**4. Results**

Images of the microcapsules in solution show that the capsules average size was 30 µm (Fig. 1A) and that the inside of the microcapsule is loaded with the dye/mineral oil mixture (Fig. 1B). About 80% of the mass of these microcapsules is the fluid mineral oil and about 20% of the mass consists of the gelatin shell.

Cross-Linked Gelatin Microcapsules for Drug Delivery in a Arthritic TMJ 261

the microcapsules appeared to degrade as a result of inducing an arthritic response (i.e., CFA injection), compare Fig. 4A to Fig. 4B and compare Fig. 4C to Fig. 4D. Time also reduced the number of microcapsules in the joint, compare 5 day and 10 day saline injected rats (compare Fig. 4A to Fig. 4C) and 5 day and 10 day CFA injected rats (compare Fig. 4B to Fig. 4D). A significant decrease in the number of intact beads was observed in the inflamed TMJ when comparing the 5 day time point to the 10 day time

Fig. 2. Microcapsule breakdown in vitro. Microcapsules (approximately 80,000) were placed in 200 µl of three different solutions at a temperature of 37°C and the degradation of the microcapsules was monitored by counting the number of intact microcapsules remaining after 1, 5 and 10 days. The solutions the microcapsules were placed in consisted of either 1) 0.9% saline or 2) 15 µg/ml of CFA/water homogenate or 3) a 50%/50% homogenate of 0.9% saline and 15 µg/ml CFA. At each time point a 10 µl aliquote was removed and the number of microcapsules was counted in triplicate in a hemacytometer chamber. Values are the

point (Fig. 4E).

mean ± SEM.

Fig. 1. Gelatin microcapsules containing canola oil mixed with Alexa 488 dye range in size between 15-30 micrometers. Microcapsules are loaded with canola oil (80% by weight) mixed with 1mM Alexa 488 dye. A) Bright field image of the microcapsules in solution. B) Fluorescent image of the same microcapsules before injection into the TMJ. Bar equals 100 µM.

To initially assess the degradation characteristics of the gelatin shell the microcapsules were placed in 24 well tissue culture plates containing saline and CFA solutions. Over a period of 10 days the beads did significantly degrade in a mixture of isotonic saline and 15 µg of CFA (Fig. 2).

Injection of dye loaded microcapsules (30 µl) into the upper joint space, a region between the articular fossa and condyle (Fig. 3A) produced an area of beads 2-3 mm in diameter (dotted line, Fig. 3B) in the upper joint space of the rat TMJ. Cells (e.g. neutrophils) were adjacent to the microcapsules 5 hours after injection (blue nuclei, Fig. 3C-E). In the TMJ

Fig. 1. Gelatin microcapsules containing canola oil mixed with Alexa 488 dye range in size between 15-30 micrometers. Microcapsules are loaded with canola oil (80% by weight) mixed

Fluorescent image of the same microcapsules before injection into the TMJ. Bar equals 100 µM. To initially assess the degradation characteristics of the gelatin shell the microcapsules were placed in 24 well tissue culture plates containing saline and CFA solutions. Over a period of 10 days the beads did significantly degrade in a mixture of isotonic saline and 15

Injection of dye loaded microcapsules (30 µl) into the upper joint space, a region between the articular fossa and condyle (Fig. 3A) produced an area of beads 2-3 mm in diameter (dotted line, Fig. 3B) in the upper joint space of the rat TMJ. Cells (e.g. neutrophils) were adjacent to the microcapsules 5 hours after injection (blue nuclei, Fig. 3C-E). In the TMJ

with 1mM Alexa 488 dye. A) Bright field image of the microcapsules in solution. B)

µg of CFA (Fig. 2).

the microcapsules appeared to degrade as a result of inducing an arthritic response (i.e., CFA injection), compare Fig. 4A to Fig. 4B and compare Fig. 4C to Fig. 4D. Time also reduced the number of microcapsules in the joint, compare 5 day and 10 day saline injected rats (compare Fig. 4A to Fig. 4C) and 5 day and 10 day CFA injected rats (compare Fig. 4B to Fig. 4D). A significant decrease in the number of intact beads was observed in the inflamed TMJ when comparing the 5 day time point to the 10 day time point (Fig. 4E).

Fig. 2. Microcapsule breakdown in vitro. Microcapsules (approximately 80,000) were placed in 200 µl of three different solutions at a temperature of 37°C and the degradation of the microcapsules was monitored by counting the number of intact microcapsules remaining after 1, 5 and 10 days. The solutions the microcapsules were placed in consisted of either 1) 0.9% saline or 2) 15 µg/ml of CFA/water homogenate or 3) a 50%/50% homogenate of 0.9% saline and 15 µg/ml CFA. At each time point a 10 µl aliquote was removed and the number of microcapsules was counted in triplicate in a hemacytometer chamber. Values are the mean ± SEM.

