death could be observed on the EH-1 channel, and no calcein signal was detected. As the imaging plane was moved downward and viable cells were reached, the EH-1 signal decreased dramatically, and the calcein signal increased. As one proceeded further down, the calcein signal started decreasing, either because CAM might not have diffused that far, or simply because of attenuation from intervening cell layers. All our CLSM imaging was performed in a plane positioned about 10 mm below the plane of brightest calcein signal. This was
deep enough into the biopsy to ensure that no damage from the biopsy punch was visible; yet, it was close enough to the coverslip for good intensity and resolution. Typically, using the Nikon CF Fluor DL 403 / 1.3 lens, we imaged at a depth of about 40 mm, while with the Nikon CF Fluor 103 / 0.5 (which gave a much thicker optical section), imaging could performed at a depth of about 70 mm. EH-1 images were analyzed using the particle counting capability of NIH image V. 1.60 and the downloadable confocal macros. Briefly, EH-1 images were opened in NIH Image with a red density slice set so that all EH-1-stained nuclei appeared in red. Outlining, labeling and counting of the color-coded nuclei was then performed automatically using the Analyze Particles function of the Analyze menu. The area analyzed was also measured and results were expressed as number of EH-1 stained nuclei per mm 2 of tissue imaged.
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
43
3. Results and discussion
3.1. MTT assay A significant decrease in TR index was observed with dissected buccal mucosa stored in PBS at room temperature as early as 3 h post-mortem (Fig. 2, open symbols). The influence of temperature on viability was studied in a separate experiment, using tissue from a second animal (solid symbols). As shown on Fig. 2, TR indices at 378C were significantly lower than those at room temperature (218C628C) at all time points. Additionally, the close similarity between the two sets of data obtained at room temperature from two different animals illustrate the robustness and good reproducibility of the assay. Although PBS is still used by some groups as a flow-through buffer [15], most now use KBR-type buffers in an effort to sustain the viability of the membrane. We wanted to evaluate whether the viability of dermatomed mucosa was indeed improved by such buffer systems. Moreover, because KBR buffers contain glucose, and due to the high bacterial load of the buccal mucosa, the influence of sodium azide was also investigated. Fig. 3 shows that, at room temperature, the difference in viability
Fig. 2. MTT assay of dissected buccal mucosa: Influence of temperature. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in PBS at room temperature. Buccal mucosa was carefully dissected using surgical scissors, floated on PBS at room temperature and assayed at various time points post-mortem (Mean6s.e., n53–5).
Fig. 3. MTT assay of dermatomed buccal mucosa: Influence of medium composition at room temperature. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in KBR at room temperature. Buccal mucosa was dermatomed to 800 mm, stored in various media at room temperature, and assayed 3 h and 20 h post-mortem (Mean6s.e., n53).
after 2.5 h in PBS and in KBR is significant, although not as important as we expected. After 20 h, however, PBS yielded TR index 50% lower than those obtained from KBR. Additionally, no significant differences were found between PBS and KBR lacking glucose, or when comparing media with or without sodium azide. In their MTT assay with skin, Hood et al. used a 50% decrease from control as a limit for viable samples [11]. Control samples in KBR assayed within 1 h post-mortem reproducibly yielded TR indices around 0.05 (Figs. 3 and 4). Thus, by Hood’s criteria (TR indices above 0.025 are from viable samples), after 20 h in KBR (with or without sodium azide) at room temperature, buccal samples were still viable (TR indices of 0.035 and 0.038, respectively) whereas those in PBS or in KBR without glucose were not (TR indices of 0.020 and 0.021 respectively). Because most in vitro diffusion studies are conducted at 348C, a study similar to that in Fig. 3 was conducted at that temperature. TR indices for KBR samples were maintained at 85% of the control value after 6.3 h and at 72% after 9.3 h (Fig. 4a). Contrary to Fig. 3, however, striking differences in TR indices between KBR and PBS samples were visible at all time points. As early as 3.3 h post-mortem, TR
