THE UNIVERSITY LIBRARY
PROTECTION OF AUTHOR ’S COPYRIGHT
This copy has been supplied by the Library of the University of Otago on the understanding that the following conditions will be observed:
1. To comply with s56 of the Copyright Act 1994 [NZ], this thesis copy must only be used for the purposes of research or private study.
2. The author's permission must be obtained before any material in the thesis is reproduced, unless such reproduction falls within the fair dealing guidelines of the Copyright Act 1994. Due acknowledgement must be made to the author in any citation.
3. No further copies may be made without the permission of the Librarian of the University of Otago.
August 2010 Oral Delivery of Peptides to Salmon
Greg F. Walker
A thesis submitted for the degree of Doctor of Philosophy of the University of Otago, Dunedin, New Zealand
August 1999 11
Abstract
Purpose. Oral delivery of peptide and protein drugs to fish has potential advantages to the aquaculture industry. However, the bioavailability of proteins and peptides from the intestinal tract is very low. This can be attributed inpart to the extensive lumenal and mucosa! proteolytic activities of the intestine. In this study the metabolic activities of the Quinnat salmon (Oncorhynchus tshawytscha) intestinal tract towards bovine serum albumin (BSA), human (hLHRH) and salmon (sLHRH) luteinizing-hormone releasing hormones were investigated. To overcome this metabolic barrier a number of proteolytic inhibitors were investigated to stabilise the protein and peptides drugs in the intestine, improving their opportunity for absorption across the salmon intestine.
Methods. The extent of BSA degradation was determined by the formation of acid soluble peptides from BSA, measured by the method of Lowry. The metabolic stability hLHRH and sLHRH were determined using a capillary electrophoresis (CE) assay. Binding of proteins and proteolytic substrates to the polyacrylic acid, specifically
Carbopol ® 934P (CP934P), was studied by centrifugal filtration. Gel filtration and reverse phase HPLC was used to determine the stability of trypsin in the presence of CP934P.
Results. The lower intestinal tract of the salmon is divided into three anatomically distinct sections: anterior, middle and posterior. The lumenal proteolytic activities of the posterior intestinal section towards BSA were approximately half that of the anterior and middle sections. The half-lives of the LHRH analogues in the posterior intestinal sections were 2-fold longer than for the anterior and middle sections. Intestinal lumenal proteolytic activities towards BSA and the LHRH analogues correlated with higher activities towards the synthetic substrates of chymotrypsin, trypsin and elastase. LHRH analogues were shown to be rapidly hydrolysed by bovine a-chymotrypsin, whereas bovine trypsin and porcine elastase hydrolysed the LHRH analogues slowly by companson. 111
Proteolytic activity of the posterior mucosal homogenates towards BSA was 4-fold higher than for the middle mucosal homogenates. LHRH analogues were hydrolysed by the posterior mucosal homogenate, whereas in the middle mucosal homogenate they were stable. The higher metabolic instability of LHRH analogues in .the poste1ior mucosal homogenate correlated with higher chymotryptic-lik:e activity.
Soybean trypsin inhibitor (SBTI) was the most effective inhibitor, stabilising the LHRH analogues against the lumenal proteolytic activity of the intestine and increasing their half-lives in the anterior, middle and posterior section between 7.5..: and 15-fold. CP934P
(0.35% w/v, pH 8.0 at 15°C) increased the half-lives of the LHRH analogues in the anterior, middle and posterior sections between 1. 7- and 2. 7-fold. The rates of degradation in the presence of CP934P followed pseudo-zero order kinetics, indicating that the mechanism of inhibition was not autolysis.
CP934P (0.35% w/v, pH 7 at 37°C) inhibited the rate of hydrolysis of N-a-benzyol-L arginine ethyl ester (BAEE) and hLHRH by trypsin to 34% and 28% of the control activity, respectively. CP934P completely stabilised hLHRH in the presence of bovine a chymotrypsin. Binding studies showed 68% of the trypsin and 22% of chymotrypsin;
10% ofBAEE was bound to CP934P, but no LHRH was bound. Lower specific activities of the unbound CP934P treated proteases indicate that inactivation of the proteases is also occurring. No low molecular weight autolysis products of trypsin could be identified by gel filtration. Reverse phase HPLC analysis of the unbound CP934P treated trypsin suggests a number of conformational forms of trypsin. The equilibrium binding capacity was 30 mg of trypsin to lg of CP934P. IV
Conclusions. Chymotryptic-like activity of the salmon intestinal tract is a major metabolic barrier to the oral absorption of the LHRH analogues. The chymotryptic-like activities of the intestinal lumen and the intestinal mucosa are most likely due to the pancreatic secreted chymotrypsin and the intestinal brush-border membrane protease endopeptidase 24.18, respectively. The co-administration of SBTI with LHRH analogues in oral formulations could improve the oral absorption ofLHRH analogues in salmon. The inhibition of lumenal proteolytic activity by CP934P is the result of enzyme-polymer interaction, thereby reducing the free concentration of protease and in part denaturing the protease. The posterior intestine of salmon is the most metabolically favourable site for the oral delivery of peptides and proteins; however the LHRH analogues will have to overcome the higher endopeptidase 24.18 activity present in the posterior mucosa. V
Acknowledgments
My sincere thanks to my supervisors, Prof. Ian Tucker and Dr Robin Ledger for their guidence, help and advice throughout the course of study leading to this thesis. I am especially grateful to my friend and colleague Paul O Donnell (instrumentation tech.) who frequently fixed the capillary electrophoresis machine while others ran. I would also like to thank Dr Nigel Davies for his advice, support and friendship during my time at the School of Pharmacy. I acknowledge all the past and present postgraduate students of the School of Pharmacy, in particular fellow salmon researcher Leo Schep for his support and cooperation.
I am grateful for the financial assistance given by the Otago University Targeted Research Scholarship.
I also take the opportunity to thank my friends and flatmates, who over the many years have helped me through with many good times and memories. Without them I would have been cast against a padded cell.
Finally, I would like to acknowledge the support from my family, in particular my parents who have always encouraged and supported me in what ever I have done. Vl
Publications arising from this thesis
Refereed journals Walker, G.F., Ledger, R. & Tucker, LG. (1999). Carbomer inhibits tryptic proteolysis of luteinizing hormone-releasing hormone and benzoyl-L-arginine ethyl ester by binding the enzyme. Pharmaceutical Research 16, 1074-1080.
Refereed conference abstracts Walker, G.F., Ledger, R. & Tucker, LG. (1996) Proteolytic activity in the lower gastrointestinal tract of salmon. The Australian Journal of Hospital Pharmacy 26, P50.
Walker, G.F., Ledger, R. & Tucker, LG. (1997). Distribution of proteolytic enzymes in the lumen of the lower intestinal tract of salmon. Pharmaceutical Research 14S, 1445.
Walker, G.F., Ledger, R. & Tucker, LG. (1998). Susceptibility of human and salmon luteinizing hormone-realeasing hormone (LHRH) to metabolism in the salmon intestinal lumen and mucosal homogenates and their protection by various protease inhibitors. Proceedings ofthe Australian Society of Clinical Experimental Pharmacologists and Toxicologists 5, 1-57. Vll
Table of contents
Title page 1 Abstract 11 Acknowledgements V Publications arising from this thesis Vl Table of contents Vll List of tables Xl List of figures XIV List of abbreviations XXll
Chapter 1 : General Introduction 1 ?' 1.1 Preface 2 1.2 Intestinal barriers 3 1.2.1 Biochemical barriers 3 1.2.2 Transport barrier 5 1.2.3 P-glycoprotein efflux pump 8 1.3 Strategies to improve peptide and protein delivery 9 1.3.1 Protease inhibitors 9 1.3 .1.1 Serine protease inhibitors 9 1.3 .1.2 Low molecular weight peptidase inhibitors 10 1.3.1.3 Bile salts 11 1.3.1.4 Chelating agents 12 1.3.1.5 Mucoadhesive polymers 13 1.3.2 Permeation enhancers 15 1.3.3 Chemical modification 18 1.3.4 Formulation 20 1.4 Oral delivery of peptides and proteins to teleosts 23 1.4. l Digestive tract of gastric teleosts 23 1.4.2 Biochemical barriers of the intestinal tract of teleosts 27 1 A.2.1 Lumenal proteases 27 1.4.2.2 Intestinal membrane bound proteases 31 1.4.2.3 Intracellular digestion of proteins and peptides 31 1.4.2.4 Biochemical barriers to the immunological system 32 1.4.3 Absorption of proteins and peptides 34 1.4.3 .1 Peptide carrier mediated transport 34 1.4.3.2 Paracellular route 34 1.4.3 .3 Transcytosis 34 1.4.3.4 P-glycoprotein efflux pump in teleosts 36 1.4.4 Oral peptide and protein delivery to teleosts 36 1.5 Aims of the study 40
Chapter 2: Capillary Electrophoretic Assay ofLuteinizing Hormone-Releasing Hormones 41 2.1 Introduction 42 2.2 Materials 48 vm
2.3 Methods 48 2.3 .1 Instrumentation and standard conditions 48 2.3.2 Capillaries 49 2.3 .3 Buffers 49 2.3.4 LHRH digests 50 2.3.5 Silanisation · 50 2.3.6 hLHRH standards-in the presence of a-chymotrypsin 50 2.3. 7 hLHRH standards- variability studies 51
2.4 Results 52 2.4.1 Precision 52 2.4.2 Optimisation of sample stacking 56
~ 2.4.3 Optimisation of run buffer pH 63 2.4.4 Effect of chymotrypsin on the reproducibility 66 2.4.5 Storage stability 67 2.4.6 CE analysis ofLHRH 67 2.5 Discussion 68
Chapter 3: The Metabolic Barriers 73 3 .1 Introduction 74 3 .2 Materials 75 3.3 Methods 76 3 .3 .1 General 76 3.3.1.1 Fish 76 3.3.1.2 Collection and preparation of lumenal homogenates 76 3.3.1.3 In situ pH measurements 77 3.3.1.4 Preparation ofmucosal homogenates from frozen intestine 77 3.3.1.5 Preparation ofmucosal homogenates from fresh intestine 77 3.3.1.6 Protein estimations 78 3 .3 .1. 7 Measurement of surface tension 78 3.3.1.8 Statistical analysis 79 3 .3 .2 Proteolytic assays 79 3.3.2.1 General proteolytic activity 79 3.3.2.2 LHRH analogues 79 I· 3.3.2.3 Assays usingp-nitroanilides 80 3.3.2.4 Chymotryptic activity 80 3.3.2.5 Carboxypeptidase A and B activity 81 3.4 Results 82 3 .4.1 Lumenal proteolytic activity 82 3 .4.1.1 Intestinal lumen ·pH 82 3 .4.1.2 Optimum storage conditions for lumenal extracts 82 3.4.1.3 Effect of feeding on lumenal protein content 84 .3 .4.1.4 Effect of feeding on proteolytic activity 85 3 .4.1.5 LHRH digestion by pancreatic enzymes 87 3.4.1.6 LHRH digestion by lumenal enzymes 94 3 .4.1. 7 Specific proteolytic activity 97 IX
3.4.2 Mucosal proteolytic activity 100 3.4.2.1 Collection and storage ofmucosal homogenates 100 3.4.2.2 Stability ofLHRH in mucosal homogenates 101 3.4.2.3 General proteolytic activity 103 3 .4.2.4 Distribution of proteolytic enzymes 104 3.4.3 Inhibition of proteolytic activity 105 3.4.3.1 Inhibition of proteolysis by bile salts 105 3.4.3.2 Inhibition of proteolysis by chelators 107 3 .4.3 .3 Effect of soybean trypsin inhibitor 108 3.4.3.4 Inhibition oflumenal LHRH proteolysis 109 3.4.3.5 Inhibition of mucosal proteolytic activity 113 3.5 Discussion 115 3 .5 .1 Lumenal proteolytic activity 115 3.5.2 Mucosal proteolytic activity 121
Chapter 4: Polyacrylic Acid Polymer Inhibition Studies 125 4.1 Introduction 125 4.2 Materials 128 4.3 Methods 128 4.3.1 Analytical methods 128 4.3.1.1 Protein estimatfons 128 4.3.1.2 BSA hydrolytic activity 129 4.3.1.3 BAEE hydrolytic activity 129 4.3.1.4 BTEE hydrolytic activity 129 4.3 .1.5 Capillary electrophoresis 13 0 4.3.1.6 HPLC ofLHRH 130 4.3.1.7 Gel electrophoresis 130 4.3.1.8 Gel filtration 130 4.3.1.9 HPLC of trypsin 131 4.3.2 Inhibition studies 131 4.3.3 Binding studies 132 4.3.3.1 Centrifugal filtration 132 4.3.2.2 Langmuir isotherm 132 4.3.2.3 Effect of ionic strength and pH on adsorption of trypsin to CP934P 133 4.3.2.4 Adsorption of proteins to CP934P 133 4.4 Results 135 4.4.1 Inhibition of trypsin 135 4.4.1.1 Inhibition studies 135 4.4.1.2 Binding studies 139 4.4.1.3 Gel electrophoresis 141 4.4.1.4 Gel filtration 143 4.4.1.5 HPLC of trypsin 146 4.4.1.6 Langmuir isotherm 148 4.4.1. 7 Effect of ionic strength and pH 148 4.4.1.8 Protein binding 150 X
4.4.2 Inhibition of a-chymotrypsin activity 150 4.4.2.1 Inhibition studies 150 4.4.2.2 Binding studies 153 4.5 Discussion 154 4.5.1 Inhibition studies 154 4.5.2 Binding studies 161
Chapter 5: Summary and Future Perspectives 165 5 .1 General discussion 166 5 .2 Future perspectives 173 '> References 175
,, Xl
List of tables Table 1.1 Proteases secreted into the lumen. Specificity is indicated by filled circle. 3
Table 1.2 Membrane bound peptidases. Specificity is indicated by filled circle. 4
Table 1.3 Specificity of low molecular weight peptidase inhibitors. 10
Table 1.4 Formulation designed to improve the oral delivery of peptides. 20
Table 2.1 Reproducibility of peak migration time (min), peak height (mV) /, and area (mV) for hLHRH and sLHRH. 52
Table 2.2 Effect of light mineral oil on reproducibility and percentage change in peak height (mV) and area (mV) for hLHRH and sLHRH. 54
Table 2.3 Effect of silanisation of sample vials on the reproducibility of and percental change in peak height and area for hLHRH and sLHRH. 5 5
Table 2.4 Effect of electrophoresis under constant voltage and constant current on reproducibility of CE assay for hLHRH. 56
Table 2.5 Effect of run buffer pH on reproducibility of migration time (MT), and peak height (PH) of hLHRH and sLHRH. 66
Table 2.6 Effect of chymotrypsin on reproducibility of peak height ofhLHRH. 66
Table 2.7 Effect of storage on the stability ofhLHRH in the presence of chymotrypsin. 67
Table 3.1 pH of the lumenal contents of salmon. 82
Table 3.2 General proteolytic activity of intestinal lumenal homogenates stored at 4°C for 3h, -80°C after being stored on ice and-80°C after snap frozen in liquid nitrogen (mean± s.d., n=3). 84 XU
Table 3.3 Two-way analysis of variance of the protein content of the anterior, middle and posterior intestinal sections after 12 and 24 h without food release. 85
Table 3 .4 Effect of feeding on specific activities and total activities of the anterior, middle and posterior intestinal sections. 85
Table 3.5 Two-way analysis of variance of specific proteolytic activity of the anterior, middle and posterior intestinal sections after 12 and 24 h without food release. 86
Table 3.6 Two-way analysis of variance of total activity of the anterior, middle and posterior intestinal sections after 12 and 24 h without food release. 86
Table 3. 7 Frequency· of gut section activities of individual salmon indicating a possible rank ordera 87
Table 3.8 Half-lifes (min) ofhLHRH and sLHRH in the anterior, middle and posterior lumenal homogenates. 97
Table 3.9 Two-way analysis of variance of half-lives ofhLHRH and sLHRH in the anterior, middle and posterior lumenal homogenates. 97
Table 3.10 Specific activities of the middle and posterior mucosal homogenates towards the synthetic substrates for trypsin, chymotrypsin, elastase and leucine aminopeptidase. 104
Table 3.11 Effect of sodium deoxycholate on proteolytic activities of the intestinal lumenal contents. 106
Table 3 .12 Percentage inhibition of general proteolytic activity of lumenal homogenates in the presence of 2 mM and 5 mM sodium deoxycholate. 107
Table 3 .13 Percentage inhibition of general proteolytic activity of lumenal proteolytic activity in the presence of CP934P, citric acid and EDTAa 108
Table 3 .14 Percentage inhibition of lumenal elastase proteolytic activity by soybean trypsin inhibitor at 1:50 and 1:100 intestinal extract:soybean trypsin inhibitor. 108
Table 4.1 Sedimentation of CP934P 0.35 % (w/v) in 50 mM sodium phosphate buffer pH 7 .0. Data are absorbances of samples taken from the upper region of the dispersion. 140 xm
Table 4.2 The protein conc_entration, activity and specific activity of trypsin solutions (50 µg/ml) filtered after incubation with pH 7 buffer alone (control) and after incubation with CP934P (mean± s.d., n = 3). 141
Table 4.3 Concentration of protein solutions (300 µg/ml) filtered after incubation with pH 7 buffer alone (control) and after incubation with CP934P at 37°C for 30 min in phosphate buffer pH 7.0 (mean± s.d., n = 3). 150
Table 4.4 Protein concentration, activity and specific activity of chymotrypsin (50 µg/ml) solutions filtered after incubation with pH 7 buffer alone (control) and after incubation with CP934P (mean± s.d., n = 3). 153
Table 4.5 CP934P binding, molecular weight and pl of proteins and peptides interacting with CP934P. 163 XIV
List of figures Figure 1.1 The putative mechanisms of uptake of proteins and peptides by intestinal epithelium. (i) carrier mediated, (ii) receptor mediated or transcytosis, (iii) passive diffusion, (iv) paracellular diffusion through the intercellular spaces. 5
Figure 1.2 Schematic representation of the salmonid gastrointestinal tract 24
Figure 1.3 Transverse section (200 X) of the wall of the posterior intestine of a ,' salmon ( Oncorhynchus tshawytscha). Mounted sections were stained with haemtoxylin and eosin: (A) epithelium, (B) goblet, (C) lamina propria , (D) stratum compactum, (E) stratum granulosum, (F) muscularis circularis, (G) muscularis longitudinalis and (H) tella subserosa (with permission from Phill Veillete). 26
,, Figure 2.1 Schematic of the capillary electrophoresis apparatus. 43
Figure 2.2 Schematic of electro-osmotic flow 44
Figure 2.3 Capillary electrophoretic separation ofhLHRH (A) and sLHRH (B) (O. lmg/ml in deionised water) in 50 mM sodium phosphate buffer pH 2.5 (Section 2.3 .1 ). 53
Figure 2.4 Capillary electrophoretic separation of the chymotryptic hLHRH digest inl O mM sodium phosphate buffer pH 2.5 containing sodium chloride (Section 2.3.1). Digests were diluted 2.5-fold (A), 4.0-fold (B) and 6-fold (C) in deionised water. 57 xv
Figure 2.5 Capillary electrophoretic separation of the chymotryptic sLHRH digest inlO mM sodium phosphate buffer pH 2.5 containing sodium chloride (Section 2.3.1). Digests were diluted 2.5-fold (A), 4.0-fold (B) and 6-fold (C) in deionised water. 58
Figure 2.6 Effect of diluting hLHRH (A) and sLHRH (B) chymotryptic digest in deionised water on the peak heights (triangle) and peak area (diamonds) when electrophoresed (Section 2.3 .1) in 10 mM sodium phosphate buffer at pH 2.5 containing 50 mM sodium chloride (open symbols) and 75 mM sodium chloride (closed circles). Each point represents the mean of three experiments. 59
Figure 2. 7 Effect of diluting hLHRH (closed symbols) and sLHRH (open symbols) chymotryptic digests in deionised water on sample stacking (peak height/peak area) when electrophoresed (Section 2.3.1) in 10 mM sodium phosphate buffer at pH 2.5 containing 75 mM sodium chloride (squares) and 50 mM sodium chloride (triangles). Each point represents the mean of three experiments 60
Figure 2.8 Effect of the ionic strength of the run buffer on the migration time of hLHRH (open circles) and sLHRH (closed circles) when electrophoresed in 10 mM sodium phosphate buffer pH 2.5 containing 50, 75 or 100 mM sodium chloride (Section 2.3.1 ). Each point represents the mean of three experiments. 61
Figure 2.9 Effect of diluting the digests in deionised water on RSD (%) for migration time (open bars) and peak height (closed bars) when electrophoresed (Section 2.3.1) in 10 mM sodium phosphate buffer pH 2.5 containing 50 mM sodium chloride (A) and 75 mM sodium chloride (B). 62 XVI
Figure 2.10 Capillary electrophoretic separation of the chymotryptic hLHRH digest at pH 2.5 (A), 4.0 (B) and 6.5 (C) containing 75 mM sodium chloride (Section 2.3.1). Arrows indicate hLHRH. Chymotrypsin digests analysed were not identical. 64
Figure 2.11 Capillary electrophoretic separation of the chymotryptic sLHRH digest at pH 2.5 (A), 4.0 (B) and 6.5 (C) containing 75 mM sodium chloride (Section 2.3.1). Arrows indicate sLHRH. Chymotrypsin digests analysed were •, not identical. 65
Figure 3.1 Time-course of the formation of acid-soluble peptides from BSA when incubated with the anterior (A), middle (B) and posterior (C) lumenal homogenates at pH 8.0, 15°C (Section 3.3.2.1). 83
Figure 3.2 Capillary electrophoretic separation of the chymotryptic hLHRH digest (Section 3.3.2.2). Digests were analysed after O (A), 2 (B) and 10 (C) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 89
Figure 3.3 Capillary electrophoretic separation of the chymotryptic sLHRH digest (Section 3.3.2.2). Digests were analysed after O (A), 2 (B) and 10 (C) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 90
Figure 3.4 Capillary electrophoretic separation of the elastase digest ofhLHRH (Section 3.3.2.2). Digests were analysed after O (A) and 60 (B) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 91
Figure 3.5 Capillary electrophoretic separation of the elastase digest of sLHRH (Section 3.3.2.2). Digests were analysed after O (A) and 60 (B) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 92 xvn
Figure 3.6 Capillary electrophoretic separation of the trypsin hLHRH (A) and sLHRH (B) digest (Section 3.3.2.2). Digests were analysed after 60 min incubation at pH 8.0, 15°C (Section 3.4.1.5). 93
Figure 3.7 Capillary electrophoretic separation ofhLHRH (A) and hLHRH lumenal digests by the anterior (B), middle (C) and posterior (D) intestinal section (Section 3.3.2.2). Samples were analysed 120 min after incubation at pH 8.0, 15°C (Section 3.4.1.6). 95
Figure 3.8 Capillary electrophoretic separation of sLHRH (A) and sLHRH ,, lumenal digests by the anterior (B), middle (C) and posterior (D) intestinal section (Section 3.3.2.2). Samples were analysed 120 min after incubation at pH 8.0, 15°C (Section 3.4.1.6). 96
Figure 3.9 Specific activity of trypsin (A), chymotrypsin (B), elastase (C) and carboxypeptidase A (D) and B (E) and leucine aminopeptidase (F) in the anterior, middle and posterior intestinal sections. The intestinal lumenal homogenates (10 µg of protein) were assayed for specific activity at pH 8.0 at 15°C (Sections 3.3.2.3 - 3.3.2.5). 99
Figure 3 .10 Relative distribution of general proteolytic activity, trypsin, chymotrypsin, elastase, carboxypeptidase A and B and leucine aminopeptidase activity of the lower intestinal tract. General proteolytic activity was measured using BSA as the substrate (Section 3.3.2.1) remaining proteolytic activities were determined using synthetic substrates (Sections 3.3.2.3-3.3.2.5). Anterior (•), middle (D) and posterior (•) intestinal sections. 100
Figure 3.11 Capillary electrophoretic separation ofhLHRH (A), and hLHRH posterior mucosal digest (B), sLHRH (C) and sLHRH posterior mucosal digest (D) (Section 3.4.2.3). Samples were analysed 180 min after incubation at pH 7.4 at 15°C (Section 3.4.1.6). 102 xviii
Figure 3.12 Time-course of the formation of acid-soluble peptides from BSA after incubation with the middle (open circles) and posterior ( closed circles) mucosal homogenates (Section 3.4.2.3). 103
Figure 3 .13 Change in surface tension with log concentration of sodium ·\ deoxycholate in 50 mM sodium phosphate buffer pH 8.0 (Section 3.3.1.7). 105
Figure 3.14 Half-lives ofhLHRH in anterior (A), middle (B) and posterior (C) Lumenal homogenates in the absence (control) and presence of EDTA, SBTI, DOC and CP934P. Half-life represents the mean time (min± s.d., n = 3) for the LHRH concentration to drop to half its initial concentration. N.S. not significantly different from control,(*) p < 0.01, (**) p < 0.005, (***) p < 0.0025, (****) p < 0.001 (*****) p < 0.0005, compared to control by Student's t test 110
Figure 3.15 Half-lives of sLHRH in anterior (A), middle (B) and posterior (C) Lumenal homogenates in the absence (control) and presence ofEDTA, SBTI, ' I DOC and CP934P. Half-life represents the mean time (min± s.d., n = 3) for the LHRH concentration to drop to half its initial concentration. N.S. not significantly different from control,(*) p < 0.01, (**) p < 0.005, (***) p < 0.0025, (****) p < 0.001, compared to control by Student's t test. 111
Figure 3 .16 Time-course of hLHRH proteolysis when incubated with the anterior lumenal homogenate in the presence ( open circles) and absence of CP934P (0.35% w/v) (closed circles) at pH 8.0 at 15°C (Section 3.4.3.4). Each symbol represents the peak height (mean± s.d., n=3). 112 XIX
Figure 3.17 Half-lives ofhLHRH (A) and sLHRH (B) in posterior mucosal Homogenates in the absence (control) and presence ofEDTA. DOC, bestatin and SBTI. Half-life represents the mean time for the concentration ofLHRHs to fall to half its original concentration (min± s.d., n = 3). N.S. not significantly different from control, (**) p < 0.005 compared to control by Student's t test. 114
Figure 3 .18 Potential cleavage sites of human and salmon LHRH by trypsin (T), chymotrypsin (C) and elastase (E). 117
Figure 3 .19 Schematic representation of enzyme-substrate complex. The active subsites (S) of the enzyme are located on both sides of the catalytic site (C) of the substrate (P). 118
Figure 4.1 Time-course of the formation of acid-soluble peptides from BSA after incubation with bovine trypsin in the presence (closed circles) and in the absence (open circles) of CP934P (0.35% w/v) at pH 7.0 (Section 4.3.1.3). Each point represents the mean± s.d. of three experiments. 136
Figure 4.2 Formation of BA from BABE (20 mM) by trypsin (10 µg/ml) in the presence ofCP934P (0.35% w/v) (closed circles) and in the absence of CP934P ( open circles) at pH 7.0 (Section 4.3.2). Each point represents the mean± s.d. of three experiments. 136 ,,
Figure 4.3 Formation of the principal metabolite ofhLHRH following incubation of hLHRH (0.4 mM) with trypsin (10 µg/ml) in the presence of CP934P (0.35% w/v) (closed circles) and in the absence of CP934P (open circles) at pH 7.0 (Section 4.3.2). Each point represents the mean± s.d. of three experiments. 136 xx
Figure 4.4 Capillary electrophoretic separation of incubations ofhLHRH with trypsin (10 µg/ml): in the presence of CP934P (0.35% w/v) at pH 7.0 immediately after addition of trypsin (A), after 24 h incubation in the presence of CP934P (B) and after 24 h incubation in the absence of CP934P (C) at 3 7°C (Section 4.4.1.1). 138
Figure 4.5 Absorbance spectrum of CP934P 0.005% (w/v) in 50 mM sodium phosphate buffer pH 7.0 (Section 4.4.1.2). 139
:- Figure 4.6 SDS-PAGE of trypsin filtrates after silver staining. Trypsin filtrates after incubation in the presence of CP934P (0.35% w/v) (lane 1) in the
,, absence of CP934P (lane 2) in 50 mM sodium phosphate buffer pH 7.0.