Cross-Linked Gelatin Microcapsules for Drug Delivery in a Arthritic TMJ 263

Fig. 4. Microcapsule breakdown in a non-arthritic and arthritic TMJ. Saline or 15 µg of CFA in 15 µl was injected into the TMJ followed by a second injection 24 hours later of 30 µl of dye loaded microcapsules. Microcapsules in the upper joint space/synovial tissue were imaged in sagittal sections of the TMJ tissue 5 and 10 days after injection of microcapsules. Images were captured 5 days after injecting the TMJ with microcapsules and saline (A) or CFA (B). Microcapsules in the upper joint space/synovial tissue was imaged 10 days after injection of microcapsules in a joint previously injected with saline (C) or CFA (D). The histogram below indicates the number of microcapsules counted per joint in rats sacrificed 0, 5 and 10 days after injection. There were 5 animals per group. Asterisk indicates P<0.05.

Fig. 3. Injection of gelatin microcapsules containing canola oil with Alexa 488 dye were injected into the upper joint space of a rat temporomandibular joint (TMJ). A) Sagittal view of a rat skull, arrows point to the TMJ condyle and articular fossa. In this experiment the TMJ upper joint space was injected with 30 µl of dye loaded microcapsules (90 mg microcapsules [dry weight] in 1 ml of Tris buffer or 0.9% saline and then the rats were sacrificed 5 hours later. B) Saggittal section of a rat TMJ 5 hours after injection of microcapsules into the upper joint space (arrow) stained with hemotoxylin and eosin. Dotted line in the upper joint space outlines the region containing the injected microcapsules. C) High magnification image of boxed region in panel B. D) Fluorescent image of Alexa 488 loaded microcapsules 5 hours after injection into the TMJ. Green is the Alexa 488 loaded microcapsules and blue is the DAPI stained cell nuclei. E) High magnification of image from panel D. Bar in panel B equals 1 mm and bar in panel D equals 100 µm.

Fig. 3. Injection of gelatin microcapsules containing canola oil with Alexa 488 dye were injected into the upper joint space of a rat temporomandibular joint (TMJ). A) Sagittal view of a rat skull, arrows point to the TMJ condyle and articular fossa. In this experiment the TMJ upper joint space was injected with 30 µl of dye loaded microcapsules (90 mg microcapsules [dry weight] in 1 ml of Tris buffer or 0.9% saline and then the rats were sacrificed 5 hours later. B) Saggittal section of a rat TMJ 5 hours after injection of microcapsules into the upper joint space (arrow) stained with hemotoxylin and eosin.

microcapsules. C) High magnification image of boxed region in panel B. D) Fluorescent image of Alexa 488 loaded microcapsules 5 hours after injection into the TMJ. Green is the

magnification of image from panel D. Bar in panel B equals 1 mm and bar in panel D equals

Dotted line in the upper joint space outlines the region containing the injected

Alexa 488 loaded microcapsules and blue is the DAPI stained cell nuclei. E) High

100 µm.

Fig. 4. Microcapsule breakdown in a non-arthritic and arthritic TMJ. Saline or 15 µg of CFA in 15 µl was injected into the TMJ followed by a second injection 24 hours later of 30 µl of dye loaded microcapsules. Microcapsules in the upper joint space/synovial tissue were imaged in sagittal sections of the TMJ tissue 5 and 10 days after injection of microcapsules. Images were captured 5 days after injecting the TMJ with microcapsules and saline (A) or CFA (B). Microcapsules in the upper joint space/synovial tissue was imaged 10 days after injection of microcapsules in a joint previously injected with saline (C) or CFA (D). The histogram below indicates the number of microcapsules counted per joint in rats sacrificed 0, 5 and 10 days after injection. There were 5 animals per group. Asterisk indicates P<0.05.

Cross-Linked Gelatin Microcapsules for Drug Delivery in a Arthritic TMJ 265

Fig. 6. Effect of metalloprotease inhibitor 2 and 9 on microcapsule breakdown in vivo Rats were injected with saline or CFA or saline with an MMP inhibitor or CFA with an MMP inhibitor added and then 24 hours later these groups of rats were injected with 30 µl microcapsules or microcapsules plus MMP inhibitor. All injections were into the superior joint space of the TMJ. The inhibitor was a 1 µM solution of the MMP-2/9 inhibitor IV. The histogram shows the total number of beads per joint after 0, 5 and 10 days after bead

Fig. 7. Nociceptive response after injecting microcapsules along with metalloprotease inhibitors Rats were injected into the upper joint space of the TMJ with 30 µl saline or 15 µg CFA (CFA inj), some in combination with a MMP-2/9 inhibitor. One day after this injection a second 30 µl injection of microcapsules or microcapsules containing MMP-2/9 inhibitor was performed. Before and after injection the meal duration was recorded. No inj= before animal were injected. There were 8 animals per group. \*=P<0.05, \*\*=P<0.01, \*\*\*=P<0.001

when comparing CFA injected rats to the saline injected rats.

injection. There were 3-5 animals per group. \*=P<0.05.