44
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
(Fig. 6). In PBS, there was a significant increase in nuclear staining with time for about 9 h post-mortem. In some cases, this increase was highly linear (Fig. 6, closed circles, R 2 50.90); in others, however, the trend was not as clearly marked, particularly at higher temperatures (Fig. 6, open squares). Nevertheless, in all cases, about 9 h post-mortem, the number of EH-1 nuclei in the KBR samples was significantly lower (|50%) than that in the PBS samples, which is remarkably consistent with the MTT assay. After 10 h in PBS, loss of tissue structure could be observed by CLSM, and translated into an apparent decrease of the number of EH-1 nuclei. Despite these encouraging results, however, the EH-1 signal acquisition and quantification was tedious and often provided noisy results at early time points. Therefore, we concentrated our efforts on the calcein results. CLSM images of the calcein fluorescence detected at the basal level of buccal mucosa samples stored at 348C in PBS or in KBR are shown in, respectively, Figs. 7 and 8. After about 9 h post-mortem, no calcein fluorescence could be elicited from the PBS samples as illustrated on Fig. 7d. In KBR samples, however, calcein fluorescence was detected more than 24 h post-mortem (Fig. 8d). Fig. 4. MTT assay of dermatomed buccal mucosa: Influence of medium composition at 348C. Tissue was excised from the animal immediately after sacrifice and transfered to the laboratory in KBR (a) or PBS (b) at room temperature. Buccal mucosa was dermatomed to 800 mm (a) or dissected (b), stored in various media at 348C, and assayed at various time points post-mortem (Mean6s.e., n53).
indices from PBS samples were down to 0.03 (i.e., 37% decrease from the KBR control). A 50% decrease was reached at 6.3 h. These results were confirmed in a separate experiment (Fig. 4b).
3.2. CLSM assay In the CLSM assay, EH-1 images of buccal mucosa kept at room temperature in PBS showed an increase in nuclear staining with time (Fig. 5). Using standard image analysis techniques, the number of EH-1 stained nuclei per unit surface area imaged was quantified and plotted as a function of time postmortem for tissue samples stored in PBS or KBR
3.3. Discussion Viability is a parameter that is gaining increased attention in studies where freshly excised membranes are used to mimic in vivo conditions, as in in vitro diffusion experiments. With skin, viability may not critically affect in vitro diffusion behavior, provided that the stratum corneum is the main barrier to diffusion, an hypothesis which holds true in many circumstances. However, in instances where metabolism is of interest, where the compound’s diffusion is epidermis-limited, and where non-keratinized epithelia (like buccal mucosa and cornea) are studied, viability should be carefully evaluated as it has the potential to directly affect results obtained in vitro. Numerous viability assays have been developed for a large array of cell lines and cytotoxic events, including 51 Cr release, 3 H-thymidine uptake [16], trypan blue exclusion, ATP and glucose utilization, tetrazolium salts colorimetric assays [16,17] and fluorescent vital staining. Only a limited number of these assays, however,
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
45
Fig. 5. CLSM imaging of EH-1 fluorescence in the upper layers of the buccal mucosa at 2.5 h (left) and 7.0 h (right) post-mortem. Scale bar represents 100 mm.
have been or can be adapted to the study of thick tissue sample viability. The tetrazolium salts assays [10,11,18–21] and the glucose utilization assays
Fig. 6. Quantitative assessment of viability by CLSM. Surface density of nuclei stained by EH-1 as a function of time postmortem at the end of staining. Buccal mucosa was dermatomed to 800 mm and stored in PBS at room temperature (d), PBS at 348C (h), or KBR at 348C (s). Each experiment was conducted separately using tissue from a different animal. One biopsy was imaged throughout the thickness of the epithelium at each time points, at 2–4 different locations. Straight lines indicate regression lines for PBS at room temperature (d, 2–9 h, R 2 50.90) and KBR at 348C (s, 2–11 h, R2 50.27).