\ Migration and molecular weights of protein standards indicated on figure 142
Figure 4.7 Gel filtration of trypsin (50 µg/ml) stored on ice (A) and after incubation at 60°C for 45 min (B) in 50 mM sodium phosphate buffer pH 7.0 (Section 4.3.1.8). 143
Figure 4.8 Gel filtration ofcentrifugal filtrates of trypsin (50 µg/ml) in the absence of CP934P (0.35% w/v) (A) and in the presence of CP934P (0.35% w/v) (B) and of CP934P 0.35% (w/v) in the absence of trypsin (C) in sodium phosphate buffer pH 7.0 after incubation at 37°C for 30 min (Section 4.3.1.8). 145
Figure 4.9 Reverse-phase HPLC of centrifugal filtrates of trypsin (50 ~Lg/ml) maintained at 4°C (A), trypsin incubated at 37°C in the absence of CP934P (0.35% w/v) (B), and incubated in the presence of CP934P (0.35% w/v) at 37°C (C) in 50 mM sodium phosphate buffer pH 7.0 for 30 min (Section 4.4.1.5). 147 XXl
Figure 4.10 Langmuir absorption isotherm of trypsin onto CP934P (0.35% w/v) incubated in 50 mM sodium phosphate buffer pH 7.0 at 37°C for 30 min (Section 4.3.3.2). Each point represents the mean of three experiments. 148
Figure 4.11 Effect of sodium chloride concentration on binding of trypsin (300 µg/ml) to CP934P (0.35% w/v) at pH 7 .0 (Section 4.4.1.6). Each point represents the mean± s.d. of three experiments. 149
~
Figure 4.12 Effect of pH on binding of trypsin (300 µg/ml) to CP934P (0.35% w/v) at pH 6.0, 7.0 and 8.0 (Section 4.4.1.6). Each point represents the mean ± s.d. of three experiments. 149
Figure 4.13 Time-course of the hydrolysis ofhLHRH by chymotrypsin in the presence (closed circles) and absence (open circles) of CP934P (0.35% w/v) at pH 7.0 (Section 4.4.2.1). 151
Figure 4.14 Reverse phase chromatograms ofhLHRH bovine chymotrypsin \ \ digests in the presence of CP934P (0.35% w/v) at pH 7 after 0.5 min (A) and 2.0 min (B) incubation at 37°C and after 2 min incubation in the absence of CP934P at 37°C (Section 4.4.2.1). 152
Figure 4.15 Proposed kinetic mechanism for protein adsorption and denaturation (Soderduist and Watson, 1980). 158
Figure 4.16 Proposed routes of the inhibition of proteolytic activity in the presence of
;, polyacrylic acid. A straight arrow denotes a change in state, a dashed arrow denotes proteolytic attack (autolysis). 159 xxii
List of abbreviations
ANOVA Analysis of variance ATP adenosine 5' -triphosphate AU absorbance units BABE N-a-benzyol-L-arginine ethyl ester BAPNA N a-benzoyl-DL-arginine p-nitroanilide BSA bovine serum albumin BTEE N-a-benzyol-L-tyrosine ethyl ester CE capillary electrophoresis CMC critical micelle concentration CP934P Carbopol 934P® DMSO dimethylsulfoxide DOC deoxycholate EDTA ethylenediaminetetra-acetic acid EOF electro-osmotic flow FITC fluorescein isothiocyanate GI gastrointestinal tract hLHRH human luteinizing hormone releasing hormone Hipp-L-Arg hippuryl-L-phenyl-arginine Hipp-L-Phe hippuryl-L-phenyl-phenylalanine HPLC high performance liquid chromatography HRP horseradish peroxidase I.D. inner diameter LH luteinizing hormone MW molecular weight N.S. not significantly different at the 5% level O.D. outer diameter pl isoelectric point PMSF phenylmethylsulphonyl fluoride xxm
psi pounds per square inch RSD relative standard deviation SBTI soybean trypsin inhibitor s.d. standard deviation SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis sLHRH salmon luteinizing hormone releasing hormone TCA trichloroacetic acid TEER trans epithelium electrical resistance TLCK N-tosyl-lysine chloromethyl ketone TPCK N-tosyl-phenylanine chloromethyl ketone
'., TRH thyrotropin-releasing hormone
)I
\ ,\
'{
•)
> 1
Chapter 1: General Introduction
·} 2
1.1 Preface
The Food and Agriculture Organisation of the United Nations (1997) believes that the increases in world wide demand for fish will be met by the aquaculture industry. It has been predicated that by the year 2020 the aquaculture industry will be bigger than the worlds entire fisheries (Holmes, 1996). However the farming of fish in high concentrations is associated with a number of problems, including poor reproduction and disease. With an increased demand for aquaculture there is an associated increase in the demand for better methods of peptide and protein of drug delivery to control these problems. Much of the research has focused on the delivery of the reproductive hormone luteinizing hormone-releasing hormone (LHRH) and its analogues (Van Der Kraak et al., 1983, Peter et al., 1988, Zohar et al., 1989), growth hormone and various vaccines (Austin, 1984).
Typically these compounds are delivered to fish by injection, implant or dip. However these treatments require removal and frequent handling (intensive manipulation) of fish ,,\ which can create sites for bacterial infection and cause stress to fish resulting in loss of appetite and poor reproductivity. One strategy to overcome these problems is to incorporate the drug into the fish diet. However to achieve sufficient absorption of peptide or protein drugs across the intestinal tract a number of barriers have to be overcome. This Introduction reviews the barriers to the oral delivery of peptides and proteins, in particular those of the teleost fish (fish with bony skeletons) and strategies used to improve their oral absorption across the intestinal tract. 3
1.2 Intestinal barriers
1.2.1 Biochemical barriers
Firstly, the metabolic barrier of the intestinal lumen plays an important role in the inactivation of the peptide drugs before reaching the systemic circulation. The lumen contains proteases secreted from the pancreas and peptidases sloughed off the mucosal membrane. The pancreatic proteases include the serine endopeptidases trypsin, chymotrypsin and elastase and also the exopeptidases, carboxypeptidases A and B. Specificities of these proteases are shown in Table 1.1 (Woodley, 1994). The lumenal concentrations of the pancreatic enzymes decrease towards the colon (Layer et al., 1990). It is believed that low amounts of pancreatic enzymes in the colon are due to their adsorption to the wall of the small intestine and their inactivation during transit, possibly by autolysis (Bahe et al., 1986). The proteolytic activity of the colonic contents towards a number of proteins is considerably lower than the ileum (Gibson et al., 1989). The major proteases present within the colon are believed to be of bacterial origin (Macfarlane et al., 1988, Gibson et al., 1989). Proteases can be secreted by bacteria or released from the cytosol after lysis of the bacterium. Microbial proteases have not been well characterised although serine, cysteine and metalloproteases have been identified in the colonic .contents of humans (Macfarlane et al., 1988, Gibson et al., 1989).
'I Table 1.1 Proteases secreted into the lumen. Specificity is indicated by filled circle.
Trypsin H 2NO-O--o-e±-o-ocooH Arg, Lys
Chymotrypsin H 2NO-~o-e±-o-ocooH Phe, Tyr
Elastase . H 2NO-O--o-e±-o-ocooH Ala, Gly, Ile, Leu, Val
Carboxypeptidase A H 2N0-0-0-0-0-,i,ecooH Tyr, Phe, Ile, Thr, Glu, His, Ala
Carboxypeptidase B H 2NO-O--O-o-olecooH Lys, Arg 4
Brush-border membranes of the enterocytes contain a number of peptidases (Figure 1.1). These peptidases form the second biochemical barrier to peptide absorption. At least twelve exo- and endopeptidases have been identified so far in the brush-border membranes of mammals (Table 1.2) (Woodley, 1994). The activities of these brush border enzymes have been shown to differ along the intestinal tracts of rat, rabbit and human (Bai et al., 1995a). ·
Table 1.2 Membrane bound peptidases. Specificity indicated by filled circle ,,
:, Aminopeptidase N H2N.±-o-0-o-0-ocooH Many .,
Aminopeptidase A H2NJ:-o-o-o-0-0cooH Asp, Glu ,I, Aminopeptidase P H2Ne-O-O-O-O-OcooH Pro
Aminopeptidase W H2N.±-o-0-0-0-0cooH Trp, Tyr, Phe
Dipeptidyl peptidase IV H 2NO-e!--0-o-0-0cooH Pro, Ala
y-glutamyl transpeptidase H2N.Lo-o-o-0-0cooH y-glutamic acid \ Peptidylpeptidase A H2NO-O-O-oi...... cooH His-Leu
Carboxypeptidase M H2N0-0-0-o-olecooH Lys, Arg
Carboxypeptidase P H2N0-0-0-0-olecooH Pro, Gly, Ala ,, y-glutamyl carboxypeptidase H2N0-0-0-0-ol.cooH y-glutamic acid I,
Endopeptidase-24.11 H2No-o--+e-o-0-0cooH Hydrophobic amino acids :}' Endopeptidase-24.18 HzNO-oJ.-!:-0-0-0cooH Aromatic amino acids .\
,\
), If a peptide is absorbed into the enterocyte cell it can come in contact with cytosolic peptidases. The cytosol of the mammalian enterocytes contains di- and tri- peptidases (Bai and Amidon, 1992) as well as proteases capable of hydrolysing larger peptides (Bai, 1994a and 1995). Peptides that enter the enterocyte by endocytosis can be released into the cytosol (Walker, 1981) or enter the lysosomes which 5
contain a large number of peptidases (Barrett and Heath, 1977, Walker, 1981, Marcon Genty et al., 1989). The lysosomal peptidases are known as cathepsins; so far eight exopeptidases and nine endopeptidases have been identified (Bohley and Seglen, 1992).
1.2.2 Transport barrier Peptides can be absorbed across the epithelial cell surface by a number of routes or
.\ mechanisms. In general peptides have poor intrinsic permeability across biological membranes due to their size, charge, conformation, hydrogen bonding capacity and hydrophilic nature (Burton et al., 1991). There are four possible routes or mechanisms for the passage of peptides across intestinal mucosa; these are shown schematically in Figure 1.1. The first three routes are considered transcellular transport since the molecule travels through the epithelial cell and the fourth route utilizes the paracellular pathway.
11 111 IV Domain
Brush-border
Tight junctions bI I Ii I t 11 Ill 0 t 0 Basolateral \
·,. Figure 1.1 The putative mechanisms of uptake of proteins and peptides by intestinal
:, epithelium. (i) carrier mediated, (ii) receptor mediated or transcytosis, (iii) passive diffusion, (iv) paracellular diffusion through the intercellular spaces. 6
i) Membrane carrier systems have been identified which facilitate the absorption of di (Cheesman and Parsons, 1974), tri- (Sleisenger et al., 1979) and tetra-peptides (Smithson and Gray, 1977) in the mammalian small intestine. A number of studies have shown that
+ peptide transport is Na -dependent and occurs against the concentration gradient, while a variety of other studies have shown the transport to be Na+ -independent (Cheeseman and Parsons, 1974, Ganapathy et al., 1981). Dipeptides and tripeptides were shown to be
-~ transported by a proton gradient-energised transport system in brush-border membrane vesicles (BBMV) isolated from the rabbit small intestine (Ganapathy and Leibach, 1983).
ii) Endocytotic uptake can be divided into two processes, pinocytosis, which involves the ingestion of fluid and solutes including proteins and peptides via small vesicles(< 150 nm in diameter) and phagocytosis, which involves the ingestion oflarge particles(> 250 nm in diameter).
Specific transepithelial mechanisms which facilitate the uptake of a number of proteins including growth factors and immunoglobulins have been identified in the immature intestine of humans (Chu and Walker, 1991, Koldovsky et al., 1991).
In the mature intestine, uitrastructure studies of horseradish peroxidase (HRP) provided direct evidence of non-specific transport of proteins through the vesicular compartments ,, of the enterocyte (Cornell et al., 1971). It is believed that the macromolecules can enter the cell in a non-specific manner by two mechanisms: by binding to the apical membrane followed by endocytosis or by being in solution close to the invaginating membrane \ which engulfs the protein in the developing vesicle. Gonnella and Neutra (1984) observed that proteins that enter the enterocyte when bound to the apical surface bypass the lysosomal system whereas proteins that enter from solution are exposed to the lysosomal enzymes within the enterocyte.
Specialised epithelial cells known as "M" cells have been shown to transport proteins such as ferratin (Boclanan and Cooper, 1973) and HRP (Owen, 1977) across the 7
intestinal epithelium. The proteins are incorporated into membrane compartments, transferred across the cell, and extruded from the serosal surface into the interstitium containing lymphoid cells (Owen, 1977).
iii) Passive diffusion of a peptide will depend on its ability to be incorporated into and diffuse through the lipid bilayer matrix of the membrane. Hydrophobicity of the molecule has been shown to be a good predictor of passive diffusion across cell membranes (Smith et al., 1975, Leo et al., 1977, Artursson, 1990, Artursson and Karlsson, 1991). Banks and Kastin (1985) showed that the flux of peptides across the blood-brain barrier can be related to its partitioning behavior between n-octanol and water, which is a measure of hydrophobicity. However, in vitro studies using human intestinal Caco-2 cell monolayers showed that the desolvation energy (hydrogen bond potential) required to break the hydrogen bonds in the lumenal fluid limits the entry of the peptide into the membrane (Conradi et al., 1991). This relationship was extended to in vivo studies which also showed that the absorption was inversely correlated to the number of calculated hydrogen bonding sites for model peptides (Karls et al., 1991).
iv) The paracellular route involves diffusion through the extracellular space between adjacent enterocytes. Tight junctions are the rate-limiting barriers that restrict flow through the paracellular pathway (Powell, 1981 ). The calculated pore radius of the :, intestinal tight junction is 5 nm (Pappenheimer, 1987, Pappenheimer and Reiss, 1987). Results with hamster intestine suggest that polypeptides of 11 amino acids long (MW 1900) can cross via the tight junctions but for larger macromolecules such as HRP (MW 40 000) this transport route is not possible under the same physiological conditions (Atisook and Madara, 1991). This route has two main advantages over the transcellular route; the peptide avoids biochemical barriers of the enterocytes and the absorption of the peptide is not limited by its desolvation energy (Conradi et al., 1991). 8
1.2.3 P-glycoprotein efflux pump
Drugs absorbed from the intestine may be secreted back into the intestinal lumen by a P-glycoprotein efflux carrier system present in the brush border membranes of enterocytes and colonocytes (Thietbaut et al., 1987, Burton et al., 1993, Fricker et al., 1996). The P-glycoprotein efflux carrier system was shown to be saturable and ATP dependent. Burton et al. (1993) showed for the first time the participation of P-glycoprotein efflux system in the active resistance to peptide transport in Caco-2 cells.
The transport of the model peptide AcPhe(NMPhe)2NH2 in the basolateral to apical direction was greater than the apical to basolateral direction in Caco-2 cell monolayers. The basolateral transport was reduced in the presence of the P-glycoprotein inhibitor verapamil. Burton et al. (1993) speculated that P-glycoprotein may present an additional barrier to the non-specific intestinal absorption of peptides. Other studies have shown peptides to be substrates for the P-glycoprotein (Toppmeyer et al., 1994, Lang et al., 1997); however the importance of the P-glycoprotein transport system in the intestinal absorption of peptides in vivo has yet to be determined. 9
1.3 Strategies to improve peptide and protein delivery
1.3.1 Protease inhibitors
Protease inhibitors should 1mprove the oral absorption of therapeutic peptides and I•.;- proteins by reducing their proteolytic breakdown in the intestinal tract. A number of investigators have attempted to enhance the oral delivery of peptides and proteins by using protease inhibitors, notably the pancreatic protease inhibitors.
,, 1.3.1.1 Serine protease inhibitors
~ Aprotinin (MW 6 500) has inhibitory activity towards both trypsin and chymotrypsin
, (Kassell et al., 1965). A number of investigators have shown that the co-administration of this inhibitor with peptide has improved the absorption of the peptide across the intestine.
·\ Co-administration of aprotinin with insulin and sodium salicylate (an enhancer) potentiated the amount of insulin absorbed across the colon of the rat (Nishiata et al., 1983). Aprotinin and sodium cholate (an enhancer) improved the absorption of intact insulin across the ileal intestine of rat 200-fold (Ziv et al., 1986). No damage to the intestine was observed by the presence of sodium cholate and aprotinin in the intestinal lmnen. The inclusion of aprotinin in insulin-containing capsules also increased the :I,-'" absorption of insulin across the colon of pigs (Elias et al., 1992).
