Pepsin and Pancreatin Performance in the Dissolution of Crosslinked Gelatin Capsules from pH 1 to 8

Jean Gallery,* Jian-Hwa Han, and Chiramel Abraham

ABSTRACT The authors conducted a systematic study of the use of pepsin and pancreatin in dissolution of crosslinked gelatin capsules in order to understand the digestive performance ability of the enzymes across the pH range from 1 to 8. The work presented here includes a description of the experimental design and controls, followed by actual dissolution profiles obtained with pepsin from pH 1 to 6.8 and pancreatin from pH 4 to 8. The results generally agree with published activity curves for the enzymes and bring into question the appropriateness of the recommended pancreatin level set forth in USP á711ñ.

INTRODUCTION

Since the publication of the two-tier dissolution testing stimuli article in 1998 (1) and its subsequent adoption in USP á711ñ, there has been discussion in industry about the enzyme and activity recommendations based on the pH of the medium. Much of this discussion has been anecdotal, but published proteolytic activity curves for pepsin (1, 2) and pancreatin (trypsin and chymotrypsin components) (1, 2) confirm that peak pepsin activity occurs around pH 2 and peak pancreatin activity appears around pH 8. Unfortunately, the proteolytic activity of these enzymes in the pH range of 5–6 is significantly less than the peak activity.

The intention of this study was to characterize the proteolytic activity of pepsin and pancreatin through actual dissolution testing on crosslinked capsules across the pH range in which a dissolution laboratory typically operates.

EXPERIMENTAL DESIGN AND CONTROLS

Experimental Design

Tier 1 and Tier 2 (Tier 1 + enzyme) dissolution profiles of crosslinked gelatin capsules across the pH range of 1 to 8 have been collected according to the experimental plan outlined in Table 1.

Table 1. Experimental plan

	Pepsin				Pancreatin
Enzyme	No	Yes	No	Yes	No	Yes	No	Yes
pH	Control		Crosslinked		Control		Crosslinked
1	X	X	X	X
4	X	X	X	X	X	X	X	X
6	X	X	X	X	X	X	X	X
6.8	X	X	X	X	X	X	X	X
8					X	X	X	X

Dissolution profiles of six capsules were collected at each test condition. Each profile presented graphically in this article represents a mean of 6 capsules tested. Crosslinking of substrate capsules was achieved by treatment with formaldehyde vapor. The control capsules were untreated.

Dissolution was performed in medium containing aqueous hydrochloric acid or phosphate buffers to achieve the specified pH. The enzymes were added at the level recommended in USP á711ñ. That is, pepsin was added to the medium at not more than 750,000 units per liter, and pancreatin was added at not more than 1750 units of proteolytic activity per liter. Enzymes were freshly purchased from Sigma-Aldrich, and the labeled activities (from certificates of analysis) were used to determine the amount of enzyme to weigh for the media preparations to provide the maximum allowed activity levels. Each medium was prepared immediately before use to minimize potential autolysis and subsequent activity loss over time.

Substrate Capsules

The substrate capsules chosen for this study have a dissolution profile that is independent of pH across the range studied (pH 1–8). The chosen substrate capsules are soft gelatin capsules with a hydrophilic liquid fill (development product at Abbott Laboratories). The dissolution profiles in Figure 1 show that dissolution is rapid, and opening of the gelatin shell dictates the release profile. Capsule opening is faster at lower pH, but this is to be expected well below the isoelectric point (pI) of the gelatin.

Fig.1. Substrate capsule dissolution profiles (untreated, Tier 1).

In addition, the untreated substrate capsule dissolution profiles are not affected by the addition of pepsin or pancreatin, as shown in Figure 2.

Fig. 2. Substrate capsule dissolution profiles (untreated, Tier 2).

Induced Crosslinking

The goal of the induced crosslinking was to prepare capsules that would not meet Q=80% in 30 minutes under Tier 1 conditions but would meet the requirement under Tier 2 conditions (with enzyme). Crosslinking was induced by treatment with formaldehyde vapor in a manner similar to a process previously published (6). The capsules were placed on a tray inside a vacuum desiccator, and approximately 5 µL of formaldehyde was added to a separate dish inside the chamber. A vacuum was pulled briefly to increase the vapor pressure of the formaldehyde, and the flask was left closed for a length of time for the crosslinking reaction to occur.

Residual formaldehyde vapor will remain in the gelatin shell and further crosslink the capsules after they are removed from the reaction chamber. For this reason, it was decided to standardize the exposure time so that capsules could be prepared immediately before testing in the different media. Figure 3 depicts the exposure selection process. The profiles were generated in 0.1 N HCl medium with pepsin at ~750,000 units per liter.

Fig. 3. Aldehyde exposure study (0.1 N HCl Tier 1, Tier 2 with pepsin).

A 2-hour exposure time was insufficient to cause a Tier 1 failure. However, crosslinking induced by overnight exposure (15.5 hours) was too severe for pepsin to overcome at the USP activity level. A 4-hour exposure provided the desired failure and allowed complete recovery with the digestive enzyme. This aldehyde exposure time was used to crosslink capsules for the study outlined in Table 1 above. The Tier 1 and Tier 2 capsules for each test condition were prepared and tested on the same day for direct comparability. Figure 4 demonstrates the reproducibility of the 4-hour crosslinking process for each experiment in Table 1.

Fig. 4. Tier 1 capsule profiles after 4 hours of aldehyde exposure

Figure 4 does show some variability in the degree of crosslinking, which is to be expected. However, with the exception of the pH 1 curve, all other profiles show approximately 60% or less released at 30 minutes. Although the higher solubility of gelatin at pH 1 again is apparent, the profile does not meet Q of 80% at 30 minutes. There is no trend among the pH 4 to pH 8 Tier 1 profiles. One can observe variability at the same pH on different days.