[4,22,23] are two popular skin viability assays. Thymidine uptake [24] and trypan blue exclusion [25,26] were used recently with cornea. Ferrera et al. determined heart biopsies viability using the MTT bioassay [20]. Recently, Potter and coworkers [27] used nuclear fluorescence staining to estimate in vivo tissue injury following laser surgery. However, fluorescent viability probes, although widely used in cell cultures, have only found a limited number of application with thick samples [28]. Typically, buccal tissue viability has been assessed through a combination of ATP measurements [29] and histopathological examinations [30], complemented (or not) by linearity of transport data [30,31]. One could argue that the latter is a test of integrity rather than viability, as are electrophysiological measurements (transepithelial potential and resistance), which are also commonly used not only for buccal mucosa [32–34], but for other epithelia as well [33,35]. We have addressed this point in detail elsewhere [36,37]. Nevertheless, despite this awareness, only limited viability data are published in these studies. This is of concern with respect to the interpretation of flux data when most laboratories using porcine buccal mucosa get their tissue from slaughterhouses, and the time of death (often unknown or rarely reported) seldom coincides with the
46
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
Fig. 7. CLSM imaging of calcein fluorescence in the basal layers of dermatomed buccal mucosa stored in PBS at 348C, 2.8 h (a), 4.8 h (b), 6.8 h (c), and 8.8 h post-mortem. Scale bar represents 100 mm.
time of tissue excision. Furthermore, once excised, although it is a common practice to keep buccal tissue in KBR or PBS (ice-cold during transfer from the slaughterhouses, and at 348C during diffusion experiments), limited data are available to document the extent to which viability is actually maintained with these media [38]. In our laboratory, the MTT assay and CLSM vital staining with the EH-1 / CAM probe pair were used earlier to assess the viability of excised rabbit corneas. In this study, we adapted these two assays for the study of buccal mucosa viability. Contrary to the common belief that buccal mucosa viability is relatively stable for about 6 h [32], we have shown that TR activity in dissected mucosa decreases sharply in the first 10 h post-mortem at room temperature (Fig. 2). These experimental conditions are close to those encountered in the slaughterhouse between the time the animals are sacrificed
and the time the tissue is actually excised by the researchers. The decrease in viability is sharper as the temperature is increased to 348C (Figs. 3 and 4) and 378C (Fig. 2). We have not investigated the rate of TR decrease at 48C, however it is likely to be slower than at the above temperatures. This would constitute relevant information, however, for laboratories obtaining tissue from carcasses stored overnight in cold rooms [39]. Having established that the MTT assay can successfully discriminate between various experimental parameters (time and temperature) and shows a good inter-animal reproducibility, buccal mucosa viability was then determined under conditions closer to those used in in vitro experimental setups. With a few exceptions [15,40], most in vitro buccal studies are conducted using dermatomed mucosa with KBR rather than PBS as diffusion cell medium [41,33]. At room temperature, the difference between KBR and
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
47
Fig. 8. CLSM imaging of calcein fluorescence in the basal layers of dermatomed buccal mucosa stored in KBR at 348C, 2.7 h (a), 8.7 h (b), 10.7 h (c), and 25 h post-mortem. Scale bar represents 100 mm.