Two serine protease inhibitors have been isolated from soybean, Kuntiz inhibitor,
~ .A MW 22 000 (Kruntiz, 1947) and the Bowman-Birk type inhibitor (MW 7 975) (Seidl and Liener, 1972). Kuntiz type inhibitor is known as a strong inhibitor of bovine trypsin and weak inhibitor of bovine chymotrypsin, while the Bowman-Birk type inhibitor is a strong inhibitor of both serine proteases. Kidron et al. (1982) observed that the direct injection of insulin into the rat ileal lumen only .resulted in significant lowering of blood glucose when co-administered with the Kuntiz type soybean trypsin inhibitor. 10
1.3.1.2 Low molecular weight peptidase inhibitors
There are peptidases present in the intestine which have a large number of specificities (Tables 1.1 & 1.2). A number of compounds have been shown to have specific inhibitory
.. ( activity towards these peptidases in vitro (Table 1.3). A major limitation is that a number
', of these low molecular weight inhibitors including elastinal and bestatin can cause systemic side effects when given orally (Yagi et al., 1980, Ito et al., 1983) .
. \ In vitro studies showed that thiorphan inhibited 85% of the hydrolytic activity of rat intestinal brush-borders membranes towards an anti AIDS octapeptide drug (Su and Amidon, 1995), however the effect of this inhibitor on improving oral bioavailability of the octapeptide was not tested.
Puromycin and bestatin (0.5 mM) both protected metkephamid from proteolytic degradation in the perfused small intestine ofrat (Taki et al., 1995). However only puromycin significantly improved the absorption of metkephamid across the small intestine. It was suggested that puromycin is able to diffuse into the intestinal cell with metkephamid protecting it from proteolytic attack during permeation across the cell.
Table 1.3 Specificity oflow molecular weight peptidase inhibitors. - ' Inhibitor Specific inhibitory activity Reference Bestatin leucine aminopeptidase, Suda et al. (1976)
·'· aminopeptidase B, and aminopeptidase N A:,
Amastatin aminopeptidase A and leucine Aoyagi et al. (1978) aminopeptidase
,\. Puromycin aminopeptidase B aiid N Taki et al. (1995) _.J- 7 {
Elastinal elastase Umezawa et al. (1973)
Thiorphan endopeptidase24.11 Bai (1994b) 11
1.3.1.3 Bile salts
In vitro experiments have shown bile salts inhibit proteolytic enzymes of the intestinal lumen (Bai et al., 1995b) and mucosa in mammals (Yamamoto et al., 1990, Bai, 1994a). Bile salts are amphipathic compounds and therefore at a certain concentration they self associate to form micelles. This concentration is known as the critical micelle concentration (CMC). The CMC of a molecule can change depending on the solution pH, temperature, counter ion and electrolyte concentration (Florence and Attwood, 1998). Roda et al. (1990) reported the CMCs of deoxycholic, glycocholic, glycodeoxycholic and taurocholic acids in deionised water to be 10, 12, 6, and 10 mM respectively, while in 150 mM sodium chloride the respective CMCs were 3, 10, 2 and 6 mM.
Deoxycholate at 25 mM was shown to be more effective than the enzyme inhibitors
,, aprotinin (0.015 mM) or p-.chloromercuriphenylsulfonic acid (0.05 mM) in protecting \ insulin and proinsulin from the proteolytic activity of the ileal and rectal mucosa! homogenates of rabbits (Yamamoto et al., 1990). Glycocholate at 20 mM also inhibited the degradation of the peptide ebiratide in the intestinal lumen and mucosal homogenates ofrat (Okagawa et al., 1994). The mechanism of enzyme inhibition by bile salts above the CMC may be due to the entrapment the protein or peptide protecting it from the proteolytic enzymes (Lee et al., 1991). However, studies at high concentrations ofbiie salt(> 5 mM) have resulted in permanent damage to the intestinal cells (Mekhjian and Phillips, 1970). Damage of cells is believed to be due to the detergent characteristics ~f bile salt molecules (Shiau, 1981, Lester et al., 1983).
The mechanism of enzyme inhibition by bile salts at concentrations below CMC is not well understood. Bile salts at 2 mM are able to inhibit the insulin degrading activity by lumenal contents of the colon of rats. Glycodeoxycholate, deoxycholate and taurodeoxycholate were described as weak to moderate inhibitors, inhibiting insulin degrading activity by 36, 24 and 20% of the control activity, respectively (Bai et al., 1995b). Interestingly, taurocholate and glycocholate at 2 mM were not effective inhibitors of proteolytic degradation of insulin. 12
Bile salts at 2 mM have also been shown to inhibit the proteolytic activities of the rat intestinal brush-border peptidases (Bai, 1994a). Bile salts sodium deoxycholate and sodium glycodeoxycholate, inhibited endopeptidase-24.11 by 17 and 29% of the control activity. Endopeptidase-2 was inhibited by sodium glycocholate (20%), sodium taurocholate (47%) and sodium taurodeoxycholate (41 %). Dipeptidyl peptidase IV was inhibited by sodium taurocholate (32%) and sodium taurodeoxycholate (19%), and '- carboxypepitdase P was inhibited by sodium deoxycholate (50%) and sodium taurodeoxycholate (57%). No bile salts inhibited the activity of aminopeptidases W
'( andP.
Since endopeptidase 24.11 and endopeptidase-2 have similar substrate specificities and are inhibited by different bile salts, enzyme inhibition cannot be due to competition ·' between substrate and bile salt for the substrate binding site (Bai, 1994a). The author speculated that bile salts at concentrations below their CMC bind to the enzymes resulting in conformational changes to the three-dimensional protein structure rendering it inactive. Previously, Milov et al. (1992) showed bile salts bind at 3 to 6 sites on human carbonic anhydrase I and II, resulting in conformational changes and inhibition of enzymatic activity.
C .1
'( 1.3.1.4 Chelating agents Complexing agents inhibit enzymes by sequestering divalent cations from the enzyme structure. Divalent cations are co-factors important for the maintenance of the protein structure required for their activity. Zinc requiring enzymes such as carboxypeptidase A, Band aminopeptidase N are effectively inhibited by complexing agents such as EDTA (Ikesue et al., 1993, Bernkop-Schnilrch and Krajicek, 1998).
1_ .\
EDTA improved the stability of the neuroleptic peptide, Des-enkephalin-y-endorphin in the rectal lumen and ligated colon of rat (van Hoogdalem et al., 1989). However EDTA was only effective in improving oral bioavailability of the peptide in the presence of a permeation enhancer (van Hoogdalem et al., 1989). 13
Bacitracin, a cyclic dodecapeptide (MW 1423) obtained from Bacillus licheniformis
2 inhibits the activity of aminopeptidase N by chelation of Zn + from the enzyme . (Sangadala et al., 1994). In situ experiments showed bacitracin to be more effective than soybean trypsin inhibitor and aprotinin in promoting the absorption of biologically active insulin across the large intestine of rat (Yamamoto et al., 1994). However, bacitracin was ineffective in promoting absorption of biologically active insulin in the small intestine of rat.
;( Since the effect of chelating agents depends on the relative affinity of chelator and enzyme for the divalent ion, not all enzymes requiring divalent cations for their stability are inhibited by chelating agents. Binding of calcium to trypsin has been shown to
'/ increase the stability of trypsin against autolysis (Bode and Schwager, 1975). However .\ incubation ofEDTA with trypsin had no effect on the enzymatic activity of trypsin (Russin et al., 1974, Lue.Ben et al., 1996, Bernkop-Schnurch and Krajicek, 1998). Russin et al. (1974) showed that EDTA reduced the autolytic stability of bacterial trypsin only in ) the presence of 8 M urea. This suggests that the calcium is tightly bound and only accessible to chelation by EDTA after some relaxation of the protein conformation .
...... ) 1.3.1.5 Mucoadhesive polymer$ Polymers of acrylic acid have been shown to inhibit a broad range of proteolytic enzymes of the gastrointestinal tract in vitro (Bai et al., 1995b, Lue.Ben et al., 1996c). These r- ~-~· polymers are used in oral formulations for their mucoadhesive properties (Gu et al., 1988,
j Harris et al., 1990). The inhibitory activity of these polymers was first observed when dispersions of polycarbophil (1 % w/v) completely inhibited the proteolytic degradation of 9-desglycinamide, 8-arginine vasopressin by the mucosal homogenate of rat (Lehr et al., 1992a). Junginger and co-workers further demonstrated that the acrylic acid polymers, polycarbophil and Carbopol 934P were able to inhibit a large range oflumenal proteolytic enzymes including: trypsin, a-chymotrypsin, carboxypeptidase A and cytosolic leucine aminopeptidase (Lue.Ben et al., 1996c) and brnsh-border peptidases of 14
rat (LueBen et al., 1996a). It was postulated that the inhibitory effect of these polymers is based on their ability to complex essential divalent ions from the enzyme structure (Lue13en et al., 1995).
( I It has also been suggested that inhibition of proteases by polyacrylic acid was dependent on the polymers ability to reduce the local pH below the optimum pH for the proteolytic activity (Bai et al., 1995b). Polyacrylic acids, including Carbopol 974P, 971P and 934P were shown to inhibit the hydrolysis of insulin, calcitonin and insulin-like growth factor by trypsin, chymotrypsin and microbial proteases of the rat colon (Bai et al., 1995b). The ;( degree of inhibition correlated with the ability of the polyacrylic acid to lower the pH in vitro (Bai et al., 1995b).
These polymers offer a number of advantages over other protease inhibitors. Polyacrylic .' acids are not absorbed across the intestinal wall and therefore are not likely to have systemic toxicity. They have inhibitory activity towards a broad range of enzymes including those of the lumen and the brush borders.Also having mucoadhesive properties, the inhibitory activity of the polymer may be localised.
More recently, Bernkop-Schnilrch and co-workers have developed several mucoadhesive polymer-inhibitor conjugates that protect co-administered peptides from specific
'( proteolytic attack, for review Bernkop-Schniirch (1998. The immobilization of protease inhibitors on mucoadhesive polymers offers several potential advantages. Conjugated inhibitors will not reach the systemic circulation and the mucoadhesive properties of the polymers will provide intimate and prolonged contact of the inhibitor to the adsorbing intestinal membrane.
Chymotryptic inhibitors, chymostatin- and soybean trypsin inhibitor (Bowman-Birk) were conjugated to polyacrylic acid (Carbopol 940). The soybean trypsin inhibitor - polymer conjugate was 17% more effective in the inhibition of chymotrypsin than the chymostatin - polymer conjugate. However, the mucoadhesive strength of the soybean 15
trypsin inhibitor - polymer conjugate was 35% lower than the unmodified polymer and the chymostatin - polymer conjugate (Bernkop-Schniirch and Apprich, 1997).
The aminopeptidase inhibitor, bacitracin was conjugated to Carbopol 940 and had greater
'\ inhibitory effect on aminopeptidase activity compared to the unconjugated bacitracin ., (Bernkop-Schriurch and Marschtitz, 1997). However, in this study it was also observed that the neutralised Carbopol 940P inhibited the activity of aminopeptidase N, but weakly " by comparison. ·e '1
EDTA was covalently attached to chitosan, Carbopol 934P and polycarbophil (Bernkop Schntirch and Krajicek, 1998). Chitosan-EDTA conjugates had the strongest inhibitory '.1 effect on carboxypeptidase A and aminopeptidase N. The chitosan-EDTA conjugate i exhibited the highest affinity and lowest capacity for both calcium and zinc, suggesting binding affinity for divalent ions to be a more important parameter in enzyme inhibition.
1.3.2 Permeation enhancers I ,) Penetration enhancers improve oral absorption by their action on the transcellular and/or paracellular transport routes (Figure 1.1 ). The two main classes of permeation enhancers are chelating agents and detergents. Most permeation enhancers are detergents including bile salts, sodium dodecyl sulfate and Triton X-100. It is believed that their detergent
·( properties alter the membrane lipid organisation, increasing transport via the transcellular route. Chelating agents effect the paracellular pathway by disrupting the integrity of occluding junctional complexes by chelating calcium or magnesium from the tight junctions (Tindall, 1964, Gumbiner, 1987). There is an enormous literature on enhancement of drug permeability across the intestinal tract. It is beyond the scope of this
;.. "f, introduction to review it all but recent reviews include Aungst (1993) and Sayani and Chein (1996).
Numerous studies have shown that bile salts improve the absorption of peptides and proteins across the intestinal tract (Ziv et al., 1981, Kidron et al., 1982). Model 16
_, membrane studies have shown bile salts at low concentration (below the CMC) increase
~ ~ ,,. bilayer permeability, however at higher concentrations (usually above the CMC) bile salts cause disruption to the bilayer (Walde et al., 1987). Common bile salts can be divided into two groups: the dihydroxy- and the trihydroxy-bile salts such as deoxycholate (DOC) -< and cholate, respectively. The dihydroxy-bile salts have been found to be more effective _, ' absorption enhancers than the tri - hydroxy bile salts ( Chadwick et al., 1979, Ziv et al., 1983); however the dihydroxy-bile salts are more damaging morphologically (Teem and Philips, 1972, Chadwick et al., 1979). Direct injection of insulin with 2 mg of either -·- ·-;..
;( cholate or DOC into the colon promoted absorption of insulin across the intestinal epithelium in :q_ormal and diabetic rats (Kidron et al., 1982). DOC was shown to be more effective than the trihydroxy cholate in promoting insulin absorption (Kidron et al., --- \ 1982).
', Chelating agent, citric acid has been reported to enhance the membrane absorption of heparin (Sue et al., 1976) and LHRH (Okada et al., 1982). Citric acid was shown to act as - _, a permeation enhancer at concentrations as low as 0.1 mM in Caco-2 cell monolayers (Cho et al., 1989). EDTA increased the permeation of thyrotropin-releasing hormone and Met-enkephalin analogues across Caco-2 cell monolayers (Lang et al., 1997). -1
.. ,_; Acrylic acid polymers have been shown to increase the permeation of peptides via the -r paracellular route (Lue13en et al., 1994). Junginger and co-workers reported that (,, polyacrylic acids reduced the trans epithelium electrical resistance (TEER) of Caco-2 cell monolayers; which suggested opening of the tight junctions. Junginger and co-workers postulated that the polyacrylic acids disrupted the integrity of the tight junctions by
I ;-'( chelating calcium from the tight junctions (Lue13en et al., 1994, Borchard et al., 1996).
)- ' However the opening of the tight junctions to the paracellular transport of FITC-dextran
'·\ (MW 4 400) only occurred at pH 4.0, but not at pH 7.4. It is not known whether the paracellular transport is due to the action of the polyacrylic acid or due to the damage of cells at the low pH. Only when polyacrylic acid was applied to the basolateral side did the TEER drop sufficiently for paracellular transport ofFITC-dextran. Similarly, it has 17
previously demonstrated that EDTA i~ more effective in opening tight junctions when applied to the basolateral side of Caco-2 monolayers (Noach et al., 1993).
Chitosan is a linear polycationic polysaccharide composed of poly(2-deoxy-2-amino glucan). It was reported to enhance the absorption of insulin and calcitonin across the nasal epithelium (Illum et al., 1994). The enhanced absorption of the peptide drugs was believed to be due the mucoadhesive properties of the chitosans and the enhancement of the paracellular transport route. Application of chitosan to the apical side of Caco-2 cell · also increased the permeability of the paracellular transport marker mannitol (Artursson et al., 1994). The influence of chitosan on transport of the marker molecules was strongest when the pH was well below 6.5. At pH 6.0, 4.9, and 4.0 the apparent
. \ permeability coefficients ofmannitol were 1.7, 2.9, and 4.9 emfs x 107, respectively. As chitosan has an average pKa of 6.5, it was postulated that charge density is important for enhancement of the intestinal mucosa."Work by Schipper et al. (1996) showed that enhancement is dependent on the interaction of positively charged chitosan polymer with
J, the negatively charged surface of the epithelial cell membrane. .\
Intraduodenal administration of the peptide analogue buserelin in the presence of chitosan
-- J resulted in a 5-fold increase in its systemic absorption (Lue.Ben et al., 1996b). Chitosan
·( was more effective than polyacrylic acids in improving the absorption ofbuserelin. Since •/ chitosan did not inhibit trypsin and carboxypeptidase B activity it was concluded that the permeation barrier is greater than the metabolic barrier for the oral absorption of buserelin.
In vitro studies by Lue.Ben et al. (1997) demonstrated that chitosan (1 % w/v) was more effective than the polyacrylic acid, polycarbophil (1 % w/v) in enhancing the absorption
.\ of 9-desglycinamide, 8 -arginine vasopressin (DGV AP) across Caco-2 cell monolayers. However in the rat vertically perfused intestinal loop, both polymers enhanced the absorption of DGAVP similarly. It was concluded that chitosan enhanced the transport of 18
DGVAP via the paracellular route whereas polycarbophil enhanced transport by protecting the peptide from the metabolic activity of the intestine.
Chitosan is poorly soluble at neutral and alkaline pH. More recently a water soluble derivative of chitosan, N-trimethyl chitosan chloride (TMC) was synthesised (Kotze et ·( al., 1997). The N-trimethyl chitosan at pH 7.4 reduced the TEER, presumably by opening the tight junctions and enhanced the transport of mannitol, FITC-Dextran 4400 and -, buserelin across the Caco-2 cell monolayers. Kotze et al. (1997) compared the enhancing effects of two chitosan salts, hydrochloride and glutamate and TMC at 1.5% (w/v) at pH values between 4.4 and 6.2. All chitosans improved the adsorption of peptide drugs across Caco-2 cell monolayers. The rank order of effectiveness at 1.5% (w/v) was .- \ chitosan hydrochloride> chitosan glutamate> TMC. However due to TMCs higher solubility, permeation enhancement could be improved by increasing the concentration of TMC to 2.5% (w/v) .
• .I,, 1.3.3 Chemical modification .\ Chemical modification of peptides offers a number of advantages, including increased
./ potency, resistance to proteolysis and increased in membrane permeability. The modified - , ' peptide may be active in its own right or active only after regeneration in vivo after ( adsorption.
The LHRH analogue leuprolide [D-Leu6, Pro6NEt] LHRH, was chemically modified at positions 1, 2, or 3 and N-methylated at positions 1, 2, or 4 (Raviv et al., 1992). The analogues were tested for receptor binding, LH release from rat pituitary cells, clearance,
stability towards chymotrypsin and rat intestinal degradation. The analogues illAe-Ser4,
)- ·' NMe-Phe2, NAc-Sarl and 1-NaI3 -LHRH showed up to 60-fold increase in stability towards chymotrypsin and a 20-fold increase in stability towards intestinal lumenal degradation. Several of the analogues showed potency in the range of leuprolide. There was greater clearance of analogues that were more hydrophobic than leuprolide, 19
.( suggesting a relationship between clearance and hydrophobicity, Even though these analogues were stabilised against enzymatic degradation there was no significant increase in intestinal absorption, indicating that intestinal absorption of these analogues is not limited by enzymatic stability. '\
( Recently, Fujita et al. (1996) improved the intestinal absorption of human calcitonin (hCT) by chemical modification with short-chain length fatty acids such as acetic acid
18 and caproic acid. The fatty acids were attached to s-amino group ofLys • The biological
>' activities of the acetyl (Ac-hCT) and caproyl (Cap-hCT) derivatives were approximately 70% and 40% of the native peptide, respectively. The chemical modification of calcitonin with fatty acids improved its lipophilicity and metabolic stability in the intestinal mucosa -~ ofrats (Fujita et al., 1996). Chemical modification improved small and large intestine absorption of calcitonin from 0.5% to 3.5% and 4.5% for Ac-hCT and Cap-hCT, respectively.
" Physiological transport pathways including vitamin B 12 (Russell-Jones, 1998) and bile salt (Kramer et al., 1994) have been used to improve transport of peptides across biological membranes. Hexa- and heptapeptides ofD-amino acids couple to modified bile -I acids showed a 14-fold increase in adsorption across perfused rat ileum (Kramer et al.,
·( 1994). The uptake was shown to involve an active transport process.
,}
.J.. :,
,\ 20
1.3.4 Formulation
Formulations of peptide and protein drugs are designed to protect the drug from the proteolytic enzymes of the intestinal lumen and/or release the drug at the site of the GI
-<\ tract most favorable for absorption. A number of different formulations have been_ investigated to improve the absorption of peptide drugs (Table 1.4).
Table 1.4 Formulations designed to improve the oral delivery of peptides. Delivery system Peptide Reference
>- liposomes insulin Patel et al. (1982) calcitonin Fukunaga et al. ( 1991)
polyacrylate-coated soft gelatin insulin Touitou and Rubinstein (1986) -· capsules for targeted delivery to the colon
azopolymer-coated hard gelatin insulin Saffran et al. (1986) capsules for targeted delivery to the colon .I
/ -\ . nanoparticles insulin Michel et al. (1991)
microemulsions (water-in-oil) insulin Constantinides et al. (1994)
polyacrylate gel containg long insulin Morimoto et al. (1983) chain fatty acids (mucoadhesion) ·(
Many investigators have speculated that the colon is a superior region for peptide absorption. This is based on the assumption that the overall proteolytic activity of the ·,..... > colon is low compared to the small intestine (Saffran et al., 1986, Yamamoto et al., 1994) and the greater effectivity of permeation enhancers (Ishizawa et al., 1987, Tomita et al., 1988). To take advantage of the characteristics of the colon a number of different
1, ' potential formulations have been developed to target the colon (Rubinstein et al., 1997). •\
Time-controlled and pH-dependant release dosages have been examined for the specific delivery of drugs to different regions of the intestine (Murray and Tucker, 1990, Ashford et al., 1993). In time-controlled release dosage forms the drug is released after a specific 21
time interval believed to be the time required for the dosage form to reach the colon. The pH-dependant dosage forms are dependant on the enteric-coating polymer being stable or insoluble in the low pH of the stomach or neutral pH of the small intestine but dissolve at the higher pH of the colon. Soft gelatin capsules coated with the pH sensitive acrylic
. ', polymer (Eudragit RS, Land Sat various ratios) containing insulin were shown to release \ the peptide only above pH 7.0. Oral administration of insulin in these capsules to rats resulted in significant drop in blood glucose compared to control (Touitou and Rubinstein, 1986). However pH-dependant and time-controlled systems lack reliability -- J due to large variations in transit times and pH due to diet, food intake, intestinal mobility and disease state (Pozzi et al., 1994, Wilding et al., 1994) .