Results and Discussion

The Tier 1 and Tier 2 dissolution profiles of crosslinked capsules were carried out as described in Table 1. The profiles obtained with pepsin are presented in Figure 5, and the profiles obtained with pancreatin are presented in Figure 6.

The data in Figure 5 confirm the wide pH range of published pepsin activity (2). The enzyme easily overcomes the crosslinking at pH 4, where the profile is equivalent to the uncrosslinked profile presented in Figure 1. The Tier 2 release in pH 1 is slightly reduced at 20 minutes compared to the uncrosslinked profile; however the Q value of 80% is easily met at 30 minutes. This agrees with the published activity curve (2) that shows a somewhat lower activity at pH 1 than at pH 4. It is clear that pepsin begins to suffer activity loss at pH 6. The chosen Q is met at 30 minutes, but the release at 20 minutes is only ~40%. This indicates a delay in capsule shell opening that could lead to a failure at 30 minutes for products that are not immediately solubilized. Pepsin is unable to overcome the crosslinking at pH 6.8, where the activity of the enzyme is greatly reduced.

The pancreatin data in Figure 6 are more surprising. The lack of response at pH 4 is expected based on the published activity curves (4, 5). However, the reduced ability to overcome crosslinking from pH 6 to 8 was not expected. The pancreatin activity at pH 8 was barely sufficient to bring the 30-minute release value above 80%, which is shown in part by the large standard deviation obtained (87±14%). The Tier 2 profile at pH 6.8 is somewhat better than at pH 8 insofar as all 6 capsules achieve complete release in 30 minutes. The activity at pH 6 is not sufficient to allow the Q of 80% to be met at 30 minutes; however it is very close (79%).

Pancreatin was not able to bring the profiles back to the uncrosslinked appearance at any of the pH values studied. Delayed capsule shell opening was observed at each pH. Considering these capsules were crosslinked to the same extent as those in the pepsin portion of the study, it appears that the pancreatin level allowed in General Chapter á711ñ is set too low. This statement is supported by a confirmation of the labeled activity of the lot of pancreatin used in the study. The value determined by the monograph test was 98% of the labeled amount of proteolytic activity from the vendor certificate of analysis.

Fig. 5. Demonstration of pepsin activity: Tier 1 and Tier 2 dissolution profiles of crosslinked capsules from pH 1 to 6.8.

Fig. 6. Demonstration of pancreatin activity: Tier 1 and Tier 2 dissolution profiles of crosslinked capsules from pH 4 to 8.

Further studies were performed to find the level of pancreatin that would bring the Tier 2 profiles back to the same appearance as the uncrosslinked capsule profile. Figure 7 shows the profiles obtained at pH 8 (expected peak activity) using the USP recommended pancreatin level of ~1750 activity units/L, along with profiles obtained with 2×, 3×, and 5× levels of pancreatin.

Fig. 7. Dissolution profiles at pH 8 with multiple levels of pancreatin.

The profiles in Figure 7 indicate that a three-fold increase in pancreatin is required to bring the profile back to an uncrosslinked appearance. There is improvement at the two-fold level; however the standard deviation is nearly 20% at the 20-minute time point. At the 3× level, the standard deviation at 20 minutes is ~2%.

CONCLUSION

The data presented in this study confirm that pepsin has a wide pH range of useful digestion activity in the dissolution of crosslinked capsules. The pH range of pancreatin activity also agrees with published values. However, the level allowed for dissolution testing according to USP General Chapter á711ñ is insufficient when compared to the response of pepsin on the same capsules. It has been determined that a three-fold increase in pancreatin is required to achieve the same level of digestion of crosslinked gelatin in the dissolution bath when compared to pepsin. The original study published by the Gelatin Capsule Working Group (1) showed that this level of pepsin was biologically relevant, but a rigorous study was not performed to match the activity of the enzymes. Now that this systematic study has been performed, USP is encouraged to consider formally revising the recommended level of pancreatin proteolytic activity in General Chapter á711ñ on Dissolution.

Significant delay in capsule shell opening can be very detrimental to products that are not immediately solubilized. If the contents require a certain time to dissolve after the shell is breached, then the profile could be delayed enough to cause unnecessary failure. Because many products require time to dissolve the drug inside of the capsule, matching the pancreatin level to the biorelevant pepsin level will remove an unnecessary burden for industry.

REFERENCES

Gelatin Capsule Working Group, “Collaborative Development of Two-Tiered Dissolution Testing for Gelatin Capsules and Gelatin-Coated Tablets Using Enzyme-Containing Media,” Pharmacopeial Forum 24(5): 7045–7050 1998.
D.W. Piper and B.H. Fenton, “pH Stability and Activity Curves of Pepsin with Special Reference to Their Clinical Importance,” Gut (6): 506–508 1965.
Methods in Enzymology, vol. XLIV, Academic Press, Klaus Mosbach, editor, p. 79 1976.
The Enzymes, vol. III, Academic Press, Klaus Mosbach, editor, p. 429 1971.
Methods in Enzymology, vol. XLIV, Academic Press, Klaus Mosbach, editor, pp. 403–404 1976
Thomas B. Gold, Robert G. Bruice, Jr., Robert A. Lodder, and George A. Digenis, “Determination of Extent of Formaldehyde-Induced Crosslinking in Hard Gelatin Capsules by Near-Infrared Spectrophotometry,” Pharmaceutical Research 14(8): 1046–1050 1997.

* Correspondence should be addressed to Jean Gallery, Dissolution Center, D-R4P4, Bldg. R1B, Abbott Laboratories, 1401 Sheridan Rd., North Chicago, IL 60064-6249, USA; telephone 847.938.4950; e-mail jean.gallery@abbott.com