PBS was statistically significant but practically minimal (Fig. 3). At 348C, however, the temperature of choice for in vitro diffusion studies, differences between the two media were dramatically accentuated. As early as 3.3 h post-mortem, PBS samples yielded TR indices 60% lower than those from KBR samples (Fig. 4b), whereas buccal mucosa viability in KBR was maintained above 72% of its original value for more than 9 h post-mortem (Fig. 4a). The influence of sodium azide (NaN 3 ), an antimicrobial commonly used in in vitro skin studies, was also investigated because (1) NaN 3 is known to have adverse effects on skin viability and may therefore affect buccal tissue viability, (2) buccal mucosa has a high bacterial load, and (3) media for buccal work usually contain high levels of glucose, which promotes bacterial growth. Although it appeared to have a significant effect at some time points, the addition of 0.02% sodium azide to PBS or KBR did not
consistently affect viability compared to the original media. Glucose however, was shown to be indispensable to the maintenance of TR activity in the tissue (Fig. 3). These MTT results were remarkably consistent with both the EH-1 and calcein results obtained by CLSM. Obviously, as opposed to the MTT assay, CLSM imaging cannot be used as a routine laboratory assay for viability determination and screening. Furthermore, as we showed in this study, CLSM does not readily lend itself to obtaining quantitative data. Nevertheless, CAM staining reliably confirmed the results obtained through the MTT assay. Additionally, contrary to SEM / TEM, it allows the direct visualization of unfixed and viable samples, which, while not directly valuable in this particular study, could potentially be very informative in other instances [6]. For example, one can see how CAM staining could be conducted in parallel with the use
48
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
of a second probe (an enzyme maker or a diffusion probe excitable by the 568-nm or 647-nm line of the KrAr laser) to ensure tissue viability during confocal imaging of some other event of interest. Although it is beyond the scope of this paper to address this issue in details, it should be noted that in viable CAM-stained samples, we consistently observed that calcein fluorescence was much brighter at the basal level than anywhere else in the tissue. Consequently, our CLSM assay was based primarily on fluorescence signal from the basal level of the epithelium, which we found to be reliably consistent with the EH-1 signal and the results from the MTT assay. Although the localization of calcein fluorescence did not prevent us from drawing conclusions about the viability of our biopsy samples, it was sufficiently intriguing that we are currently trying to understand it in more details. Three hypotheses are currently being investigated in our laboratory to explain this phenomenon: (1) there is a preferential diffusion pathway for CAM from the connective tissue towards the epithelium or along the basal membrane, and the area of bright calcein fluorescence actually represents CAM’s diffusion front towards the surface of the epithelium; (2) CAM is able to diffuse throughout the epithelium, but is only cleaved to calcein in the lower cell layers of the epithelium, indicating higher esterase activity in this part of the tissue, or finally, (3) CAM is able to diffuse throughout the whole epithelium, it is cleaved by esterases inside all viable cells, but calcein fluorescence is quenched (pH or concentration effect) in the upper part of the tissue. A concentration-induced quenching effect is unlikely considering that staining is performed using a 50 mM CAM solution and that calcein self-quenching was reported to occur at concentrations above 100 mM. The existence of a pH gradient within the tissue would certainly be of interest to drug delivery laboratories like ours, however, it is currently difficult to demonstrate due to the lack of pH probes excitable by a Kr /Ar laser. We are currently focusing on quantifying the esterase activity in various levels of the tissue (hypothesis 2). Hot spots of enzyme activities have been observed earlier in skin [13]. Moreover, recently, Gherzi et al. reported high expression levels of a glucose transporter in the basal cells of oral mucosa [42].
Based on the results of this study, we are currently conducting our in vitro diffusion experiments at 348C with KBR for no longer than 8 h post-mortem. As mentioned briefly earlier, the question remains whether viability is directly linked to tissue integrity, i.e. whether diffusion pathways are altered as tissue viability is gradually lost. In experiments designed to measure permeability coefficients, this may be the only relevant issue, and already some groups have focused their efforts on demonstrating linearity of transport during the time frame of the experiment [30,43]. Studies conducted in our laboratory have shown that care should be taken in choosing the model permeant for such investigation [36,37]. Moreover, as detailed earlier, for experiments with broader goals [6], this information is clearly not sufficient.