.,,;_ Many colonic-specific drug delivery systems rely on enzymes unique to the gut microflora to release active agents in the colon (Rubinstein et al., 1997). Polysaccharidases and azoreductases are the two main classes of enzymes that are produced by colon microflora considered reproducible enough to exploit in drug targeting (Ashford and Fell, 1994). The most original work in this area was by Saffran and co workers (Saffran et al., 1986, Saffran et al., 1990). Solid dosage forms of the peptide drug were coated with polymers cross-linked with azo-bonds, which are degraded by the
-~ ;' indigenous microflora of the colon. A sustained pharmacological response for insulin was
·( observed when the coated delivery systems were orally administered to rats. It was proposed that once the polypeptide reached the colon the indigenous micro flora reduced the azo bonds, thus breaking the cross-links and releasing the polypeptide for absorption. /
Tozaki et al. (1997) demonstrated that chitosan capsules were stable in the stomach and the small intestine and were degraded specifically by microorganisms in the rat colon. c.,-·':,. The mechanism by which the chitosan capsules are degraded in rat cecal contents is not known. Oral administration of chitosan capsules containing insulin, aprotinin (protease inhibitor) and sodium glycocholate ( enhancer) resulted in a significant pharmacological response to the insulin. 22
Peptides and proteins that are released by formulations at specific sites of the GI tract are moved from the site of administration by the cleaning action of the housekeeper wave and are hindered from adsorption by the mucus gel on the surface of the intestinal epithelium. Polymers which are able to interact with the mucus of the GI tract (mucoadhesives) can potentially overcome these problems (Smart et al., 1984, Gu et al., 1988). These -,' mucoadhesive polymers have three desirable features for oral drug delivery: prolonging residence time at the site of drug absorption, improving contact with the absorbing mucosa and localisation in specific regions to improve and enhance the bioavailability of the drug. Close contact with the absorbing mucosa might not only promote absorption of the drug but could also localise the action of the excipients such as protease inhibitors and permeation enhancers present in the formulation. - -~
The cross-linked polyacrylic acid derivatives polycarbophil (Noveon® AAl, weakly '- crosslinked with divinylglycol) and carbomer (Carbopol®, crosslinked with allyl sucrose) have strong mucoadhesive properties (Gu et al., 1988, Harris et al., 1990). Microspheres of poly (hydroxyethylmethacrylate) coated with the polycarbophil (Noveon® AAl) or Carbopol934 (Carbopol®) were shown to have greater adhesion to pig intestinal mucosa than uncoated microspheres in vitro. However, in situ intestinal transit experiments in the rat ileum showed that microspheres coated with polycarbophil showed only significant
i mucoadhesive behavior during the first hour of delivery (Junginger et al., 1990). A number of other studies have shown that the cross-linked polyacrylic acid derivatives improve the absorption of peptides in vivo (Lehr et al., 1992a, Ilan et al., 1996, Baluom et al., 1997). However because of their apparent protease inhibitory and mucosal permeation enhancement activity (Sections 1.3 .1.5 and 1.3 .2) it is difficult to quantify the contribution of the polymers mucoadhesive properties to the improved absorption of peptides. ;\ 23
Chitosan has mucoadhesive properties which are believed to be mediated through ionic interaction between positively charged amino groups in chitosan and negatively charged sialic acid residues in mucus or on cell surface (Lehr et al., 1992b). Chitosan-coated liposomes (CS-liposomes) were shown to have greater adhesiveness to the intestinal -/
•" ( surface of rat compared to uncoated liposomes (Takeuchi et al., 1996). The oral delivery
1 of CS-liposomes containing insulin resulted in a significant drop in blood glucose level in rats. When 5 IU of insulin was delivered orally in chitosan-coated liposomes the drop in blood glucose was equivalent to only 10% of the biological response observed when the -~ ( formulation was delivered by injection. Oral administration of non-coated liposomes
-; resulted in a biological response less than one fifth of the response for CS-liposomes.
>
- ( 1.4 Oral delivery of peptides and proteins to teleosts
1.4.1 Digestive tract of gastric teleosts In order to achieve oral delivery of peptides an understanding of the anatomy and physiology of the intestinal tract is required. A schematic of the gastrointestinal tract of the salmon is shown in Figure 1.2. The digestive tract starts with the short oesophagus, which joins the J-shape stomach. The stomach has muscular closures at each end, the
- .....- I cardiac sphincter anteriorly and the pyloric sphincter (pylorus) posteriorly. Posterior to ! the pylorus is the start of the anterior section of the intestine and adjacent are the hepatic and pancreatic ducts which enter into the anterior intestine. At the beginning of the / } anterior section there are clusters of finger like diverticula, called pyloric ceaca. The numerous diverticula of the intestine increases the surface area of the intestine. The surface area of trout pyloric ceaca was shown to be three times that of the anterior gut and -~ "· twice that of the total intestine (Ezeasor and Stoke, 1980). Bergot et al (1975) referred to the pyloric ceaca in rainbow trout as being histologically identical to the anterior ,\ intestine. The intestinal section posterior to the so-called intestino-rectal valve is morphologically distinct from the anterior section. For the purpose of this study the intestinal region with ceaca attached is referred to as the anterior intestine, the remaining 24
intestine which has no pyloric ceaca is further divided into two sections, the middle and posterior which are separated by a so-called intestino-rectal valve (Figure 1.2).
Oesphagus ~ ~ Intestine
---- Posterior .c ~
>
Figure 1.2 Schematic representation of the salmonid gastrointestinal tract.
Scanning electron microscopy studies in trout showed the different topographical characteristics of the ceaca, the intestine (anterior, middle) and the rectal (posterior) mucosa (Ezeasor and Stoke, 1980). The surface of the diverticula has numerous parallel longitudinal ridges, which are believed to facilitate the passage of food in and out of the \ diverticula. The intestine is covered with numerous short blunt projections, conforming to the villi. The villi of the fish intestine appeared thicker and shorter than those described in the mammalian intestine. The rectum is deeply folded, the top of the folds are lined by non-vacuolated cells, while the sides of the folds have vacuolated cells suggestive of protein or other macromolecule uptake by pinocytosis (Ezeasor and Stokoe, 1981).
The intestinal mucosa ofteleosts is lined with a single layer of columnar epithelium and scattered among the epithelial cells are mucus-secreting goblet cells (Fange and Grove, 1979) (Figure 1.3). Closely following the basal contours of epithelial cells is the basal lamina, a narrow band of complex collagen and mucopolysaccharides. Adjacent to the basal lamina is the lamina propria, which has a rich supply of blood capillaries (Hinge and Grove, 1979, Yasutake and Wales, 1983). Adjacent to the lamina propria is a thin . layer containing collagen and elastic fibers as well as blood vessels and nerves containing 25
two distinct regions known as the stratum compactum and stratum granulosum. The stratum granulosum layer is infiltrated by eosinophilic granular cells (EGC). The muscularis layer is formed by inner circular and outer longitudinal muscles and is surrounded by a serous membrane, composed of simple flat endothelium and a layer of loose connective tissue (tella subserosa) (Takashima and Hibiya, 1995).
I
,, 'IP' l. ~ ~ ,....~ ~, !_ _.__ ;;,. r-'"'-l r- ~ I ~ ~ )"- -,.;~ 'R.~ I .;._ . ~ • ? Jt "'" . , -- ,Sr .::;... ,,."
,/ 11,c 1. . \ ' r;;;•• ;~ i·A \·.• .·_·. !\ :,'.',:.} :i Y ~.)u~· 1( "" 1''' i i ( l\1,, \.i'
E
N 0\
·H
Figure 1.3 Transverse section (200 X) of the wall of the posterior intestine of a salmon (Oncorhynchus tshawytscha). Mounted sections were stained with haematoxylin and eosin: (A) epithelium, (B) goblet cell, (C) lamina propr1a, (D) stratum compactum, (E) stratum granulosum, (F) muscularis circularis, (G) muscularis longitudinalis and (H) tella subserosa (with permission from Phill Veillete) 27
1.4.2 Biochemical barriers of the intestinal tract of teleosts
1.4.2.1 Lumenal proteases
Proteolytic digestion in the intestine can occur in the gut lumen, either at the cell membrane associated with microvillous-brush border or intra-cellularly in the enterocytes as mentioned previously (Section 1.2.1). Enzymes found in the intestinal lumen can potentially come from either the pancreas or from epithelial cells released from the gut wall; however for fish the lumenal proteases are generally considered to be predominantly of pancreatic origin (Yoshinaka et al., 1981). The exocrine pancreas is found in the fat tissue diffused around the pyloric ceaca (Anderson and Mitchum, 1974). > The digestive enzymes secreted form the pancreatic tissue are carried into the anterior intestine through a series of ducts. The pancreatic proteases that have been detected in pyloric ceaca and the intestine of teleosts include the trypsins, chymotrypsins, elastases, carboxypeptidases A and Band aminopeptidases (Fange and Grove, 1979, Cohen et al., 1981a, Clarke et al., 1985). r.1
\ Several isoforms of trypsin have been described for teleosts including Atlantic cod (Asgeirsson et al., 1989), sardine (Murakami and Noda, 1981), anchovy (Martinez et al., -r I 1988), Atlantic salmon and rainbow trout (Torrissen, 1984). Interestingly, trypsin ( enzymes isolated from cold-adapted species such as anchovy (Martinez et al., 1988). ·i Atlantic cod (Asgeirsson et al., 1989), and salmon (Taran and Smovdyr, 1992) display " substantially higher catalytic efficiencies (kcat / ~ ) than mammalian trypsins. Work by ":- 1 Fletcher et al. (1987) on rat anionic and cationic isoforms of trypsin showed that the substrate binding sites of the two isoforms had different charge distributions and therefore may have different binding preferences for protein substrates. Murakami and Noda
,\ (1981) isolated two trypsin-like proteases from the pyloric ceaca of sardine, but the soybean trypsin inhibitor and trypsin specific inhibitor N-p-tosyl-L-lysine chloromethyl ketone (TLCK) inhibited only one of the proteases. 28
Mammalian cationic trypsins have been shown to be stabilized against autolysis in the presence of calcium ions (Vithayathil et al., 1961, Walsh, 1970). Trypsin enzymes from catfish (Yoshinaka et al., 1984), anchovy (Martinez et al., 1988) and carp (Cohen et al., 1981a) were also shown to _be dependent upon calcium for their stability and activity. However trypsin enzymes from sardine (Murakami and Noda, 1981), capelin
{ (Hjelmeland and Raa, 1982) and chum salmon (Uchida et al., 1984a, Uchida et al., 1984b) were not stabilised by calcium.
"'','
·\ Chymotrypsin or chymotrypsin-like enzymes have been isolated from dogfish (Racicot and Hultin, 1987, Asgeirsson and Bjarnason, 1991), carp (Cohen et al., 1981a), Atlantic cod (Raae and Walther, 1989) and rainbow trout (Kristjansson and Nielsen, 1992). Many properties of these enzymes are similar to those of bovine and porcine chymotrypsins, however some important differences have been observed. The chymotrypsins from fish were shown to have higher catalytic activities than the bovine enzyme (Cohen et al., 1981b, Ramakrishna et al., 1987, Racicot and Hultin, 1987, Raae, 1990, Kristjansson and Nielsen, 1992). ,\
f • Two anionic chymotrypsin-like proteases were isolated from the pyloric ceaca ofrainbow ' I trout (Kristjansson and Nielsen, 1992). The catalytic efficiencies (k /K ) of the trout cat m ., chymotrypsins I and II against a chymotryptic synthetic substrate were approximately i 4 and 7-fold higher than the bovine enzyme, respectively. The trout chymotrypsins hydrolytic activities against casein were 3-fold higher than for the bovine enzyme. Trout chymotrypsins also showed different heat-stability to bovine trypsin in the presence and absence of calcium, indicating different binding affinities for calcium. The different calcium-binding properties of the two trout chymotrypsins indicate some structural differences may exist in and around the calcium binding loops of these enzymes. •\
Interestingly, fish trypsins and chymotrypsins were shown to be more sensitive to soybean trypsin inhibitor (Kuntiz type) compared to mammalian enzymes (Croston, 1960, Krogdahl and Holm, 1983). 29
The serine protease inhibitor aprotinin is a potent inhibitor of mammalian chymotrypsins (Kassell et al., 1965). Two isoforms of chymotrypsin isolated from trout had variable behaviour towards the aprotinin inhibitor (Blow, 1971). One isoform was inhibited to 86% of its control activity while the activity of the second isoform was enhanced in the presence of inhibitor, suggesting structural differences in the active sites between the isoforms. Aprotinin also had a lower inhibitory activity against cod chymotrypsins l . , compared to the bovine enzyme (Asgeirsson and Bjamason, 1991).
't Chymotrypsin-like enzymes of trout (Kristjansson and Nielsen, 1992).and carp (Cohen et al., 1981) have been shown to be more active at alkaline pH compared to the bovine enzyme. The authors suggest that there are different groups involved in stabilising the active site conformation of the fish and the bovine enzyme. Chymotrysin enzymes from .l ' trout (Kristjansson and Nielsen, 1992), carp (Cohen et al., 1981a), and Atlantic cod (Raae and Walther, 1989, Asgeirsson and Bjamason, 1991) were shown to consist of a single
. ,\ polypeptide chain. This is in contrasts to bovine a-chymotrypsin, which contains three ( polypeptide chains linked together by disulphide bonds (Blow, 1971).
Pancreatic elastase has been isoiated from gummy shark, red stingray, rainbow trout, ., carp, eel, bluefin tuna, yellowtail, sea bass and angler, confirming that elastase is an ( important enzyme of digestion in fish (Y oshinaka et al., 1985). In anchovy, Martinez and Serra (1989) assayed crude intestinal extracts for different proteolytic activities using synthetic substrates. Even though proteolytic enzymes have different catalytic activities towards substrates, they speculated that elastase was the dominant digestive enzyme in the pyloric ceaca and intestine.
·\ A pancreatic elastase was purified and characterized from a pancreatic extract of the catfish (Yoshinaka et al., 1984). The catfish elastase was found to be a serine proteinase and showed a high degree of similarity to porcine pancreatic elastase with respect to pH stability and temperature dependent activity. Amino acid composition of elastase purified 30
from carp pancreas showed high similarities to the respective bovine and porcine enzymes (Cohen et al., 1981). The kinetic properties of carp elastase were also found to be similar to the respective rat and porcine enzymes (Cohen et al., 1981b). Inhibition of carp elastase by turkey ovomucoid was also similar to porcine elastase (Cohen et al., 1981b).
An intestinal elastase purified from Atlantic cod had a catalytic efficiency (kcaJK111J 2-fold greater than observed for porcine elastase (Raae and Walther, 1989). The elastase ·'( had the greatest activity towards the carbonyl side of the alanine residues, but also hydrolysed valine and leucine residues. The elastase had low heat stability and towards acidic pH compared to porcine elastas~.
2 Metalloproteinases are a group of proteases that require Zn + for activity. This can be demonstrated by the loss of activity on the addition of the bivalent chelators EDTA and EGTA. Metalloproteinases have been found in bluefi.n tuna, yellowtail (Yoshinaka et al., 1985), and catfish (Yoshinaka et al., 1984). Interestingly, the metalloproteinase enzyme
from catfish hydrolyses the elastase substrates elastin and N-succinyl-Ala-Ala-Ala p nitrophenyl ester (Yoshinaka et al., 1984).
There has been relatively little work characterising fish metalloexopeptidases such as I carboxypeptidase A and Band aminopeptidases. Using synthetic substrates Martinez and Serra (1989) demonstrated relatively low activities of carboxypeptidases A and Bin crude extracts from the pyloric ceaca and intestine of anchovy. Cohen et al. (198 la) also observed low levels of carboxypeptidase B in the pancreas of carp. 31
1.4.2.2 Intestinal membrane bound proteases
Intestinal brush-border enzymes constitute a further biochemical barrier to the intestinal absorption of peptides and proteins (Section 1.2.1). Although this barrier has been well
'\ researched in mammals there has been minimal work on the isolation or characteristion of these enzymes in teleost species. It has been suggested that enzymes associated with teleosts intestinal cells are similar to those isolated from higher vertebrates, but this remains to be confirmed (Ash, 1985, Sire and Vernier, 1992) . • ,c
The activity of the brush-border membrane leucine aminopeptidase was found to be evenly distributed along the entire length of the intestine of cottids (Westem, 1971 ), catfish and carp (Fraisse et al., 1981). In perch a slight decrease in leucine aminopeptidase activity was observed near the rectum (Hirji and Courtney, 1982). In rainbow trout Di Constanzo et al. (1983) observed that the leucine aminopeptidase activities of the anterior brush-border membranes were 3.5-fold higher than in the posterior brush-border membranes.
:\
An alanine aminopeptidase was purified and characterised from the pyloric caeca of tuna (Hajjou and Le Gal, 1994). In contrast to alanine aminopeptideases of mammalian species the tuna aminopeptidase was not inhibited by puromycin and only partially inhibited by . y bestatin (Hajjou and Le Gal, 1994) . (
1.4.2.3 Intracellular digestion of proteins and peptides
Cytosolic peptidases are involved in the metabolism of internalised peptides and proteins in mammals (Section 1.2.1). However little work has been done on identifying cytosolic proteases of the intestinal enterocytes of fish. Hydrolytic activity towards three specific dipeptides, glycyl-leucine, glycyl-phenylalanine and phenylalanyl-glycine was •\ determined in the cytosol of trout enterocytes (Ash, 1980). They found that the N terminal glycyl dipeptides were predominantly hydrolysed by the cytosolic peptidases. 32
Rombout et al. (1985) and Georgopoulu et al. (1988) demonstrated the uptake of proteins by the posterior intestine of carp and trout, respectively. However in rainbow trout only 6% of the in vivo administration of horseradish peroxidase (HRP) escaped the lysosomal degradation of in the posterior enterocytes and reached systemic circulation '\
( (Georgopoulu et al., 1988). In trout the tracer proteins HRP and ferratin were localised in supranuclear vacuoles where acid phosphatase and cathepsin activities were observed (Georgopoulou et al., 1985, Georgopoulou et al., 1986). Antigenic sites of both cathepsin B and D indicated the presence of these capthepsins in the supranuclear vaculoar system, whereas cathepsin D was also found in the secondary lysosomes of the posterior intestinal
-'( enterocytes of trout (Georgopoulou et al., 1985, Georgopoulou et al., 1986). Cathepsin D
'> and cathepsin B have been shown to play an important role in the lysis of absorbed .\ proteins in mammals (Aronson and Barrett, 1978, Barth and Afting, 1984). The posterior intestine of the rainbow trout has characteristically high catheptic activity, which increases considerably after ingestion of food (Georgopoulou et al., 1985).
In carp absorbed HRP and ferratin are hydrolysed by the posterior enterocytes in different ( ways (Rombout et al., 1985). HRP is mainly released in the intercellular space where it \_, comes into contact with intra-epithelial lymphoid cells; some of the HRP is localised in secondary lysosomes of enterocytes. In contrast most of the ferratin appears to end up in
-( large supranuclear vacuoles where it is digested slowly. i
1.4.2.4 Biochemical barriers of the immunological system
During the transepithelial passage proteins and peptides can come in contact with the lymphoid tissue associated with both the enterocyte and underlying connective tissue. The gut associated lymphoid tissue (G.A...LT) includes lymphocytes, macrophages and eosinophilic granulocytes (EGCs) and has been described for goldfish, (Temkin and ,',\ Millan, 1986) channel catfish, (Krementz and Chapman, 1975) salmonids (Ezeasor and Stakoe, 1980) flounder (Davina et al., 1980) and carp (Rombout et al., 1989). 33
The posterior intestine ofteleosts has been shown to play a significant role in immunity. The absorbing enterocyte of the posterior intestine has been likened to the M cell of the mammalian intestinal epithelium with the existence of secretory immunoglobulin forming cells. Proteins absorbed by the posterior enterocytes of trout have been identified \
f in the macrophages (Rombout et al., 1985, Georgopoulou and Vernier, 1986).
Rombout et al. (1985) identified ferratin containing macrophages between the epithelial cells, in the lamina propria of the posterior intestinal epithelium, after oral intubation of ferratin to carp. It is generally accepted that macrophages are antigen-presenting cells in
-'( the immune system and degrade most of the phagocytosed antigen (V aresio et al., 1980, Rombout and van der Berg, 1989) .
. I \
EGCs are also able to internalise macromolecules absorbed by the posterior enterocytes. \ Le Bail et al. (1989) showed that the small amount of intact bovine growth hormone that absorbed across the enterocyte into the lamina propria was trapped in EGCs and the immune cells in trout yearling. Isolated EGCs from the lamina propria of posterior of rainbow trout were shown to endocytose FITC-coupled proteins in vitro (Dorin et al., J._ . 1993). Further in vitro work showed that EGCs only endocytose foreign (heterologous) proteins while homologous proteins remained associated with the plasma membrane and
.Y did not enter the cell (Dorin et al., 1993). Innnunoblots showed that the internalised I heterologous proteins undergo hydrolysis. The activity of cathepsin D increased significantly after endocytosis of the heterologous protein.
Since colonisation of the lamina propria with EGCs occurs during the first year of development in rainbow trout (Bergeron and Woodward, 1982), it was suggested that this innnunologica~ barrier could be avoided by the oral administration of peptides to \ yearlings (Le Bail et al., 1989). 34
1.4.3 Absorption of proteins and peptide
1.4.3.1 Peptide carrier mediated transport There are few examples of intestinal peptide transport systems that have been identified in teleosts. A proton gradient-energised dipeptide system was demonstrated in the intestinal brush border membrane vesicle (BBMV) of the eel (Verri et al., 1992). Two proton dependent peptide transport systems were identified in BBMVs of the intestine and pyloric ceaca of both tilapia and rockfish (Thamotharan et al., 1996). One of the transport systems was high affinity and saturable while the other was low affinity '< non-saturable.