We next asked the question, would injection of microcapsules into the TMJ induce a pain response in the rat with a non-arthritic or an arthritic joint. To test the response to microcapsule injection we measured a behavioral correlate of nociception (meal duration) before and after injecting microcapsules into the TMJ of a rat previously injected with saline (non-arthritis) or CFA (arthritic). Microcapsule injection did not significantly increase the pain response in either the non-arthritic or arthritic joint (Fig. 5).

Fig. 5. Nociceptive response after injecting microcapsules into a non-arthritic and arthritic TMJ. Rats were injected with 30 µl 0.9% saline or 15 µg CFA (CFA inj) and then 24 hours later the rats were injected with either saline or 30 µl microcapsules in 0.9% saline. After injection of the microcapsules the meal duration was recorded for 10 days (1 through 10). There were 8 animals per group. \*=P<0.05, \*\*=P<0.01 when comparing the CFA + saline group to the saline + saline group and when comparing the CFA + microcapsule group to the saline + microcapsule group.

To test that the breakdown of beads in vivo may be due the presence of MMP-2 and MMP-9 we injected the microcapsules with an MMP-2/9 inhibitor. The number of microcapsules decreased over time in vivo but addition of a MMP-2/9 inhibitor had no significant effect on microcapsules degradation after injection into a rat TMJ (Fig. 6). The inhibition of the MMP-2/9 also did not affect the nociceptive response after injecting the TMJ with microcapsules (Fig. 7).

We next asked the question, would injection of microcapsules into the TMJ induce a pain response in the rat with a non-arthritic or an arthritic joint. To test the response to microcapsule injection we measured a behavioral correlate of nociception (meal duration) before and after injecting microcapsules into the TMJ of a rat previously injected with saline (non-arthritis) or CFA (arthritic). Microcapsule injection did not significantly increase the

Fig. 5. Nociceptive response after injecting microcapsules into a non-arthritic and arthritic TMJ. Rats were injected with 30 µl 0.9% saline or 15 µg CFA (CFA inj) and then 24 hours later the rats were injected with either saline or 30 µl microcapsules in 0.9% saline. After injection of the microcapsules the meal duration was recorded for 10 days (1 through 10). There were 8 animals per group. \*=P<0.05, \*\*=P<0.01 when comparing the CFA + saline group to the saline + saline group and when comparing the CFA + microcapsule group to

To test that the breakdown of beads in vivo may be due the presence of MMP-2 and MMP-9 we injected the microcapsules with an MMP-2/9 inhibitor. The number of microcapsules decreased over time in vivo but addition of a MMP-2/9 inhibitor had no significant effect on microcapsules degradation after injection into a rat TMJ (Fig. 6). The inhibition of the MMP-2/9 also did not affect the nociceptive response after injecting the

the saline + microcapsule group.

TMJ with microcapsules (Fig. 7).

pain response in either the non-arthritic or arthritic joint (Fig. 5).

Fig. 6. Effect of metalloprotease inhibitor 2 and 9 on microcapsule breakdown in vivo Rats were injected with saline or CFA or saline with an MMP inhibitor or CFA with an MMP inhibitor added and then 24 hours later these groups of rats were injected with 30 µl microcapsules or microcapsules plus MMP inhibitor. All injections were into the superior joint space of the TMJ. The inhibitor was a 1 µM solution of the MMP-2/9 inhibitor IV. The histogram shows the total number of beads per joint after 0, 5 and 10 days after bead injection. There were 3-5 animals per group. \*=P<0.05.

Fig. 7. Nociceptive response after injecting microcapsules along with metalloprotease inhibitors Rats were injected into the upper joint space of the TMJ with 30 µl saline or 15 µg CFA (CFA inj), some in combination with a MMP-2/9 inhibitor. One day after this injection a second 30 µl injection of microcapsules or microcapsules containing MMP-2/9 inhibitor was performed. Before and after injection the meal duration was recorded. No inj= before animal were injected. There were 8 animals per group. \*=P<0.05, \*\*=P<0.01, \*\*\*=P<0.001 when comparing CFA injected rats to the saline injected rats.