4. Conclusions CLSM used in combination with CAM and EH-1 (used as ‘live’ and ‘dead’ probes respectively) is a powerful tool to assay the viability of unfixed, thick buccal mucosa samples submitted to various storage conditions. Although CLSM cannot realistically be the method of choice when viability is the only variable of interest, it becomes a powerful tool, however, when viability probes are used in combination with other reporting probes to monitor two events at once. The MTT assay, on the other hand, proved to be an easy, reliable, sensitive, and reproducible assay that can be routinely performed in most laboratories. To our knowledge, it is the first time that this assay is used for the determination of buccal tissue viability. It is a more convenient alternative than histology, electrophysiology, or ATP level determination, all of which were used earlier with this tissue. Recently, we have also shown it can be reliably used for toxicity screening purposes [6]. Using these two assays, we showed that buccal mucosa viability is significantly altered in PBS in the first 8 h following tissue collection. In KBR, viability was maintained well above 75% of the original value for about 6 h. Work is now under way in our laboratory to determine whether the loss of viability of buccal tissue in PBS is associated with alterations in in vitro permeation behavior, and vice-versa,
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
whether maintenance of viability in KBR is synonym with preservation of in vivo diffusion pathways [36,37].
[9] [10]
Acknowledgements The authors wish to thank Harolyn Hood (FDA, Washington DC) for sharing her MTT protocol, Ron Baireuther and Gabriela Fuentes from UCSF Experimental Surgery for providing pig tissue on a regular basis, and Willy Fong and Joanne Lee (UC Berkeley) for their help with some of the MTT assays. NIH Image is a public domain program developed at the U.S. National Institutes of Health and available on the Internet at http: / / rsb.info.nih.gov / nih-image /. This study was supported by NIH-DE11275 (CC and DI), Fondation Singer Polignac (DI), and an AACP New Investigator Grant (CC).
[11]
[12]
[13]
[14]
[15]
[16]
References ´ A.G. de Boer, J.A. Bouwstra, H.E. [1] J.C. Verhoef, H.E. Bodde, Junginger, F.W. Merkus, D.D. Breimer, Transport of peptide and protein drugs across biological membranes, Eur. J. Drug Metab. Pharmacokinet. 15(2) (1990) 83–93. [2] A.J. Hoogstraate, J.C. Verhoef, A. Pijpers, L.A.M.G. Van ´ In Leengoed, J.H.M. Verhijden, H.E. Junginger, H.E. Bodde, vivo buccal delivery of the peptide drug buserelin with glycodeoxycholate as an absorption enhancer in pigs, Pharm. Res. 13(8) (1996) 1233–1237. [3] M.J. Rathbone, B.K. Drummond, I.G. Tucker, The oral cavity as a site for systemic drug delivery, Adv. Drug Deliv. Rev. 13 (1994) 1–22. [4] S.W. Collier, N.M. Sheikh, A. Sakr, J.L. Lichtin, R.F. Stewart, R.L. Bronaugh, Maintenance of skin viability during in vitro percutaneous absorption / metabolism studies, Toxicol. Appl. Pharmacol. 99(3) (1989) 522–533. [5] P.W. Wertz, C.A. Squier, Cellular and molecular basis of barrier function in oral epithelium, Crit. Rev. Ther. Drug Carrier Syst. 8(3) (1991) 237–269. [6] D. Imbert, C. Cullander, Assessing penetration enhancers for buccal drug delivery: tools for simultaneous efficacy and toxicity screening using viable tissue samples. Proceed. Int’l. Symp. Control. Rel. Bioact. Mater., 25 (1998), 26–27. [7] P.L. Moore, I.C. Maccoubrey, R.P. Haugland, A rapid pH insensitive two color fluorescence viability cytotoxicity assay, J. Cell Biol. 111(5 part 2) (1990) 58A. [8] D. Imbert, C. Cullander, Assessment of cornea viability by