1.4.3.2 Paracellular route
\ There is no direct evidence of absorption of peptides and proteins across the intestinal epithelia by the paracellular route. However, Schep et al. (1998) recently showed that the salmonid posterior intestine in vivo was permeable to the low molecular weight hydrophilic markers fluorescein and mannitol. Since peptides (Nellans, 1991) are thought
•, to follow the paracellular route as for fluorescein and mannitol (Ma et al., 1993, Sakai et '" al., 1997), Schep et al. (1998) speculated that peptides could also be absorbed across the salmonid posterior intestinal epithelia by the paracellular route.
·\ 1.4.3.3 Transcytosis A number of investigators have shown that orally administered proteins cross the intestinal epithelium into the systemic circulation intact or partially hydrolyzed (Watanabe, 1982, Georgopoulou et al., 1985, McLean and Ash, 1986, Suzuki et al., 1988). Studies have shown there to be regional differences in the protein absorption in the \ intestine of teleosts. The posterior section of most, if not all, teleosts appears to be specialised in the uptake of macromolecules (Noaillac-Depeyre and Gas, 1979, Stroband and van der Veen, 1981, Iida and Yamamoto, 1985, Rombout et al., 1985, Georgopoulou et al., 1986). Rombout et al. (1985) demonstrated that intact HRP and ferritin are 35
absorbed by the anterior and posterior intestine of carp, although uptake by the anterior intestine was relatively low.
In vivo oral administration of proteins to trout confirmed that protein absorption takes
I place in the posterior intestine (Georgopoulou et al., 1985). The rate of protein uptake was dependent on the size of the protein (Georgopoulou et al., 1985, Georgopoulou et al., 1986). HRP (MW 40 000) was visualised in mucosal epithelium2 h after oral administration whereas human IgG (MW 150 000) and ferritin (MW 460 000) were visualised 10 and 18 h after administr~tion. In rainbow trout, cells of the posterior section where shown to be structurally differentiated for the absorption of proteins and carrying out intracellular digestion (Georgopoulou and Vernier, 1986) like those of the ileum cells of suckling rat (Gonnella and Neutra, 1984).
In carp, HRP and ferritin were shown to be absorbed and processed by the enterocytes in different ways (Rombout et al., 1985). A high portion of absorbed HRP was transported in vesicles to branch endings of lamellar infoldings of the basal cell membrane . .( Consequently, most of the HRP was released into the intercellular space where it came in ' ' contact with the intra-epithelial lymphoid cells. Small amounts of absorbed HRP were also was found in the secondary lysosomes of the enterocytes. Rombout et al. (1985)
-( suggested that HRP is transported across the trout intestinal epithelium by the mechanism described for the new-born-rat intestine (Abrahamson and Rodewald, 1981). HRP bound to receptors on the apical cell membrane and is transferred into the cell in clathrin coated vesicles. The clathrin coat protects these vesicles from fusion with lysosomes. I ,\ Preliminary studies by (Rombout et al., 1985) show that the portion ofHRP found in the supranuclear vacuoles decreases with age. Ferritin in contrast was taken up in a nonselective manner. Ferritin was taken up by fluid phase endocytosis and is accumulated '\ in small vesicles or vacuoles that fuse with lysosomal-like bodies (Rombout et al., 1985). 36
1.4.3.4 P-glycoprotein efflux p~mp in teleosts
Work in mammalian and intestinal epithelial models suggest that physiological role of the P-glycoprotein efflux system is to restrict transcellular flux of some molecules, including peptides (Section 1.2.3). Southern blot analysis of a winter flounder genomic library indicates that the teleost possess two P-glycoprotein genes. Sequence analysis of two key areas of the teleost genes indicate more than 70% sequence identity with mammalian genes at the amino acid level (Chan et al., 1992).
The presence of P-glycoprotein in the intestine ofteleosts was supported by the immunoreaction between mammalian polyclonal and monoclonal antibodies against ;;,.. P-glycoprotein and the lumenal surface of intestinal epithelium of guppy (Hemmer et al., 1995). This suggests that the P-glycoprotein efflux system could play a role in the poor
>, absorption of peptides across the intestinal epithelia of teleosts.
1.4.4 Oral peptide and protein delivery
i Early attempts to orally deliver proteins and peptides to salmon generally involved the intubation of crude pituitary homogenates that contain various hormones whose absorption could only be monitored indirectly by their biological response (Pickford and "l Atz; 1957). Suzuki et al. (1988a) reported that the oral intubation of crude pituitary extracts to goldfish induced spermiation and ovulation in goldfish, indicating the absorption of physiologically active salmon gonadotropin. The absorption of intact
;/ gonadotropin was further supported by immunological techniques.
More recent technological advances in peptide and protein biotech.riology have enabled the large scale production of synthetic-or recombinant hormones and their superactive -\ analogues for drug delivery.
Adrenocorticotropic hormone (ACTH), a single chain linear peptide containing 39 amino acids (MW 4 541) was detected in the plasma of yearling salmon by RIA after oral 37
intubation (McLean et al., 1990). The peak plasma concentration of ACTH was observed 120 min after administration, but the amount of absorbed intact peptide was very low.
A number of investigators have orally administered superactive analogues of LHRH to induce ovulation and spawning in fish. Thomas and Boyd (1989) showed that oral
10 6 ,( administration of the superactive analogue ( des-Gly -[D-Ala ]-hLHRH (1-9) ethylamide (hLHRHa) in dead shrimp induced ovulation in spotted seatrout. They estimated that an
') oral administration of 1.0-2.5 mg/kg of hLHRHa would be sufficient to induce ovulation in spotted seatrout, in comparison to an intramuscular injection of only 100 µg/kg body
'J for the same species (Thomas and Boyd, 1988).
Orally induced ovulation of stablefish by hLHRHa required two administrations totaling 1.5 mg/kg (Solar et al., 1990). However hLHRHa was intubated with sodium bicarbonate 'j and L-lysophopsphatidylcholine to elevate the gastric pH and enhance the permeability of the intestine, respectively.
-\ McLean et al. (1991) compared directly the oral absorption ofhLHRH and hLHRHa in 17B-oestradiol-primed juvenile salmon over a 360 min period. The plasma profiles after oral intubation indicate that the superactive hLHRH analogue is absorbed more rapidly
··{ and in greater quantities (100-fold) than the natural hLHRH hormone. The absorption of intact LHRH and hLHRHa was further supported by a significant increase in the plasma concentration of gonadotropin.
r/ Oral administration of the superactive analogue (des-Gly1°-[D-Ala6]-salmon LHRH (1-9) ethylamide (sLHRHa) of only 50 µg/kg was required to induce spawning in Thai carp (Sukumasavin et al., 1992). This dose is considerably lower than for 17B-oestradiol primedjuvenile salmon (2-20 mg/kg) (McLean et al., 1991), spotted seatrout (1 mg/kg) (Thomasand Boyd, 1989) and stable fish which required a priming and a secondary dose totaling 1.5 mg/kg (Solar et al., 1990). Unlike these species, since Thai carp do not have stomach (agastric) it was suggested that the absence of a stomach reduced the metabolic 38
barrier of the gut, therefore increased the opportunity for peptide absorption across the gut.
To increase the oral absorption of intact proteins and peptides a number of investigators have used various formulation strategies. The earliest attempt to formulate was by delivering salmon pituitary extracts in polyacrylic acid (PAA) gel bases to goldfish by oral intubation (Suzuki et al., 1988b). However there was no improvement in the oral bioavailability of gonadotropin in goldfish unlike that described for mammals (Section 1.3.1.5).
Immunological and biological cell assays demonstrated the absorption of intact
' \ recombinant hi.nnan growth hormone (hGH, MW 22 000) after intubation to carp (Hertz et al., 1991). Intubation of the hormone in the presence of 50 mM deoxycholic acid (DOC) resulted in up to a 1000-fold increase in serum levels of the hormone. The peak plasma concentration of bioactive hGH was 30 min after oral intubation both in the presence and absence of DOC. Because of the large increase in the absorption ofhGH in the presence of DOC the authors suggest that DOC acts as a permeation enhancer rather then a protease inhibitor. Carp. Hertz et al. (1991) showed there was a difference in absorption ofhGH in the presence of DOC between fed fish and fish fasted for 12 days.
·( For the fed fish, it was estimated that at least 2% of intubated hGH reached the systemic · circulation while for fish fasted only 0.5% reached systemic circulation.
When antacid (sodium bicarbonate) and permeation enhancer (Mega-9) were included with recombinant porcine growth hormone (rpGH) in the feed of diploid and triploid salmon the growth enhancement by rpGH was potentiated (McLean et al., 1993). After sixteen weeks the weight of the triploid fish fed the prGH-only diet increased by 29% compared to control, while the weight of the fish fed prGH in the presence of excipients increased 60%. The weight of the diploid fish increased 17 and 50%, on the respective diets. 39
Interestingly, when the protein HRP was intubated with SBTI (1 mg/kg) and Mega-9 (1 % w/v) there was an enhanced uptake of the protein into tissues of the liver and spleen (McLean and Ash, 1990). However this cocktail did not increase the plasma concentration ofHRP, suggesting that the absorbed protein was delivered to the lymphatic system (Section -1.4.2.4).
More recently micoencapsulation has been used to protect peptide drugs from the lumenal proteolytic activity of the intestine and the low pH of the stomach in teleosts.
-, Recombinant chum salmon growth hormone (rsGH) entrapped in an enteric polymer matrix ofhydroxypropylmethyl cellulose phthalate (GH-EPM) improved the oral absorption of intact rsGH into the systemic circulation of juvenile rainbow trout
I \ (Moriyama et al., 1993). The higher absorption of the microencapsulated rsGH was
' , believed to be due to the peptide being protected from the acidity and proteolytic activity of the stomach. However, this formulation may also allow rsGH to reach the posterior intestine where protein and peptide absorption is high (Georgopoulou and Vernier, 1986).
\
6 9 A salmon LHRH analogue [D-Ala - Pro NET] LHRH (sLHRHa) formulation induced ovulation in carp, catfish and trout (Breton et al., 1995). In vivo experiments showed that permeation eru1iancer (Tween 80 and oleic acid) and protease inhibitors EDTA and
-i ) eggwhite trypsin inhibitor enhanced absorption of sLHRHa in the agastric species carp
·( (Breton et al., 1995). To bypass the stomach of the gastric species catfish and trout, sLHRH-a and excipients were entrapped in a silica gel coated with cellulose-acetate phtalate to form microcapsules. Successful ovulation of rainbow trout was achieved with :... ~, a dose of 90 µg/kg (Breton et al., 1995).
This introduction reviewed the pancreatic enzymes isolated from teleosts. In comparison to the mammalian pancreatic enzymes the teleosts pancreatic enzymes had different; catalytic efficiencies, requirements for divalent ions and activity towards protease inhibitors. A number of protease inhibitors have been used to increase the oral bioavailability of peptides and proteins in mammalian species; however since the 40
pancreatic enzymes of teleosts differ to mammalian enzymes it is not known how effective these protease inhibitors are at reducing the lumenal proteolytic activity in teleosts.
In comparison to lumenal proteases there has been minimal work on characterising the membrane-bound proteases of the intestinal tract of teleosts. Since the proteolytic activity of the membrane-bound proteases can.account for upto 90% of the peptide metabolism in mammalian species (Banga, 1995), investigations must focus on these enzymes and ways to inhibit their activity in teleosts. i
A number of researchers have targeted the teleost posterior intestine for peptide delivery as it is expected to have lower levels of pancreatic enzymes; however no researchers have characterised the regional metabolic barriers of the intestine towards peptide oral delivery. A systematic study of the regional metabolic barriers of the intestinal lumen and membrane-peptidases would allow for the development of formulations which would target the region of the gut with the most favorable metabolic environment for the absorption of peptides.
1.5 Aims of the study f ·" The aim of the thesis was to investigate the metabolic barriers of the lower intestinal tract ( of the salmon to the oral absorption of peptides and protein. The thesis investigates the c> specificity, potency and mechanism of action of protease inhibitors towards intestinal j proteases isolated from salmon. This thesis also investigates the effectiveness of several protease inhibitors in improving of metabolic stability of the human and salmon LHRH analogues in tissue homogenates extracted from the intestine of salmon. 41
I Chapter 2: Capillary Electrophoretic Assay of Luteinizing Hormone Releasing Hormones
" !
>- ;1
-( ·. ·,
I \
-,
\..
\
-.\
'I
1--<
·( t t } '·
'.\ 42
2.1 Introduction
•\ Luteinizing Hormone-Releasing Hormones (LHRH) is a neuropeptide that links the brain
with the pituitary gland, conveying all reproductive-related information from the central
nervous system to the endocrine system in animals. LHRH analogues have been used for
the induction of ovulation and spawning in several fish species (Donaldson and Hunter,
1983). Human LHRH (hLHRH) is composed of 10 amino acids, pyroGlu-His-Trp-Ser ;1 Tyr-Gly-Leu-Arg-Pro-Gly-NH2, while salmon LHRH (sLHRH) differs by two amino
7 8 substitutions Trp and Leu . The aim of this section is to report the development of a
, quick, simple capillary electrophoresis (CE) assay to measure the metabolic stability of \ human and salmon LHRH analogues in various intestinal homogenates from salmon.
CE offers a number of advantages when compared to the well established and widely
used technique of high-performance liquid chromatography (HPLC). Principle
advantages are likely to include high separation efficiency, shorter analysis times, , \ cheaper running costs, low sample consumption, robustness and simplicity (Reiger,
1992). There is a considerable literature on the use ofHPLC to study the metabolic
-< _) stability ofLHRH and analogues (Goren et al., 1990, Han et al., 1995, Peter et al., 1996). ·( However analysis by HPLC is time consuming, as separation requires gradient elution
followed by re-equilibration of the column. For CE, the time between elution of the
~ , analyte and analysis of the next sample is much shorter. Once the analyte of interest has
eluted the remaining analytes can be flushed from the capillary by a short run buffer wash
(2 min) ready for the next analysis. Another major benefit of CE is the reduced costs of
analysis. HPLC analysis requires the purchase of expensive columns, organic solvents
and the disposal of organic solvents. The cost of CE silica capillaries is a fraction of
reverse phase HPLC columns and CE analysis requires small volumes of the electrolyte
water based buffers, typically 100-250 ml per day. However to the best of our knowledge 43
no specific investigation has been carried out to determine the usefulness of CE for the
analysis and separation of LHRH analogues from its proteolytic metabolites.
CE is a relatively new technique that uses electrical forces to drive separation in capillary
tubes (Jorgenson and Lukacs, 1983). The essential components of a CE apparatus are
illustrated in Figure 2.1. CE system consists of a buffer-filled capillary placed between
two buffer reservoirs with a potential field applied across the capillary. The separation in
. capillary electrophoresis is based on the different migration velocities of analytes in the
electrophoretic buffer under applied potential.
. \
·( High Voltage
Capillary
\
"(
·l Detector
' ' Source vial Destination vial
Figure 2.1 Schematic of the capillary electrophoresis apparatus. 44
The riet mobility of an analyte depends on its electrophoretic motility ( charge to mass
ratio) and the electro-osmotic flow (EOF) associated with the buffer system. EOF is the
•\ bulk flow of liquid in the capillary (Figure 2.2).
Static layer [ (j) cb cb (j) d5 c5 d5 d5 d5 d5 d5 ( (Stem) EB EB EB EB EB_@ EB EB EB EB EB CD CD CD OT CD CD CD f7l EB EB cr;, EB (±1 u\ Mobile layer ""cr;, q:,cr;, CD EB CD ~ 8-. \J:.J (Outer Helmoltz) EB EB EB EB EEl . . Anode EB EB EB Cathode CD CD +. CD CD CD CD EEJ @EB' EB EB EB -EEJ EB EEJ EB EB EEJ Capillary wall CD ~CD CD cp cp cp cp cp q) cp
1 . ' Figure 2.2 Schematic of electro-osmotic flow.
·(
The bulk movement of solvent is caused by a small Zeta potential at the shear plane
between static and mobile phases. The Zeta potential is a result of ionised silanol groups
(pKa of 6-7) on the surface of the capillary and cationic counterions which are attracted
to the negatively charged surface to form an electrical double layer. This Stem layer of
the double layer is static and adjacent to the Stem layer are the diffuse or mobile cations.
{ This region is known as the Outer Helmoltz plane. Under an electrical field the cations of -r the diffuse layer migrate towards the cathode. The water molecules solvating the cations
also move causing a net solution flow along the capillary. This process is called the EOF
"' and enables all molecules ( cationic, neutral and anionic) to move along the capillary
towards the detector.
The magnitude of the EOF can be expressed by equation (2.1).
µEOF = (E Qri) (2.1)
where µEOF = the movement of molecules due to the EOF, E = dielectric constant of the
buffer, s = Zeta potential and ri = viscosity of the buffer. It obvious from equation 2.1 45
that altering any of these parameters will result in changes in EOF. Since Zeta potential is
essentially determined by surface charge of the wall and surface charge of the capillary is
dependent on the pH of the buffer, the magnitude ofEOF varies with pH of the run
buffer. Typically for fused silica capillaries the EOF increases with increasing pH above
,, pH 4.0 (McLaughlin et al., 1992). The Zeta potential is also dependent on the ionic
strength of the buffer, as described by double layer theory (McLaughlin et al., 1992).
Increased ionic strength results in double-layer compression, decreased Zeta potential, '( -~ and reduced EOF. Under conditions ofEOF, the order of elution will follow: cations of t '(' the lowest mass / charge ratio > cations of a higher mass / charge ratio > neutral species > anions with higher mass/ charge ratio > anions of a lower mass / charge ratio. A 1 ' f significant feature ofEOF is that instead of parabolic flow profiles as in pressure driven
< ( HPLC systems, the driving force originates from the wall of the capillary producing a
plug like flow that has a flat velocity. This increases the resolution of separations by
.', reducing the band broadening of the analyte peak during migration along the capillary !, (Huang et al., 1989). The theory for electrophoretic separations has been discussed and l. reviewed more extensively by others (Jorgenson and Lukacs, 1983, McLaughlin et al.,
1992). { >
\ CE has now been applied in many areas of peptide analysis (Landers, 1993) and has been
shown to be a powerful tool in resolving peptides (Grossman et al., 1988, McCormick, ,, ·, 1988, Grossman et al., 1989, Camilleri et al., 1991). Although CE has been used for
peptide analysis it has been used rarely to measure the kinetic stability of peptides in the
,'. presence of proteases. This is in part due to problems associated with the analysis of
peptides in aqueous solutions containing proteins and electrolytes. Proteins and peptides
can interact with the negatively charged surface of the capillary wall, resulting in peak
broadening and changes in the EOF (Hjerten, 1985). Changes to the EOF have been
shown to dramatically reduce the reproducibility of an assay system (Hjerten, 1985). 46
Electrolytes present in the sample can result in uneven migration and zone spreading due
to poor sample stacking (Weinberg et al., 1990, Satow et al., 1991). Sample stacking
causes the sample to be concentrated into a smaller zone which then migrates along the
capillary under the electrophoresis conditions of the run buffer (Burgi and Chein, 1991).
Sample stacking requires significantly lower concentrations of electrolytes in the sample
relative to the run buffer. Under such conditions when an electrical field is applied a
proportionally greater field will develop across the sample zone causing the ions to . ( migrate faster. Once the ions reach the run buffer boundary the electrical field decreases
and they migrate slower, thereby concentrating the sample into a small zone (Burgi and
Chein, 1991). 1 '
There are various approaches by which these difficulties may be overcome. High ionic
strength buffers suppress the ion-exchange effects between analytes and the capillary
surface (Green and Jorgenson, 1989) and also increase the conductivity difference
' between the sample buffer and the capillary run buffer resulting in more efficient sample
stacking (McLaughlin et al., 1992). The cost of using high ionic strength buffers is longer
analysis times due to suppressed EOF and Joule heating (Jorgenson and Lukacs, 1983).
~ I\., Joule heating causes a temperature gradient within the capillary, creating greater ' electrophoretic mobilities at the center of the capillary resulting in non-symmetrical
',c peaks, lower theoretical plates and reduced reproducibilities. Joule heating can be
' . reduced in four ways. Firstly, by performing electrophoresis at lower field strengths at the
cost of increased elution time of the analyte. Secondly, capillaries with smaller inner
diameter can be used as their high surface to volume ratio means heat is dissipated more
efficiently. Capillaries with smaller inner diameter have a shorter path length for
detection and therefore decrease the sensitivity of the assay. Thirdly, increasing column
length provides greater electrical resistance to lower Joule heating under constant voltage.
The cost of longer capillaries is longer analysis times as elution is proportional to the 47
square of the column length. Longer capillaries also result in reduced peak efficiency, as
there is greater opportunity for analyte-capillary wall interaction. Finally, Joule heating
can be reduced by cooling the outer surface of the capillary by high velocity air or cooled
liquid. :~
)
It was shown that capillary surface interactions can also be reduced by using
electrophoteric buffers at low pH (2-3) (Grossman et al., 1988, Bergman et al., 1991). At , . ( low pH the silanol groups of the fused silica are essentially protonated and uncharged,
thereby reducing the ionic interaction between the positively charged peptides and
capillary. Again the cost is the reduction in EOF which increases the analysis time. !,
)"
Accordingly, in this study the sample and buffer ion concentration and buffer pH were
varied. As sensitivity was not an issue in these in vitro metabolic studies, capillaries of
> \ 50 µm I.D. were used to minimise Joule heating. Further, to reduce Joule heating, l capillaries were cooled with recirculating water maintained at 20°C. To reduce the .J... analysis time, the minimum recommended length of capillary was used ( effective length
,! of 60 cm). The effect of sample stacking and pH of the run buffer on the LHRH
proteolytic digests were examined with respect to the reproducibility of peak heights,
peak area, migration times and selectivity. This thesis reports a CE assay that is relatively
~ ) simple and quick and also reports intra- and inter- day variability for analysis of both
human and salmon LHRH.