Cross-Linked Gelatin Microcapsules for Drug Delivery in a Arthritic TMJ 267

articularly injected gelatin microcapsules on the nociceptive response as well as on the inflammatory response. Loading the gelatin microcapsules with the ibuprofen or morphine was accomplished because we hypothesized that intra-articular injection of drug loaded beads would reduce the nociceptive response. We determined that gelatin microcapsules do not increase IL-1β levels in the disc and synovial tissues after injection into the TMJ. Nor do the capsules increase the nociceptive response in an arthritic joint. Importantly, injection of morphine but not ibuprofen loaded microcapsules decreased the nociceptive response of a rat with an arthritic TMJ. Because injection of microcapsules loaded with morphine decreased the nociceptive response of a rat with inflammatory arthritis we expect that the

The premise of our animal (i.e., meal duration) model is that the CFA induced TMJ pain would affect the rat, such that, when a hungry animal initiated a meal the animal would eat slowly due to the TMJ pain associated with the movement of the mandible during the chewing process. This is exactly what we observed following bilateral TMJ CFA injections in males and females, i.e., the rats had longer meal durations (Harper, Kerins, Talwar, Spears, Hutchins, Carlson, McIntosh & Bellinger, 2000; Kerins, Carlson, Hinton, Grogan, Marr, Kramer, Spears & Bellinger, 2005; Kerins, Carlson, McIntosh & Bellinger, 2003). Support that meal duration is a measure for orofacial pain it that when ibuprofen is administered directly into a inflamed, CFA-injected TMJ meal duration was normal in both male and female rats (Kerins, Carlson, McIntosh & Bellinger, 2003). Our selection of CFA was made because it produces a persistent arthritic pain response that last for over two weeks (Hill, Bellinger, Spears, Hutchins, Kerins & Kramer, 2007; Ren, 1999). Thus, of all the agents, CFA was best when trying to establish long duration pain. Injection of CFA into the TMJ significantly lengthened meal duration in rats, while the same amount of CFA in the knee did not affect meal duration (Kerins, Carlson, Hinton, Grogan, Marr, Kramer, Spears & Bellinger, 2005) indicating meal duration is a specific measure of orofacial pain. Interestingly IL-1β remained significantly elevated in the TMJ of the ibuprofen treated animals injected with CFA (Kerins, Carlson, McIntosh & Bellinger, 2003), suggesting that some inflammation from the CFA injection remained. In another study, cyclooxgenase-II (COX-2) inhibitors normalized meal duration in rats after CFA injection (Kerins, Carlson, McIntosh & Bellinger, 2004). In this study the COX-2 inhibitor also attenuated the inflammation, i.e., TMJ tissue IL-1 normalized (Kerins, Carlson, McIntosh & Bellinger, 2004). In still another study, rats were given capsaicin or vehicle at 2 and 10 days of age; capsaicin permanently destroyed afferent nociceptive fibers in these animals (Bellinger, Spears, King, Dahm, Hutchins, Kerins & Kramer, 2007). When these male rats reached adulthood saline or CFA was injected into the TMJ and their meal duration was measured. Capsaicin treatment alone had no effect on meal duration, because saline injected, non-capsaicin treated rats had the same meal duration as saline injected, capsaicin treated rats. Non-capsaicin treated rats injected with CFA had longer meal durations than rats that were pre-treated with capsaicin, which demonstrated meal duration after CFA injection was normalized due to a capsaicin-induced loss of afferent nociceptive neuronal fibers (Bellinger, Spears, King, Dahm, Hutchins, Kerins & Kramer, 2007). The lack of change in meal duration in these capsaicin treated male rats occurred despite CFA inducing greater TMJ swelling, which demonstrated that the physical and mechanical changes in the inflamed TMJ synovial joint did not affect meal duration measurements. Another rationale for suggesting that meal duration is a measure of

microcapsules will amealorate the pain response in patients with TMJ disease.

**5.1 TMJ intra-articular model for testing microcapsules** 

Ibuprofen drug loaded microcapsules did not significantly affect the nociceptive response (Fig. 8A) or the immune response (Fig. 8B) of CFA injected rats. Morphine loaded microcapsules did significantly decrease the nociceptive response (p<0.05, days 4 and 5) and post-hoc testing showed a significant decrease on the fourth day post-CFA injection (Fig. 8A).

Fig. 8. Nociceptive and inflammatory response after injecting microcapsules containing ibuprofen or morphine. Rats TMJs were injected with a 30 µl solution of microcapsules (bead inj). The microcapsules contained oil (Blank microcapsules) or oil with 15% ibuprofen (Ibuprofen microcapsules) or oil containing 1% morphine (Morphine microcapsules). 24 hours after microcapsule injection the rats were subdivided further for TMJ injection of either saline or 15 µg CFA (CFA inj). A) Before and 8 days after injection of the saline/CFA the daily meal duration was recorded. There were 8 animals in each of the four treatment groups. B) The amount of IL-1β in the TMJ retrodiscal, synovial and disc tissue 7 days after injection with microcapsules. There were 3-4 animals in each of the four treatment groups for the cytokine analysis.