[17]
[18]
[19]
[20]
[21] [22]
[23]
49
confocal laser scanning microscopy and MTT assay, Cornea 16(6) (1997) 666–674. R. Supino, MTT assays, Methods Mol. Biol. 43 (1995) 137–149. F.B. Hershey, C.N.D. Cruickshank, L.I. Mullins, The quantitative reduction of 2,3,5-triphenyl tetrazolium chloride by skin in vitro, J. Histochem. Cytochem. 6 (1958) 191–196. H.L. Hood, M.G. Robl, R.L. Bronaugh, Skin absorption methodology: Improving the lipophilicity of the receptor fluid and the accuracy of the skin viability assay, Fundam. Appl. Toxicol. 30(1) (1996) 168. M.V. Berridge, A.S. Tan, K.D. McCoy, R. Wang, The biochemical and cellular basis of cell proliferation assays that use tetrazolium salts, Biochemica 4 (1996) 15–20. P. Boderke, H.P. Merkle, C. Cullander, M. Ponec, H.E. ´ Localization of aminopeptidase activity in freshly Bodde, excised human skin: direct visualization by confocal laser scanning microscopy, J. Invest. Dermatol. 108(1) (1997) 83–86. M.E.M.J. Meuwissen, J. Janssen, C. Cullander, H.E. Junginger, J. Bouwstra, A cross-section device to improve visualization of fluorescent probe penetration into the skin by confocal laser scanning microscopy, Pharm. Res. 15(2) (1998) 352–356. C.A. Squier, M. Kremer, P.W. Wertz, Continuous flow mucosal cells for measuring the in-vitro permeability of small tissue samples, J. Pharm. Sci. 86(1) (1997) 82–84. R.S. Gieni, Y. Li, K.T. HayGlass, Comparison of [ 3 H]thymidine incorporation with MTT- and MTS-based bioassays for human and murine IL-2 and IL-4 analysis. Tetrazolium assays provide markedly enhanced sensitivity, J. Immunol. Methods 187(1) (1995) 85–93. T. Mosmann, Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays, J. Immunol. Methods 65 (1983) 55–63. J.N. Kearney, L.A. Wheldon, G. Gowland, Cryopreservation of skin using a murine model: validation of a prognostic viability assay, Cryobiology 27(1) (1990) 24–30. R.F. Brown, A.R. Groves, Succinate dehydrogenase as an index of thermal injury to skin, Br. J. Exp. Pathol. 54(2) (1973) 117–122. R. Ferrera, A. Larese, F. Berthod, J. Guidollet, C. Rodriguez, G. Dureau, A. Dittmar, Quantitative reduction of MTT by hearts biopsies in vitro is an index of viability, J. Mol. Cell. Cardiol. 25(9) (1993) 1091–1099. R.F.R. Brown, J.V. Harvey Kemble, Tetrazolium reductase as an index of viability of stored skin, Burns 1 (1975) 179–185. M. Kietzmann, W. Loscher, D. Arens, P. Maass, D. Lubach, The isolated perfused bovine udder as an in vitro model of percutaneous drug absorption. Skin viability and percutaneous absorption of dexamethasone, benzoyl peroxide, and etofenamate, J. Pharmacol. Toxicol. Methods 30(2) (1993) 75–84. R.C. Wester, J. Christoffel, T. Hartway, N. Poblete, H.I. Maibach, J. Forsell, Human cadaver skin viability for in vitro percutaneous aborption: Storage and detrimental effects of heat-separation and freezing, Pharm. Res. 15(1) (1998) 82– 84.