,,'
,l 48
2.2 Materials
Human and salmon luteinizing hormone-releasing hormones were purchased from '\
) ( Bachem (California, USA). All other chemicals were analytical grade purchased from either Ajax chemicals Pty. Ltd. (Auburn, NSW, Australia) or BDH chemicals Ltd. (Poole, I } England) unless otherwise specified. Deionised distilled water was produced with a Millipore Milli-Q Reagent Water System (Continental Water Systems, USA). Micro vials C.E.S. (0.5 ml) were purchased from Dionex Corporation (California, USA). Fused . ( silica capillary tubing (50 µm I.D. with a 360 µm 0.D.) was purchased from Supelco Inc. (Bellefonte, PA, USA). Dimethylchlorosilane (Lot 589198) was obtained from Pierce I · 1 (Rockford, IL, USA). Light mineral oil with a specific gravity of 0.875 - 0.885 at 77°F and a.-chymotrypsin (EC 3 .4.21.1, TLCK treated Type VII) were purchased from Sigma " Chemical Co. (St Louis, MO, USA). Nylon filter membranes (0.45 µm) were purchased from Alltech (Auckland, New Zealand). Sample tubes of 0.5 ml and 1.5 ml were purchased from Eppendorf (EppendorfNetheler Hinz GmbH Hamburg, Germany).
) \ J. \ 2.3 Methods
,-! I 2.3.1 Instrumentation and standard conditions The capillary electrophoresis system was an Dionex CES I (Dionex Corporation, _,"( Sunnyvale, CA, USA) interfaced to a computerised data handling system (Dionex AI-450 Chromatography Automated Software v 3.3.2). Data were collected at the rate of 5 Hz l '/ and electrophoresis was carried out towards the cathode. Neslab RTE-111 (NESLAB Instruments, Inc. Portsmouth, USA) recirculating bath maintained the temperature of the water in the capillary cooling jacket at 20 ± 0.1 °C. The cooling jacket covered 33 cm of i ~ \ the capillary. Sample vials containing the capillary inlet were elevated to 150 mm higher than the destination capillary outlet th~n lowered to the height of the destination outlet; the loading process took 60 s to complete. Unless otherwise specified, electrophoresis was carried out at constant current by ramping the current from O to 60 µA over 1 min 49
and holding the current at 60 µA for the remainder of the run. The peptides were detected by a UV detector set at 215 nm.
2.3.2 Capillaries ,, Capillary preparation and installation was according to that described in the Dionex User's Manual. Polyimide-coated fused silica capillary tubing (50 µm I.D. with a 360 µm O.D.) was cut to a length of 60 cm. Approximately 5 cm from the destination outlet of the capillary a narrow segment of the outer coating was burnt and cleaned with isopropyl ' ( alcohol to make a window for the UV-Vis detector. The capillary was inserted into the capillary water cooling jacket which was connected to a water recirculating bath, N eslab
(RTE-111). New capillaries were reactivated by flushing with 0.1 M H3P04, followed by filling the capillary with 0.5 M NaOH and waiting for 10 min before flushing with ;I deionised water and run buffer. At the start of each day the installed capillary and the source and destination reservoirs were washed by with 0.1 M NaOH, deionised water and the run buffer. The rinse time at 6 psi for the capillary was 120 s, the rinse and refill times for the source and destination vial was 6 s. Prior to each run the capillary, source and .\ destination reservoirs were rinsed with the run buffer as described above.
2.3.3 Buffers
Phosphate buffer at pH 2.5 was prepared by diluting a 0.1 M stock of NaH2P04 10-fold in deionised water containing ionic strength adjuster sodium chloride. The pH was adjusted with phosphoric acid. Phosphate buffers at pH 6.5 and 7.5 were prepared by ;, diluting 0.1 M stocks ofNaH2P04 and NaH2P04 10-fold in deionised water containing sodium chloride. The diluted stocks were mixed to give the desired pH. Sodium acetate buffer at pH 4.0 was prepared by diluting a 0.1 M stock of sodium acetate 10-fold in deionised water containing variable concentrations of sodium chloride. Glacial acetic acid was added to adjust the pH to 4.0. Buffers were filtered through 0.45 µm membrane filters and degassed before use. 50
2.3.4 LHRH digests ), Chymotryptic digests of human and salmon LHRH were prepared by incubating LHRH analogues (0.2mM) with chymotrypsin (2.5 µg of protein in 250 µl) in 50mM sodium
phosphate buffer pH 8.0 at 10°C. Digestion of human and salmon LHRH was stopped by addition of equal volumes of 0.2 M HCI after 5 and 10 min, respectively. Samples were then centrifuged at 20 000 g for 5 min at room temperature to remove particluates. Samples (50 µ1) of the supernatant were transferred to 0.5 ml sample tubes and stored at ,,, -20°C.
2.3.5 Silanisation Dionex sample vials were placed in a peaker inside a desiccator and a 5% (v/v) solution -( of dimethylchlorosilane in heptane was added to the beaker, placed in the desiccator. The desiccator was evacuated until the dimethylchlorosilane had evaporated (1-2 h). The sample vials were rinsed once in methanol and then several washes with deionised water. ti, The sample vials were dried in an oven at 60°C.
\ 2.3.6 hLHRH standards - in the presence of a-chymotrypsin Human LHRH stock solution (1.2 mg/ml) was diluted with 50 mM sodium phosphate buffer pH 8.0. Standards (20 µ1) were chilled on ice and diluted with an equal volume of ice cold chymotrypsin (20 µg/ml) in 50 mM sodium phosphate buffer pH 8.0 and
0 immediately diluted 1: 1 with 0.2 M HCI and stored at -20 C for five days. Prior to analysis standards were centrifuged at 20 000 g for 5 min and 20 µ1 of the supernatant transferred into a silanised Dionex micro-vial containing 60 µI of deionised water and mixed. The solutions were immediately covered with approximately 1 mm layer of light mineral oil to prevent evaporation and analysed by capillary electrophoresis. 51
2.3. 7 hLHRH standards - variability studies ,, Triplicate standard curve curves for solutions of hLHRH and sLHRH in 50 rnM sodium phosphate buffer pH 8.0 analysed by CE were plotted. For hLHRH three independently prepared stock solutions (1.5, 1.6, 1.6 mg/ml) were diluted with 50 rnM sodium phosphate buffer pH 8.0 to give final concentrations of60.8, 81, 121.4 and 242.7 µg/ml. Three independently prepared stock solutions of sLHRH (1.3, 1.5, 1.6 mg/ml) were prepared and diluted in 50 rnM sodium phosphate buffer pH 8.0 to give final concentrations of 75.2, 150.4, 180.8, 240 and 304 µg/ml. LHRH standards were then '\ diluted with equal volume of 02 M HCl to give a total volume of 40 µ1 and centrifuged at ·.I
'( 20 000 g for 5 min. Supematants (20 µl) were placed in silanised Dionex sample vials containing 60 µ1 of deionised water and mixed. The solutions were immediately covered , '\ with approximately 1 mm layer light mineral oil to prevent evaporation and analysed by capillary electrophoresis. Linear regression analysis was carried out using the computer program Minitab (MinitablOXtra, Minitab Inc. PA, USA). The intraday variation of the assay for hLHRH at 15 µg/ml and was determined by injecting each of the three independently prepared standards six times. Interday variation was also determined at 15 µg/ml by injecting one of the standard solutions a further six times on two subsequent days.
-/
,, 52
2.4 Results .,,
2.4.1 Precision The precision of peak areas and peak heights were determined. LHRHs at lmg/ml in deionised water were analysed by capillary electrophoresis in 50 mM sodium phosphate i' buffer at pH 2.5 (Section 2.3.1) (Figure 2.3). A constant voltage of 15 kV gave an approximate current of 60 ± 4 µA. Relative standard deviations (RSD%) for migration times, peak heights, and areas for both analogues from six consecutive runs are shown in Table 2.1. The.RSD for migration times are low(< 1.0%) however the RSD for peak '.I
·~ height and peak area are relatively high. The reason lies in a significant trend which can be recognised by looking at the percentage increase in peak height and peak area between the first and last injection. The percentage increase in peak height and area suggested
I\ I sample evaporation.
Table 2.1 Reproducibility of peak migration time (min), peak height (mV) and area (mV) \. for hLHRH and sLHRH. ' .\ I, hLHRH sLHRH Injection Migration Peak Peak Migration Peak Peak I number time (min) height area time (min) height area I ,, 1 21.43 3800 49830 35.35 5250 11744 i 2 21.65 3890 51000 35.20 5356 11905 3 21.56 3999 52400 35.48 5447 12344 4 21.67 4235 55545 35.93 5556 12410 j 5 21.80 4356 57150 35.80 5888 13022 6 21.75 4378 57495 35.69 5978 13560 Mean 21.64 4110 53903 35.58 5581 12498 s.d. 0.13 247 3269 0.28 292 686 RSD (%) 0.62 6.0 6.1 0.79 5.2 5.5 Percentage (+) 14 (+) 14 (+) 13 (+) 15 ·, change 53
.·,
(A) 0.06 0.05 1·· 0.04 ,, 0.03 0.02 -.y ,-._ ~ 0.01 l j '-'-< 1.) 0.00 Cll A 0 -0.01 0.. Cll 0 10 20 30 40 1.) .. '' ~ I-< (B) ....0 0.06 ~ I u ~ 1.) ! ) ....1.) 0.05 Cl 0.04 " 0.03 ,\ 0.02
.\ 0.01 0.00 -0.01 0 10 20 30 40
Times (min)
I, Figure 2.3 Capillary electrophoretic separation of hLHRH (A) and sLHRH (B) (0.1 mg/ml in deionised water) in 50 mM sodium phosphate buffer pH f5 (Section 2.3.1) ·"
To prevent evaporation the samples were covered with a film of approximately 1 mm light mineral oil. Table 2.2 shows that for samples covered with a 1 mm layer of mineral there is a decrease in peak height and area with consecutive injections. The light mineral oil had no effect on peak shape or migration time. The percentage decrease in peak height and area suggested that hLHRH is being adsorbed to the wall of the sample vial overtime. 54
Table 2.2 Effect of light mineral oil on reproducibility and percentage change in peak height (mV) and area (mV) for hLHRH and sLHRH. hLHRH sLHRH Injection Migration Peak Peak area Migration Peak Peak area number time (min) height time (min) height 1 . 21.45 3865 50685 35.35 5456 12687 2 21.70 3976 50566 35.20 5120 11906 3 21.57 3700 49018 35.48 4890 10895 4 21.67 3706 46789 35.93 4789 10990 5 21.87 3600 46778 35.80 4705 10987 6 21.73 3479 45759 35.69 4678 10870 Mean 21.64 3721 48266 35.58 4940 . 11390 sd 0.13 179 2116 0.28 300 746 RSD (%) 0.62 4.8 4.4 0.79 6.1 6.6 Percentage (-) 10 (-) 9.7 (-) 14.3 (-) 14.3 change
It has been documented previously that LHRH analogues at low concentrations adhere to glass and other surfaces (Anik and Hwang, 1983, Mistry; 1993). Dionex sample vials were silanised (Section 2.3.5) for the analysis of hLHRH and sLHRH (0.1 mg/ml in
.,I deionised water). Table 2.3 shows that the percentage decrease of both analogues over the analysis was reduced by approximately half when sample vials were silanised. All subsequent experiments used silanised Dionex sample vials.
l•
" 55
Table 2.3 Effect of silanisation of sample vials on the reproducibility of and percentage change in peak height and area for hLHRH and sLHRH hLHRH sLHRH Injection Migration Peak Peak area Migration Peak Peak area number time (min) height time (min) height 1 21.43 3865 50685 35.78 5301 118404 2 21.65 3940 54179 35.45 5243 118063 3 21.56 3854 51100 35.48 5025 117121 4 21.67 3865 50685 35.60 4985 111336 5 21.80 3800 48833 35.55 4944 110340 '< 6 21.75 3689 48378 35.65 4982 110221 Mean 21.64 3836 50643 35.60 5080 114248 sd 0.13 84.5 2055 0.12 152 4001 RSD (%) 0.62 2.2 4.1 0.34 3.0 3.5 Percentage (-) 4.6 (-) 4.6 (-) 6.0 (-) 6.9 change
Electrophoretic mobility of analytes is dependent upon the field strength. High voltages are often applied in CE; however this can generate Joule heat causing the analyte zone to be dispersed. The maximum recommended current before Joule heating of the run buffer .l occurs is reported to be 60 µA (Dionex, 1989). Electrophoresis can be carried out at either constant voltage or constant current. To determine which gave the better reproducibility, human LHRH (0.1 mg/ml in deionised water) was injected in twelve r,- consecutive runs alternating between electrophoresis at constant voltage of 15 kV and constant current of 60 µA using 50 mM sodium phosphate buffer at pH 2.5 (Section 2.3.1). There were no significant differences in peak heights (p > 0.4) or peak areas (p > 0.1) between samples electrophoresed at constant current and constant voltage (Table 2.4). In order to maximise the applied field without Joule heating of the run buffer, all subsequent ·experiments were performed at constant current of 60 µA.
r, 56
Table 2.4 Effect of electrophoresis under constant voltage and constant current on reproducibility of CE assay for hLHRH (n=3). Constant voltage Constant current MT PH PA MT PH PA Mean 22.36 1071 13898 22.02 1092 14511 s.d. 0.22 28.87 499 0.17 27.5 432 RSD (%) 0.98 2.7 3.6 0.77 2.5 2;98 MT = migration time of the analyte, PH = peak height, PA = peak area
It was found consistently that the RSD of peak height was less than the RSD of peak area during the analysis ofLHRH by CE (Tables 2.1 - 2.4). We concluded that under our electrophoresis conditions peak height gives better precision than peak area. In the subsequent experiments peak heights only were used.
2.4.2 Optimisation of sample stacking Chymotryptic digests ofhLHRH and sLHRH (Section 2.3.4) were used to investigate the effect of sampie stacking on resolution and reproducibility of the CE assay. The effect of sample electrolyte concentration on stacking was determined by diluting human and salmon LHRH chymotryptic digests 2.5-, 4- and 6-fold in deionised water (Figures 2.4 & 2.5). For both hLHRH and sLHRH lower sample dilution (or high electrolyte concentration) reduced sample stacking and decreased resolution. High ionic strength of the run buffer improved the reproducibility of the CE assay by improving stacking and reducing interaction between analyte and wall. Figure 2.6 summarises the effect of dilution of the digests with deionised water and the run buffer ionic strength on peak height and area. Diluting the digests resulted in reduction of both peak height and area; however the reduction in peak area was greater than for peak height. In other words, diluting the digests with deionised water resulted in improved stacking which can be seen by the greater peak height/peak area ratios (Figure 2.7). Stacking was the greatest for digests electrophoresed in buffer containing 75 mM sodium chloride. Run buffer containing 25 :tpM sodium chloride concentration was insufficient to stack the analyte bands for elution as symmetrical peaks ( data not shown). 57
0.002 (A) hLHRH• 0.001
0.000 0 5 10 15 20 25
0.002., (B) ,-... :::i $ ii) IZI l=:i 0 0.001 0.. IZI ii) ~ 1-s .....0 () ii) t 0.000 Q \ 0 5 10 15 20 25
0.002 ~ (C)
,) 0.001
0.000 +----,-----,-----,-----,------, > 0 5 10 15 20 25
Time (min)
Figure 2.4 Capillary electrophoretic separation of the chymotryptic hLHRH digest in
10 mM sodium phosphate buffer pH 2.5 containing 7 5 mM sodium chloride (Section
2.3.1). Digests were diluted 2.5-fold (A), 4.0-fold (B) and 6-fold (C) in deionised water. 58
0.005 (A)
0.004
0.003
0.002 sLHRH
0.001
".•v--.;• 0.000 k ; I ~~I I 0 5 10 15 20 25 30 35
,-... 0.005 (B) ~ 0.004 11) .::Ul 0 0.003 ~ 0.002 ~ l-s 0.001 ~ t Q.000 l . j 0 '' -0.001 +-----.---,------.------,,---~------. 0 5 10 15 20 25 30 35
0.005 (C) 0.004
.; 0.003 0.002 · 0.001 0.000 -1------..J
-0.001 +------T---r---r-----.-----.------,----, 0 5 10 15 20 25 30 35
Time (min) Figure 2.5 Capillary electrophoretic separation of the chymotryptic sLHRH digest in
10 mM sodium phosphate buffer pH 2.5 containing 75 mM sodium chloride (Section
2.3.1). Digests were diluted 2.5-fold (A), 4.0-fold (B) and 6-fold (C) in deionised water. 59
140T T 2000 (A) 120 1500 ~ 100 I:)£) ti:! ..... (!) (!) ,..q 80 ta 1000 ~ ~ (!) (!) 60 ~ ~ ~ -> [o ::t I I I 2.5 4 5 6
SOOT .At ... (B) T 10000
400 8000 ~ ~~ I:)£) ti:! f::s. (!) ..... 300 6000 ~ ~ ta ~ ~ '! ~ 200 4000 (!) ~ ~ 100 2000
0 0 5 6 \ 2.5 4 Dilution of LHRH chymotryptic digests
Figure 2.6 Effect of diluting hLHRH (A) and sLHRH (B) chymotryptic digest in deionised water on the peak heights (triangle) and peak area (diamonds) when -) electrophoresed (Section 2.3.1) in 10 mM sodium phosphate buffer at pH 2.5 containing 50 mM sodium chloride (open symbols) and 75 mM sodium chloride (closed symbols). Each point represents the mean of three experiments. 60
as 0.18 I-< c;j • •... ~ • ... 8, 0.12 •...... E-- D bi) D ..... D /::,. /::,. D /::,. ] 0.06 /::,. ~ c;j 0) 11.. 0 2 3 4 5 6 Sample dilution
Figure 2.7 Effect of diluting hLHRH (closed symbols) and sLHRH (open symbols) chymotryptic digests in deionised water on sample stacking (peak height/peak area) when electrophoresed (Section 2.3 .1) in 10 mM sodium phosphate buffer at pH 2.5 containing 75 mM sodium chloride (squares) and 50 mM sodium chloride (triangles). Each point represents the mean of three experiments.
Increasing the ionic strength of the run buffer resulted in longer migration times and \ increased peak heights (Figure 2.8). At the highest ionic strength (100 mM sodium chloride) the hLHRH and sLHRH analogues eluted after 32.0 and 46.5 min, respectively. These elution times were considered to be too long for our purposes. 61
60 '2 s..... S 40 ...... ~ 0 ·-g 20 I-< .....bJ) ;;E 0 40 50 60 70 80 90 100 Sodium chloride (mM)
Figure 2.8 Effect of the ionic strength of the run buffer on the migration time ofhLHRH (open circles) and sLHRH (closed circles) when electrophoresed in 10 mM sodium phosphate buffer pH 2.5 containing 50, 75 or 100 mM sodium chloride (Section 2.3.1).
! Each point represents the mean of three experiments.
Figure 2.9 shows that diluting the chymotryptic LHRH digests 2.5- to 4-fold in deionised \ water improved the RSD (%) of peak height; however there was no improvement in RSD \ from a 4 to 6-fold dilution. The precision of the migration times of both analogues were not affected by the difference in ionic strengths between run and sample buffer. All RSD were less than 1%. Under the electrophoresis at pH 2.5 containing 75 mM sodium chloride (Section 2.4.1) it would appear that for optimum stacking conditions the digest should be diluted 4-fold in deionised water. 62
5 ' (A) I I 4
~ ~ 3 Q Cl.) ~ 2
1
0 2.5 4 6 I 2.5 4 6
(B) 51 I 4 •
') ~3~ Q Cl.) ~ 2
1 \ 0 2.5 4 6 2.5 4 6 Dilution ofhLHRH Dilution ofsLHRH chymotiyptic digest chymotiyptic digest
Figure 2.9 Effect of diluting the digests in deionised water on RSD (%) for migration time (open bars) and peak height (closed bars) when electrophoresed (Section 2.3.1) in 10 mM sodium phosphate buffer pH 2.5 containing 50 mM sodium chloride (A) and 75 mM sodium chloride (B) (n=3). 63
2.4.3 Optimisation of run buffer pH In an attempt to reduce analysis time, the pH of the run buffer was increased to 4.0, 6.5 and 7.5 to increase the EOF of the system. Increasing the pH of the run buffer reduced the migration time of the LHRH analogues, which are indicated by arrows in Figures 2.10 & 2.11. At pH 7.0 neither LHRH analogue was resolved from its chymotryptic metabolites (data not shown). However at pH 2.5, 4.0 and 6.5 both LHRH analogues were resolved from their chymotryptic metabolites (Figures 2.10 & 2.11 ). Table 2.5 shows that migration time and peak height decrease with increasing pH. Run buffer at pH 6.5 resulted in relatively high RSD for migration time and peak height for both hLHRH and sLHRH (Table 2.5).