50
D. Imbert, C. Cullander / Journal of Controlled Release 58 (1999) 39 – 50
[24] J.H. Lass, W.J. Reinhart, D.L. Skelnik, W.E. Bruner, R.P. Shockley, J.Y. Park, D.L. Hom, R.L. Lindstrom, An in vitro and clinical comparison of corneal storage with chondroitin sulfate corneal storage medium with and without dextran, Ophthalmology 97(1) (1990) 96–103. [25] M. Morton, Corneal storage at 4 degrees Celsius in a chondroitin sulphate and dextran containing medium, Aust. N.Z. J. Ophthalmol. 20(3) (1992) 211–213. [26] M. Canals-Imohr, J. Costa-Vila, D. Ruano-Gil, D. PitaSalorio, Comparative study of current corneal short-term, medium-, and long-term preservation methods, Transplant Proc. 27(4) (1995) 2418. [27] R.F. Potter, G. Peters, M. Carson, T. Forbes, C.G. Ellis, K.A. Harris, G. DeRose, W.G. Jamieson, Measurement of tissue viability using intravital microscopy and fluorescent nuclear dyes, J. Surg. Res. 59(5) (1995) 521–526. [28] C.A. Poole, N.H. Brookes, G.M. Clover, Keratocyte networks visualised in the living cornea using vital dyes, J. Cell Sci. 106(2) (1993) 685–691. [29] D. Harris, J.R. Robinson, Drug delivery via the mucous membranes of the oral cavity, J. Pharm. Sci 81(1) (1992) 1–10. [30] M.E. Dowty, K.E. Knuth, B.K. Irons, J.R. Robinson, Transport of thyrotropin releasing hormone in rabbit buccal mucosa in vitro, Pharm. Res. 9(9) (1992) 1113–1122. [31] M. Nair, Y.W. Chien, I. Buccal Delivery of Progestational Steroids, Characterization of barrier properties and effect of penetrant hydrophilicity, Int. J. Pharm. 89(1) (1993) 41–49. [32] C.R. Bland, S.S. Davis, D.A. Rawlins, Evaluation of the viability of an in vitro porcine buccal mucosa preparation, Proceed. Intern. Symp. Control. Rel. Bioact. Mater. 18 (1991) 499–500. [33] Y. Rojanasakul, L.Y. Wang, M. Bhat, D.D. Glover, C.J. Malanga, J.K. Ma, The transport barrier of epithelia: a comparative study on membrane permeability and charge selectivity in the rabbit, Pharm. Res. 9(8) (1992) 1029– 1034. ´ J.C. Verhoef, M. Ponec, W. [34] M.E. de Vries, H.E. Bodde,
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
Craane, H.E. Junginger, Localization of the permeability barrier inside porcine buccal mucosa: a combined in vitro study of drug permeability, electrical resistance and tissue morphology, Int. J. Pharm. 76(1-2) (1991) 25–35. C. Bjerre, B. Sjostrom, E. Bjork, O. Camber, Pig nasal tissue: a characterization and in vitro permeation study, Pharm. Res. 13(9) (1996) S–171. D. Imbert, C. Cullander, Permeability of buccal mucosa in vitro: viability versus integrity, J. Dent. Res. 77(Suppl. A) (1998) 290. D. Imbert, C. Cullander, Buccal mucosa in vitro experiments: II. Influence of viability on tissue permeability, J. Control. Rel., (1998) Manuscript in preparation, to be submitted. ´ Methods for assessing the A.J. Hoogstraate, H.E. Bodde, buccal mucosa as a route of drug delivery, Adv. Drug Del. Rev. 12 (1993) 99–125. V.A. Moore, I.W. Kellaway, P. Timmins, A. Dennis, Development and chraracterization of an ex vivo buccal mucosal model to assess solute permeability, Pharm. Res. 14(11) (1997) S–660. C.A. Lesch, C.A. Squier, A. Cruchley, D.M. Williams, P. Speight, The permeability of human oral mucosa and skin to water, J. Dent. Res. 68(9) (1989) 1345–1349. A.J. Hoogstraate, S. Senel, C. Cullander, J. Verhoef, H.E. ´ Effects of bile salts on transport rates Junginger, H.E. Bodde, and routes of FITC-labelled compounds across porcine buccal epithelium in vitro, J. Control. Rel. 40(3) (1996) 211–221. R. Gherzi, G. Melioli, M. De Luca, A. D’Agostino, M. Guastella, C.E. Traverso, F. D’Anna, A.T. Franzi, R. Cancedda, High expression levels of the ‘erythroid / brain’ type glucose transporter (GLUT1) in the basal cells of human eye conjunctiva and oral mucosa reconstituted in culture, Exp. Cell Res. 195(1) (1991) 230–236. B. Polentarutti, A. Sjoberg, A. Peterson, E.-K. Anderberg, A.-L. Ungell, Viability of rat intestinal segments in vitro, Proceed. Int. Symp. Control. Rel. Bioact. Mater. 24 (1997) 333–334.