·1
.,i 64
0.002 (A) .. 0.001
0.000 0 5 10 .. 15 20 25 0.0021 ,,-._ I (B)
'-"~ 0) Cll ~ 'I 0 0.001 Cll °'0) ~ I-< .....0 0 .....0) 0) 0 Q 0 5 10 15 20 25 0.0061 (C) I 0.004
0.002
0 ,..... ll..JIJ'-'W~
0 5 10 15 20 25
· Time (min)
'1
Figure 2.10 Capillary electrophoretic separation of the chymotryptic hLHRH digest at pH 2.5 (A), 4.0 (B) and 6.5 (C) containing 75 mM sodium chloride (Section 2.3.1). Arrows indicate hLHRH. Chymotrypsin digests analysed were not identical. 65
'\
0.002, II II (A) ... I II 1 II l 0.001
-.~
0.000 0 5 10 15 20 25 30 35
,..-._ 0.0021 ::J. I (B) 0.004, r I (C) .. 0.002 0.000 -f'---,----,,----,----r----r---.----, 0 5 10 15 20 25 30 35 Time (min) Figure 2.11 Capillary electrophoretic separation of the chymotryptic sLHRH digest at pH 2.5 (A), 4.0 (B) and 6.5 (C) containing 75 rnM sodium chloride (Section 2.3.1). Arrows· indicate sLHRH. Chymotrypsin digests analysed were not identical. 66 Table 2.5 Effect of run buffer pH on reproducibility of migration time (MT), and peak height (PH) ofhLHRH and sLHRH (n=3). hLHRH sLHRH buffer MT PH MT PH pH mean RSD(%) mean RSD(%) mean RSD(%) mean RSD(%) ,,,,, / 2.5 23.2 0.5 419 2.9 37.34 0.6 353 3.3 4.0 12.5 0.67 384 3.3 18.5 0.74 321 3.2 "\ 6.5 6.9 0.79 352 4.5 8.3 1.0 166 6.0 k It was concluded that the optimum conditions for CE analysis of both human and salmon LHRH digests under standard conditions (Section 2.3.1) was in 10 mM sodium acetate buffer at pH 4.0 containing 75 mM sodium chloride, with a 4-fold dilution of digest in deionised water. 2.4.4 The effect of chymotrypsin on the reproducibility j "' The effect of chymotrypsin on the reproducibility of peak height and migration time was determined. Human LHRH replicates (n = 6) at 50 and 100 µg/ml were prepared in the presence and absence ofchymotrypsin (Section 2.3 .6) and analysed by CE under the optimum conditions described in Section 2.4.3. Analysis of peak heights showed that there was no significant difference in hLHRH prepared in the presence or absence of chymotrypsin at 150 µg/ml (p > 0.4) or 25 µg/ml (p > 0.3) (Table 2.6). \ .'.I Table 2.6 Effect of chymotrypsin on reproducibility of peak height ofhLHRH. Chymotrypsin Phosphate buffer only 150 µg/ml 25 µg/ml 150 µg/ml 25 µg/ml Mean 1104 139 1090 135 s.d. 32.7 7.5 28.2 7.1 RSD% 3.0 5.4 2.6 5.3 67 2.4.5 Storage stability Human LHRH replicates (n = 9) at 25 and 150 µg/ml were prepared in the presence of chymotrypsin (Section 2.3.6) and stored at -20°C for 5 days. On the day of analysis stored samples were thawed and in addition fresh samples at 25 and 150 ~Lg/ml were prepared in the presence of chymotrypsin (Section 2.3.6). Stored and fresh samples were analysed under optimum capillary electrophoresis conditions (Section 2.4.3). Table 2.7 shows that there was no significant difference between stored and fresh samples at 150 ~Lg/ml and -I 25 µglml (p > 0.3). No LHRH metabolites were detected in electophoretograms -.( indicating no degradation of the native hLHRH occurred during storage under the conditions described. Table 2.7 Effect of storage on the stability ofhLHRH in the presence of chymotrypsin. ample Fresh 25 µg/ml rzi:g---:;:-~-± . --, 5. ± { I Stored 25 µg/ml 155 7.9 I 5.1 1265 ± 42.1 ~ ',i I Fresh 150 µg/ml I 3.3 I Stored 150 µglml 1266 ± 45.2 I 3.6 2.4.6 CE analysis of LHRH Standards ofLHRH analogues (Section 2.3.7) were analysed by CE under the optimum conditions described in Section 2.4.3. Regression analysis of the calibration curve of peak :\ height versus concentration ofhLHRH showed that it was linear over the concentration range 7.6 to 30.34 µglml, (PH= 20.14 (± 0.52) - 77.40 (± 9.40), r2 = 0.993, n=12). The intra-day coefficient of variation for the CE assay was 1.25 % at 15 ~Lg/ml. The inter-day coefficient of variation was 4.97 % at 15 µg/ml. Regression analysis of the calibration curve of peak height versus concentration of sLHRH showed that it was linear over the concentration range 9.2 to 38.0 ~Lg/ml, with PH= 34.6 (+- 0.8) - 226 (+- 11), r2 = 0.990, n=15). 68 2.5 Discussion In this study sample stacking and pH of the run buffer were identified as important { factors in controlling peak shape, separation and reproducibility of chymotryptic digests ofLHRH. In summary, electrophoresis at high pH (6.5 & 7.0) resulted in loss ·of resolution and efficiency (Section 2.4.3). Low pH (2.5 & 4.0) produced a dramatic reduction in the EOF that gave greater efficiency at the cost of longer elution times ·\ (Section 2.4.3). Sample stacking was improved by diluting the LHRH digest sample in ,, deionised water and increasing the concentration of the sodium chloride in the run buffer (Section 2.4.2). f At low pH (2.5) human and salmon LHRH migrated as sharp symmetrical peaks characteristic of CE (Figure 2.3) and exhibited large differences in migration times. \ Human LHRH eluted after 21.5 min while salmon LHRH eluted after 36.5 min. The large differences in migration time was expected as at low pH the EOF is suppressed and .f, analytes migrate according to electrophoretic velocity which is proportional to their ·' ;, charge-mass ratio. At pH 2.5, hLHRH will carry two positive charges due to the , ionisation ofHis2 and Arg8, while salmon LHRH will carry a single positively charged 2 His • For both LHRH analogues the N- and C-terminus will not be ionised as they are blocked. ,.,._, Since the Dionex sample vials could not be capped sample evaporation became a :\ ;, potential problem (Table 2.1). Obviously, for accurate determination ofLHRH concentration when no internal standard was used, sample evaporation had to be prevented. To do this and to allow for gravity loading a 1 mm layer of light mineral oil was placed on top of the sample. When present, the PH of sequential samples did not increase suggests evaporation of the solvent was eliminated or minimised. (Table 2.2). Having eliminated sample evaporation it became evident that human and salmon LHRH analogues were adsorbing to the wall of the sample vials (Table 2.2). LHRH analogues at low concentrations are known to adhere to glass and other surfaces (Anik and Hwang, 69 1983, Mistry, 1993). Silanisation of the Dionex sample vials reduced the absorption of hLHRH from 10% to 4.5%, whilst for sLHRH the absorption was reduced from 14% to 6.5 %. This equated approximately to a 50 % reduction in LHRH adsorption (Table 2.3). On the assump~ion that the small residual adsorption is constant it was considered that this assay was suitable for quantitation of the degradation ofLHRH analogues in vitro. The maximum recommended current before Joule heating of the run buffer becomes -, unacceptable is 60 µA (Dionex, 1989). There was no significant difference in ·c reproducibility between separation at constant current and constant voltage (Table 2.4). i Consequently all subsequent experiments were performed at a constant current of 60 µA in order to maximise the applied field; to reduce the analysis times without excessive Joule heating of the run buffer. In CE, peak width is a function of the velocity of the analytes passing in front of the detector. Thus peaks that migrate slowly exhibit a larger peak area irrespective of their \ concentration (Huang et al., 1989, Altria, 1993). For accuracy, correction for velocity band width based on migration time is necessary (Huang et al., 1989). However, since the RSD for migration times w:as less than 1%, this correction was insignificant. Nethertheless, for electrophoresis at pH 2.5 (Tables 2.1- 2.4) peak height was a more precise measure of concentration than peak area for both hLHRH and sLHRH. This is in agreement with Watzig (1995) who found that the precision of peak height was often better than for peak area and concluded this was due to peak height being less influenced by migration times and integration errors. Electrophoretograms ofLHRH chymotryptic digests diluted in various volumes of deionised water illustrate the efficiency of sample stacking (Figures 2.4 & 2.5). There was peak-shape distortion (asymmetry) when the chymotryptic digest was diluted 2.5-fold in deionised water. At dilutions greater than 2.5-fold, symmetrical peaks were eluted and resolution (Figures 2.4 & 2.5) and reproducibility were improved (Figure 2.9). The improvement in RSD of peak height by using high ionic strength run buffer was 70 further evidence for the utility of sample stacking. Dilution of digests in deionised water "' resulted in greater peak height/area ratios (Figure 2. 7) with minimal decreases in peak heights (Figure 2.6). Figure 2.9 shows that RSD of migration time was less than 1% under all ionic strengths indicating no analyte wall interaction. This could be attributed to . ( the combination of low pH and high sodium chloride concentration in the run buffer and the removal of the protein component (chymotrypsin) by acid precipitation prior to analysis. However higher ionic strength of run buffer resulted in slightly longer analysis -~ times due to the suppression of the EOF (Figure 2.8). However as the EOF increased the resolution and reproducibility of the assay decreased (Figures 2.10 & 2.11). At pH 7.0, human and salmon LHRH analogues were not resolved from their chymotryptic metabolites. It would appear that although the greater EOF , decreases the run time it does not contribute to the separation. Sutcliffe and Corran (1993) found that for the analysis of neurohypophyseal peptides and analogues, high EOF conditions (pH> 6.0) resulted in shorter migration times with reduced resolution. The .( best separation ofneurohypophyseal peptides and analogues was achieved at pH 2.5 , (Sutcliffe and Corran 1993). } Decreasing the elution time by increasing the pH of the run buffer also decreased the ·1 ; peak heights of the LHRH analogues (Table 2.5). This was presumably due to the higher velocities of the analytes passing the detector (Hjerten, 1985). The reproducibilities of LHRH migration times and peak heights were better at pH 2.5 and 4.0 than at pH 6.5 li (Tables 2.5 & 2.6). This is presumably because at higher pH (> 4.0) the capillary surface ,-1_ ;. becomes negatively charged resulting in higher EOF and greater analyte-:wall interaction. Adsorption of analytes to the wall surface will alter the surface charge resulting in a change in the EOF (Lambert and Middleton, 1990, Towns and Regnier, 1992). This ,_,,,,,' problem may be overcome by washing the capillary with acid or alkali between runs (Lambert and Middleton, 1990). However this would increase analysis times and since good results were obtained at lower pH this was not investigated. 71 For the chymotryptic digest of LHRH analogues, a 4-fold dilution of the .digests in deionised water and a run buffer of 10 mM sodium acetate buffer at pH 4.0 containing 75 mM sodium chloride appears to be the optimum for sample stacking and resolution using the standard conditions described in Section 2.3.1. Under these conditions, hLHRH and sLHRH migrated at 12.5 and 18.5 min, respectively. The presence of chymotrypsin in hLHRH chymotryptic digested samples had no significant effect on peak height or migration time reproducibility (Section 2.4.4). This indicates that chymotrypsin - capillary wall interaction is not significant under these conditions. Standard replicates ofhLHRH in the presence of chymotrypsin in 0.1 M HCl were found to be stable after 5 days when stored at -20°C (Section 2.4.5). Subsequently all LHRH digest samples were stored in this manner. The standard curves for peak heights versus concentrations for hLHRH and sLHRH were linear within the concentration range of 7 .6 to 34.34 µg/ml and 9.2 to 38.0 µglml, respectively. They-intercepts for both calibration curves were significantly below zero (Section 2.4.6). This phenomenon is likely due to the adsorption of the LHRH analogues to the silanised sample vial as observed in Section 2.4.1. For hLHRH the intra- and inter-day coefficient of variation for the CE assay was found to be 1.25% and 4.97% at 15 µglml (Section 2.4.6). Some of this variance in precision may be attributed to injection volume fluctuations (Altria et al., 1995). This variance in precision could be investigated by the incorporation of an appropriate internal standard (Dose and Guiochon, 1991) but this was not pursued, as the variability of the CE assay was acceptable for our purposes. The results of this work demonstrate the usefulness of CE for the analysis of the degradation of peptides in vitro; however there are few reports in the literature that describe the use of CE in a similar application. Huang and Wu (1996) used CE to monitor the degradation of thyrotropin releasing hormone and insulin during iontophoretic transdermal delivery in vitro· However they did not report the CE assay conditions or the reproducibility of the assay. 72 The most attractive features of the CE assay is its simplicity and low cost. A single inexpensive capillary is used, the run buffer is aqueous eliminating the purchase and disposal of organic solvents, and the same CE assay can be used for the analysis of both LHRH analogues. This assay was used in the subsequent work to determine the metabolic stability ofh~an and salmon LHRH analogues in the presence of intestinal extracts from salmon. I .\ 73 Chapter 3: The Metabolic Barriers \ ,\ 74 3.1 Introduction The proteolytic activity of the intestine is a major barrier to the oral delivery of proteins and peptides (Section 1.2.1). To our knowledge no investigators have characterised the metabolic barrier to oral delivery of proteins and peptides in teleosts. The purpose of this chapter is to report on the relative proteolytic activity of different regions of the lower intestinal tract of salmon. The degradation of the model protein, BSA and the model peptides, human and salmon LHRH analogues was evaluated in intestinal lumenal and mucosa homogenates in vitro. Degradation of BSA was determined by measuring the formation acid.soluble peptides, whereas the degradation of the LHRH analogues was determined using the capiHary electrophoresis assay described in Chapter 2.0. Protein and peptide degradation by lumenal and mucosa homogenates was also studied in the presence of inhibitors to evaluate their potential for improving oral bioavailability of proteins and peptides. The distribution of the major digestive proteases in the lower gastrointestinal tract was surveyed using synthetic substrates. The results presented in this section show that the lumenal proteolytic activity towards BSA and the LHRH analogues in the posterior section was significantly lower than that of the anterior and middle intestinal sections. However the proteolytic activity of the posterior mucosa! homogenate against BSA was significantly higher than for the middle mucosa! homogenate whi!e both LHRH analogues were degraded only by the posterior mucosa! homogenate. -, 75 3.2 Materials Bovine serum albumin fraction V was purchased from SERVA (Feinbiochemica GmbH & Co., Heidelberg). Dimethyl sulphoxide (DMSO) and Triton X-100 laboratory reagent were obtained from BDH limited (Poole, England). N a-benzoyl-DL-Argp-nitroanilide, N-succinyl-Ala-Ala-Ala p-nitroanilide, L-leucine p-nitroanilide, N-benzoyl-L-Tyr ethyl ester, N-a-benzoyl-L-Arg ethyl ester (BABE), N-a-benzyol-L-Arg (BA), hippuryl-Arg, hippuryl-Phe, hippuric acid, human luteinizing hormone-releasing hormone (hLHRH), salmon luteinizing hormone-releasing hormone ( sLHRH), a- chymotrypsin (EC 3 .4.21.1, TLCK treated, Type VIII from bovine pancreas), trypsin (EC 3 .4.21.4, TPCK treated type XIII from bovine pancreas), elastase (EC 3.4.21.36, Type IV from porcine pancreas), .I soybean trypsin-chymotrypsin inhibitor (Bowman-Birk inhibitor) from soybean, ethylenediaminetetraacetic acid disodium salt (EDTA), deoxycholic acid sodium salt and (.\ bestatin were all obtained from Sigma Chemical Company (St Louis, MO, USA). All other chemicals were of analytical grade and were purchased from Ajax chemicals (Auburn, NSW, Australia). Sample tubes of 0.5 ml and 1.5 ml were purchased from Eppendorf (EppendorfNetheler Hinz GmbH Hamburg, Germany). Tissue extracts were homogenised by a Ystral homogeniser (Ystral GmbH, D-7801, Dottingen, Germany). Deionised water was obtained from a reverse osmosis Mill-Q® Reagent Water system and was used for the preparation of all solutions. [, 76 3.3 Methods 3.3.1 General 3.3.1.1 Fish Juvenile chinook salmon (Oncorhynchus tshawytscha) reared at the Big Sea Foods Ltd. hatchery, South Canterbury were transferred to the Portobello Marine Laboratory (Dunedin, New Zealand) where they were adapted to sea water and maintained in a continuous circulating tank. The salmon were fed a commercial feed daily. The salmon were killed with a blow to the head followed by cervical dislocation. The fish were weighed (weighting) and the biological samples were collected at the Portobello Marine ;I Laboratory and transported to Dunedin. Ethics approval was obtained from the University of Otago Ethics Committee which is established under the "Animal Protection (Codes of Ethical Conduct) Regulation", 1987. This code is equivalent to the "Principles of Laboratory Animal Care" (NIH publication# 85 - 23, revised 1985). 3.3.1.2 Collection and preparation of lumenal homogenates The lower intestinal tracts of salmon were removed and ligated into the anterior, middle )· and posterior intestinal sections. The lumenal contents of the section were carefully squeezed into a centrifuge tube (10 ml), the intestinal section was cut longitudinally and the remaining contents were washed into the centrifuge tubes with the minimum volume of ice cold saline. Each section was diluted with cold 50 mM sodium phosphate buffer ~ pH 8.0 to give a final volume of 5 ml, homogenised at 4°C for 2 min using a Ystral homogeniser and centrifuged (3 000 g for 10 min, 4°C). The supernatant was collected and 100 µl volumes were dispensed into 0.5 ml sample tubes and stored under the different conditions described. 77 3.3.1.3 In situ pH measurements The lumenal contents of the anterior, middle and posterior intestinal sections from salmon were collected into 1.5 ml samples tubes without washing (Section 3.3.1.2). The pH measurements were carried out with a micro pH electrode (Model Z451, SCHOTT GERATE, Germany) at room temperature. 3.3.1.4 Preparation of mucosal homogenates from frozen intestine " The mucosal surface area of the anterior section was too small and fragile to collect ·\ sufficient quantities of the mucosa. The lower intestinal tract was removed, the middle and posterior intestinal sections were J ligated, cut and the lumenal contents washed from each section as described in Section 3.3.1.2. The whole intestinal sections were immediately placed in liquid nitrogen and transported to Dunedin where they were stored at -80°C. The posterior intestinal sections (20 g frozen weight) were cut into small pieces and thawed in 60 ml of cold 300 rnM mannitol in 10 rnM sodium phosphate buffer pH 7.4. The suspension was vortexed in a ,( 50 ml test tube for 5 min. The suspension was passed through a Buchner funnel with 1 mm holes and the filtrate was diluted 6-fold in 50 mM sodium phosphate buffer pH 7.4. The filtrate was homogenised at 4°C for 2 min using a Ystral homogeniser and centrifuged at 3 000 g at 4°C for 10 min. Samples of the supernatant (100 µ1) were transferred to 0.5 ml sample tubes and stored at -80°C. 3.3.1.5 Preparation of mucosal homogenates from fresh intestine ;, Middle and posterior intestinal sections were removed and lumenal contents washed from each section as described Section 3 .3 .1.2. The sections were cut longitudinally and the mucosal surface was washed with cold 50 mM sodium phosphate buffer pH 7.4 and gently cleaned with a soft moist tissue. The mucosa was removed from the intestine immediately by scraping off the epithelial cell layers with a microscope cover slip. The scrapings of the middle and posterior sections were pooled and homogenised in 5 ml of 50 mM phosphate buffer pH 7.4 at 4°C using a Ystral homogeniser. The homogenate was 78 centrifuged at 3 000 g at 4°C for 10 min to remove cellular and nuclear debris. Samples of the supematants (100 µ1) were transferred to 0.5 ml sample tubes and stored under conditions described. 3.3.1.6 Protein estimations Protein contents was estimated by the method of Lowry as modified by Peterson (1977). Trichloroacetic acid (TCA) precipitation was not required. Assays were performed in microtitre plate wells in a total volume of250 µ1, and the absorbance (750nm) was read using an Microplate Spectrophotometer (Spectra MAX 340) controlled by an external computer with SOFT-MAX PRO software (Molecular Devices Corporation, CA, USA). Protein concentrations were equilibrated using bovine serum albumin standards covering the range 1-10 µg. Determination of protein was done in triplicate. 3.3.1.7 Measurement of surface tension :t A 200 mM stock solution of sodium deoxycholate was prepared in 50 mM sodium phosphate buffer pH 8.0. From the stock solution, appropriate dilutions were made in 50 mM sodium phosphate buffer pH 8.0 producing concentrations 0.156, 0.3125, 0.625, 1.25, 2.5, 3.0, 5, 10, 20, and 40 mM. Surface tension data, as a function of concentration, were determined employing the ring-detachment method Choulis and Loh (1971) using a torsion balance (White Blee. Inst. Co. Ltd). The surface tension of distilled water was 74, 73, and 73 mN/m at room temperature in agreement with Bloor et al. (1970). All measurements were performed at room temperature and the averages of two experiments are reported. All glass apparatus used and the platinum ring were washed thoroughly with distilled water. The dish was rinsed with the sodium deoxycholate standard solution, 25 ml of fresh standard added to the glass dish and the ring submerged just below the surface before the surface tension was determined. 79 3.3.1.8 Statistical analysis The statistical analyses (two way ANOVAs) were performed by the statistical program, MINITAB Release 10 (Minitab, Inc., PA. USA). Whenever an F-test showed a significant difference in means, the least-significant differences method was used to compare means pairwise. 3.3.2 Proteolytic assays 3.3.2.1 General proteolytic activity Lumenal or mucosal homogenate samples (15-20 µg of protein) were diluted with 50 mM sodium phosphate buffer with or without inhibitor to give a total volume of 0.5 ml. After incubation at 150c for 30 min, 0.5 ml of BSA (1 mg) in 50 mM phosphate buffer was added to start the assay. At various times after the start of the reaction, 100 µ1 samples were removed and added to 5 µ1 of 50 % (w/v) TCA to stop the enzyme reaction and to precipitate protein. Reaction samples were left to stand at room temperature for at least 30 min, centrifuged at 20 000 g for 5 min and 50 µ1 of the supernatant was assayed for acid-soluble peptides by the modified Lowry method (Section 3.3.1.6). The amount of acid-soluble peptide was determined using bovine serum albumin as the standard over the range 1 - 10 µg. Units of general proteolytic activity are µg of acid-soluble peptides formed per minute. 3.3.2.2 LHRH analogues Lumenal and mucosa! homogenates (1.25 - 12.5 µg of protein) were diluted with 50 mM sodium phosphate buffer and were incubated at 15°C for 30 min (total volume 200 µI) in the presence or absence of inhibitor. To start the reaction, 50 µ1 ofLHRH was added and at predetermined times, 25 µl was withdrawn from the incubates and mixed with 25 µ1 of 0.2 M HCl to precipitate tissue proteins and terminate the reaction. The resulting mixture was centrifuged (20 000 g for 10 min) to remove precipitated protein and particulates and 40 µ1 of the supernatant was stored at -20°C for analysis by capillary electrophoresis. 80 Samples (20 µl) were transferred to a Dionex sample vials and diluted with 60 µl of water. A 1 mm layer of mineral oil was placed on top of the sample to prevent sample evaporation. Capillary electrophoresis (CE) was carried out in a Dionex Model CES-1 capillary electrophoresis instrument (Dionex, Sunnyvale, CA, USA). The electrophoresis conditions are the same as for the optimum conditions described in Section 2.4.3. ( 3.3.2.3 Assays usingp-nitroanilides Trypsin activity was assayed based on the method of Erlanger et ql. (1961) using Na benzoyl-DL-Argp-nitroanilide (BAPNA) as substrate. BAPNA (43.5 mg) was dissolved \ in 1 ml of dimethylsulfoxide (DMSO) and then diluted 100-fold in 50 mM sodium phosphate buffer at pH 7.4 for mucosal studies and at pH 8.0 for lumenal studies. ,I Elastase activity was assayed using the substrate N-succinyl-Ala-Ala-Ala p-nitroanilide (SucAla3NA). SucAla3NA (4.5 mg) was dissolved in 1 ml ofDMSO and diluted 10-fold in the appropriate sodium phosphate buffer (Bieth et al., 1974). Leucine aminopeptidase was assayed using the substrate L-Leu-p-nitroanilide (Lin and Van Wart, 1982). Homogenates (50 µ1 total volume) and the p-nitroanlilide substrates (250 µl total volume) ,,( were preincubated at 25°C for 10 min in the microtitre plate wells. To start the assay a 200 µ1 sample of the p-nitroanlilide substrate (1.0 mM) was transferred to the wells containing the homogenate. The release ofp-nitroaniline was followed by the increase in absorbance at 410 nm using a Microplate Spectrophotometer (Spectra MAX 340) plate 1' reader. The apparent extinction coefficient ofp-nitroaniline was calculated to be 6500 at pH 8.0 in the microtitre wells. > >' ,,,, 3.3.2.4 Chymotryptic activity The assay of chymotryptic activity was based on the method of Rao and Lombardi (1975) with N-benzoyl-L-Tyr ethyl ester (BTEE) as the substrate. BTEE (16 mg) was dissolved ., .. in 0.5 ml of methanol and a solution of0.2% Triton X-100 in 50 mM sodium phosphate I buffer was added slowly with shaking to give a final volume of 50 ml and substrate concentration of 1.0 mM. Enzyme samples were diluted with 50 mM sodium phosphate buffer at the appropriate pH and 50 µ1 of the samples were incubated in the presence and 81 absence of 50 µ1 of inhibitor in 50 mM sodium phosphate buffer at 25°C. After 30 min, 50 µ1 of the incubate was transferred to a 1 ml quartz cuvette of 1 cm path length, containing 150 µ1 of sodium phosphate buffer. To initiate the reaction 0.8 ml of substrate stock was added. Buffer and substrate .stock were both preincubated at 25°C for at least 10 min prior to the assay. The reaction was monitored by a Hewlett Packard 8452A diode array spectrophotometer at 256nm, with a jacketed cell maintained at 25°C by recirculating water. The molar extinction coefficient of was 964 (Rao and Lombardi, 1975). Unit definition, 1 unit hydrolysed 1 µmole ofBTEE per min at 25°C. 3.3.2.5 Carboxypeptidase A and B ·activity The measurement of carboxypeptidase A and B was based on the relative absorbance difference ofhippuric acid to their the respective substrates hippuryl-L-Phe (Hipp-L-Phe) (King et al., 1987) and hippuryl-L-Arg (Hipp-L-Arg) (Folk et al., 1960) at 254nm. The molar extinction coefficients were determined by measuring the absorbance difference between a 1 mM solution ofhippuric acid in 5 mM sodium phosphate buffer pH 8.0 and 1 mM solutions of Hipp-L-Phe or Hipp-L-Arg in 50 mM sodium phosphate buffer pH 8.0, in 1 ml quartz cuvette of 1 cm path length. The molar extinction coefficients for 1 1 Hipp-L-Phe and Hipp-L-Arg were 2 600 and 573 M- cm- , respectively. Hipp-L-Phe and Hipp-L-Arg were dissolved in 100 ml of 50 mM sodium phosphate buffer pH 8.0 to give a final concentration of 1 mM. Lumenal homogenates (0.5 ml) were incubated at 25°C for 10 min and a 0.2 ml sample was transferred to 1 ml quartz cuvettes of 1 cm path length. To initiate the reaction, 0.8 ml of substrate stock pre-incubated at 25°C for 10 min was added to the quartz cuvette. The activities were measured by following the increase in absorbance at 254nm resulting from the liberation hippuric acid. Absorbance was monitored by a Hewlett Packard 8452A diode array spectrophotometer, with a j ack:eted cell maintained at 25°C by recirculating water. 82 3.4 Results 3.4.1 Lumenal proteolytic activity 3.4.1.1 Intestinal lumen pH Lumenal contents of four salmon (354 ± 96 g, mean± s.d.) where collected 12 h after food was last released into their tank (Section 3 .3 .1.3 ). The mean pH of two readings of individual lumenal contents are shown in Table 3.1. Table 3.1 pH of the lumenal contents of salmon. Gut section Fish 1 Fish2 Fish 3 Fish4 Mean± s.d. Anterior 7.7 7.88 8.00 8.18 7.94 ± 0.2 Middle 8.20 8.25 8.15 8.00 8.15±0.1 Posterior 8.50 8.16 8.19 8.42 8.28 ± 0.2 There was no significant difference in the pH of the lumen in the three sections in the four fish analysed (p > 0.1). For all subsequent lumenal metabolic studies 50 mM sodium phosphate buffer at pH 8.0 was used for consistency. 3.4.1.2 Optimum storage conditions for lumenal extracts )' Enzyme samples had to be collected from salmon at the Portobello Marine Laboratory and transported to Dunedin for analysis. Lumenal gut extracts were collected (Section 3 .3 .1.2) from the anterior, middle and posterior gut sections of a salmon (296 g), stored under various conditions and then assayed for proteolytic activity towards BSA (Section 3.3.2.1). To determine the best method for enzyme storage lumenal extracts collected > from the salmon were either immediately snap frozen in liquid nitrogen or stored on ice until arrival to Dunedin where the homogenates stored on ice were immediately assayed for general proteolytic activity (3 hr post mortem). The remaining lumenal extracts stored on ice together with the extracts snap frozen in liquid nitrogen were stored at -80°C for later analysed for general proteolytic activity. The anterior (12.7 µg of protein), middle 83 (12.9 µg of protein) and posterior (15.8 µg of protein) were incubated with BSA (1 mg, 1 ml total volume) in 50 mM sodium phosphate buffer pH 8.0 at 15°C. Samples taken at 0, 30, 60, 120 and 180 min were assayed for acid-soluble peptides (Section 3.3.2.1). Figure 3 .1 illustrates that the formation of acid-soluble peptides followed pseudo-zero kinetics. 180 I _t (A) I, 120 60 0 0 50 100 150 200 -]-- l 180 I (B) 60 0 0 50 100 150 200 Time (min) Figure 3.1 Time-course of the formation of acid-soluble peptides from BSA when incubated with the anterior (A), middle (B) and posterior (C) lumenal homogenates at pH 8.0 at 15°C (Section 3.3.2.1). 84 Protein concentrations of the anterior, middle and posterior homogenates were 6.35, 6.45 and 4.0 mg/ml, respectively (Section 3.3.1.6). The general proteolytic activity of anterior, middle and posterior gut sections after the different storage treatments are shown in Table 3 .2. The rank order of general proteolytic activity in the three sections followed anterior > middle > posterior. Table 3.2 General proteolytic activity ofintestinal lumenal homogenates stored at 4°C for 3h, -80°C after being stored on ice and -80°C after being snap frozen in liquid nitrogen (mean± s.d., n=3). Gut section Storage treatment Fresh - 80°C Liquid nitrogen Anterior 3.71 ± 0.18a 3.46 ± 0.13 3.87 ± 0.19 Middle 3.25 ±0.13 2.08 ± 0.12 3.66 ± 0.16 Posterior I ·1.56 ± 0.1 0.87 ± 0.07 1.76 ± 0.08 a Units of proteolytic activity are mg of acid soluble peptides / h / mg of gut protein The proteolytic activities of the lumenal homogenates snap frozen in liquid nitrogen and stored at -80°C were not significantly different to freshly prepared extracts. All '~ subsequent gut lumenal extracts were immediately snap frozen in liquid nitrogen on collection and stored at -80°C. 3.4.1.3 Effect of feeding on lumenal protein content ~ The effect of feeding on protein content in the intestinal lumen of the lower intestinal tract was studied. Anterior, middle and posterior lumenal contents of fish which had no food released into the tank for 12 h (237 ± 16 g, n = 5) and 24 h (220 ± 23 g, n = 5) prior to sacrifice were collected (Section 3.3.1.2). The supematants were aliquoted (100 µ1) into 0.5 mi sample tubes. Samples (10 µl) were analysed for protein content (Section 3.3.1.5). Two-way analysis of variance of the total protein content of intestinal sections after 12 hand 24 h without food release showed there is no significant difference in amount of protein in the respective sections (p > 0.1, Table 3.3). However there was a significant difference in protein contents of the different intestinal sections. The average 85 total protein contents for the anterior, middle and posterior sections were 20.1 ± 9.3, 4.7 ± 1.62 and 4.6 ± 1.77 mg (mean± s.d., n=lO), respectively. Post anova-analysis by least significant differences showed the protein content of the anterior section to be significantly greater than the middle or posterior sections (p < 0.001) while the difference between middle and posterior sections was not significant (p > 0.5). Table 3.3 Two-way analysis of variance of the protein content of the anterior, middle and posterior intestinal sections after 12 and 24 h without food release. Source DF SS MS F-value Starvation period 1 13.6 13.6 0.39 Section 2 1586.3 793.2 22.5 Interaction 2 21.9 10.9 0.31 Error 24 844.5 35.2 Total 29 2466.3 3.4.1.4 Effect of feeding on proteolytic activity The lumenal samples of the salmon which had no feed released into the tank 12 h and 24 h prior to collection were assayed for general proteolytic activity (Section 3.3.2.1). Table 3 .4 shows the specific activities and total proteolytic activities in each section for both feeding groups of salmon. Table 3.4 Effect of feeding on specific activities and total activities of the anterior, middle and posterior intestinal sections. 12h 24h Section Sp Act. Total Act. Sp Act. Total Act. Anterior 2338 ± 468a 51156 ± 1805° 3231 ±.524 58196 ± 2345 Middle 2255 ± 366 8917 ± 2010 2230 ± 410 9819 ± 1810 Posterior 1110 ± 424 4964 ± 1093 1105±250 5245 ± 800 a µg /h/mg of protein (mean± s.d., n=5) b µg/h/gut section (mean± s.d., n=5) 86 Two-way analysis of variance showed that there was no significant difference in the proteolytic activity (Table 3.5) or total activity (Table 3.6) between salmon which had no food released for 12 hand those which had no food released for 24 h (p > 0.1). However there was a significant difference in proteolytic activities and total activities between the three gut sections (p < 0.001). Table 3.5 Two-way analysis of variance of specific proteolytic activity of the anterior, middle and posterior intestinal sections after 12 and 24 h without food release. Source DF ss MS F-value Time 1 821377 821377 1.06 :( Section 2 14700659 7350329 9.5 Interaction 2 1199766 599883 0.77 Error 24 18648618 777026 Total 29 35370420 Table 3.6 Two-way analysis of variance of total activity of the anterior, middle and •I posterior intestinal sections after 12 and 24 h without food release. Source DF SS MS F-value Time 1 3 3 0 Section 2 12351 6175 13.5 Interaction 2 18 9 0 Error 24 10999 458 Total 29 23370 C,: The mean specific proteolytic activities of the anterior, middle and posterior sections for combined fed groups were 2784 ± 490, 2242 ± 395 and 1107 ± 123 µg / h per mg of protein (mean± s.d., n = 10), respectively. Post anova analysis by least significant differences show that the proteolytic a~tivity of the posterior section was significantly lower than the anterior (p < 0.05) and middle sections (p <0.1). However there was no significant difference between the anterior and middle intestinal sections (p > 0.1 ). Total proteolytic activity of the anterior, middle and posterior sections were 54.7 ± 2.12, 9.4 ± 87 2.0 and 5.1 ± 0.9 mg/ h per intestinal section (mean± s.d., n = 10), respectively. The total activity of the anterior intestinal section was significantly greater than the middle (p < 0.05) and posterior sections (p < 0.025). There was no significant difference in total proteolytic activity between the middle and posterior sections (p > 0.1 ). The rank order of proteolytic and total activities in the anterior, middle and posterior sections of individual fish is shown in Table 3.7. For proteolytic activity and total activity no fish had the rank order middle > posterior > anterior or posterior > anterior > middle. The order of activity, anterior > middle > posterior Was observed in 5 salmon for specific proteolytic activity and 7 for total activity. The probability of such an event or a more extreme result ·, occurring randomly is 0.015 and 0.000267, respectively. Table 3. 7 Frequency of gut section activities of individual salmon indicating a possible "( rank ordera. Anter>Mid>Post Anter>Post>Mid Mid>Anter>Post Post>Mid>Anter Specific +++++ + ++ + Activity Total +++++++ +++ ; \ Activity ) a No fish showed the frequency order ofMid>Post>Anter or Post>Anter>Mid for either I I proteolytic activity and total activity f ~ 3.4.1.5 LHRH digestion by pancreatic enzymes The digestion of human and salmon LHRH by pure pancreatic enzymes trypsin, \ a-chymotrypsin and elastase was investigated. LHRH (0.1 mM) analogues were incubated with the pancreatic enzymes at 2.5 µg/ml in 50 mM sodium phosphate buffer pH 8.0 at 15°C. Electrophoretograms of human and salmon LHRH chymotryptic digests ·( are shown in Figures 3.2 and 3.3. The rate of the disappearance of the human and salmon LHRH in electrophoretograms peak followed pseudo-zero kinetics with half-lives of2.5 ± 0.3 and 2.7 ± 0.4 min (mean± s.d., n=3), respectively. Chymotryptic digestion of hLHRH resulted in the formation of2 primary fragments, which migrated at 8.2 min and 88 9.4 min, with native human LHRH eluting after 10.7 min (Figure 3.2). Salmon LHRH eluted after 18.6 min with a primary chymotryptic metabolite eluting after 14.5 min (Figure 3.3). ,( \ ," ,. 89 0.0081 (A) I 0.006 0.004 0.002 ,/ 0 0 5 10 15 ) ,-...... _,~ 0.004 I (B) (l) tl.l Q 0.003 0 0. tl.l (l) 0.002 ', ~ l-< 0 0 0.001 .....(l) (l) 0 0 0 5 10 15 0.0041 (C) 0.003 ( 0.002 0.001 ., 0 0 5 10 15 Time (min) .1' Figure 3.2 Capillary electrophoretic separation of the chyrnotryptic hLHRH digest (Section 3.3.2.2). Digests were analysed after O (A), 2 (B) and 10 (C) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 90 0.0081 (A) 0.006 0.004 0.002 0 0 5 10 15 20 ,.-... 0.008 l (B) ~ 0.006 Q) f/l ~ 0 0.004 0.. f/l ~ 0.002 J.-< .....0 \ (.) i .....Q) 0 l Q) 0 0 5 10 15 20 0.0081 (C) f 0.006 -\ 0.004 0.002 0 0 5 10 15 20 Time (min) Figure 3.3 Capillary electrophoretic separation of the chymotryptic sLHRH digest (Section 3.3.2.2). Digests were analysed after O (A), 2 (B) and 10 (C) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 91 Elastase from porcine pancreas hydrolysed human LHRH resulting in the formation of a single metabolite migrating after 8.18 min, with the native LHRH eluting after 10.7 min (Figure 3.4). Elastase digestion of the salmon LHRH resulted in the formation of a single metabolite that eluted after 13.8 min with the native salmon LHRH eluting after 17.8 min (Figure 3.5). The half-lives of human and salmon LHRH in the presence of elastase were 90 ± 15 and 114 ± 21 min (mean± s.d., n=3), respectively. When trypsin (TPCK treated) :from bovine pancreas was incubated with human and salmon LHRH, no degradation was observed after 120 min (Figure 3.6). ,( 0.008 (A) 0.006 0.004 ~ 0.002 V § o I<::'. :::: ~ ---::::::;::::4~ ~ 0 5 10 15 ~ i... ~ 0.004 t (B) !, 0 .\ 0.002 0.000 ~------,-----..------, 0 5 10 15 Time (min) · Figure 3.4 Capillary electrophoretic separation of the elastase digest ofhLHRH (Section 3.3.2.2). Digests were analysed after O_ (A) and 60 (B) min incubation at pH 8.0, 15°C (Section 3.4.1.5). 92 0.008 , (A) I 0.006 0.004 0.002 r"\ 0 0 $ 11) 0 5 10 15 20 C/l ~ ,..{ 0 0.. 0.004 C/l 11) \ ~... I (B) 0 t) .....11) 11) 0.002 Q 0.000 ---~-----,---...... ------, 0 5 10 15 20 \ Time (min) Figure 3.5 Capillary electrophoretic separation of the elastase digest of sLHRH (Section 3.3.2.2). Digests were analysed after O (A) and 60 (B) min incubation at pH 8.0, 15°C (Section 3 .4.1.5). 'I 93 0.008 (A) 0.006 0.004 0.002 ,-,, ~ 0 II) r/l 0 5 10 15 A 0 0.. r/l ,{ II) I-< 0.008 I-< (B) 0 -'( ..... (.) II) 0.006 t Q 0.004 0.002 0 0 5 10 15 (_ Time (min) Figure 3.6 Capillary electrophoretic separation of the trypsin hLHRH (A) and sLHRH (B) digest (Section 3.3.2.2). Digests were analysed after 120 min incubation at pH 8.0, 15°C (Section 3.4.1.5). :;- ( C· 1\ ), ,\ 94 3.4.1.6 LHRH digestion by lumenal enzymes Lumenal contents were collected from salmon (234 ± 34 g, n=6) and intestinal sections dissected, prepared and pooled (Section 3.3.1.2). Anterior (1.3 µg of protein), middle (1.3 µg of protein) and posterior (1.35 µg of protein) lumenal extracts were incubated in 50 mM sodium phosphate pH 8.0 (200 µl total volume) for 30 min at 15°C. LHRH was added (0.2 mM, 250 µl total volume), samples taken at 0, 60 and 120 min were analysed by capillary electrophoresis (Section 3.3.2.2). Electrophoretograms of human and salmon LHRH analogues digested by the anterior, middle and posterior lumenal extracts are shown in Figures 3.7 and 3.8. The human and salmon LHRH analogues eluted after 15.5 and 24.5 min, respectively. Each LHRH analogue digested to give a unique metabolic profile that was common to each of the lumenal sections. Degradation followed pseudo I zero order kinetics. Table 3.8 shows the half-lives of both hLHRH and sLHRH in the anterior, middle and posterior intestinal sections. J ( ·~ ·> ·, 95 0.012 . (A) I I 0.008 0.004 0.000 0 5 10 15 20 0.004, (B) 0.002 ,.-... ( ::::i $ 0 0.000 C/l Q 0 0 5 10 15 20 0.. C/l 0 ~ 0.004 I ~ (C) 0 +-'u Q * 0.002 J 0.000 0 5 10 15 20 0.008 (D) / . ' . 0.006 0.004 0.002 0.000 0 5 10 15 20 Time (min) Figure 3.7 Capillary electrophoretic separation ofhLHRH (A) and hLHRH lumenal digests by the anterior (B), middle (C) and posterior (D) intestinal section (Section 3.3.2.2). Samples were analysed 120 min after incubation at pH 8.0, 15°C (Section 3.4.1.6). 96 0.008 . (A) I I 0.006 0.004 0.002 0 0 5 10 15 20 25 -; 0.004, (B) 0.002 ~ Q) 0.000 -----.-----.------,.------,---~ :1- ~ 0 0 5 10 15 20 25 0.. rt.l ~ 0.004 (C) lo< ~ 0* 0.002 \ 0,000 ----.----.------r---.---~ 0 5 10 15 20 25 0.006 (D) ' ) '{ 0.004 0.002 " \ 0.000 ~ I I 4!~\I._.,..._.;:, 0 5 10 15 20 25 Time (min) ,; Figure 3.8 Capillary electrophoretic separation of sLHRH (A) and sLHRH lumenal -' digests by the anterior (B), middle (C) and posterior (D) intestinal section (Section 3.3.2.2). Samples were analysed 120 min after incubation at pH 8.0, 15°C (Section 3.4:1.6). 97 Table 3.8 Half-lives (min) ofhLHRH and sLHRH in the anterior, middle and posterior lumenal homogenates. Half-lives (minl LHRH analogue I Anterior Middle Posterior hLHRH I 28.8 ± 2.9 27.4 ± 4.2 58.4 ± 5.0 sLHRH I 26.9 ± 2.9 26.7 ± 4.1 52 ± 3.6 a Half-lives represents the mean± s.d., n = 3. Two-way anova analysis showed that there was no significant difference between the half-lives ofhLHRH and sLHRH (p > 0.1) in each intestinal section, however the half lives of the LHRH analogues were significantly different between intestinal sections ·'· (p < 0.001, Table 3.9). Post anova analysis by least-significant differences methods showed that there were no significant differences between half-lives ofLHRH analogues in the anterior and middle extracts (p > 0.5), however the half-lives in the anterior and middle extracts were significantly shorter than in the posterior section (p < 0.001). Table 3.9 Two-way analysis of variance of half-lives ofhLHRH and sLHRH in the anterior, middle and posterior lumenal homogenates. Source .DF ss MS F-value LHRH analogue 1 29.6 29.6 2.5 Section 2 28565 1432.5 122.4 Interaction 2 9.7 4.9 0.42 Error 12 140.8 11.7 >- Total 17 3045.3 3.4.1.7 Specific proteolytic activity The lumenal activity towards specific substrates of different proteases was determined in the anterior, middle and posterior of the intestinal tract. The intestinal lumenal homogenates (10 µg of protein) were assayed for specific activity using synthetic substrates at pH 8.0 at 25°C (Sections 3.3.2.3-3.3.2.5). Figure 3.9 shows that highest proteolytic activity of trypsin, chymotrypsin, elastase, and carboxypeptidase A and B resides in the anterior and middle intestinal sections, whereas the highest lumenal activity 98 ofleucine aminopeptidase was found in the posterior section. The relative distribution of trypsin, chymotrypsin, elastase, carboxypeptidase A & Band aminopeptidase (total units) among the intestinal sections is shown in Figure 3 .10. -/ ·', ....., ' ·, ,, \ .'·, 99 1000 750 500 250 0 Anterior Middle Posterior 10000 7500 5oo'o 2500 -. 0 c:: ·o:;.... Anterior Middle Posterior 0 ;... 300 0. t+-. 0 200 sbl) ;... 100 100 0 Anterior Middle Posterior 10] (F) 8 6 :r 4 ·c 2 0 Anterior Middle Posterior Figure 3.9 Specific activity of trypsin (A), chymotrypsin (B), elastase (C) and carboxypeptidase A (D) and B (E) and leucine aminopeptidase (F) in the anterior, middle and posterior intestinal sections (mean± s.d., n=3). The intestinal lumenal homogenates (10 µg of protein) were assayed for specific activity at pH 8.0, 15°C (Sections 3.3.2.3 - 3.3.2.5). 100 70 -5('j (I.) .....i:: 60 .2'.;.f' 50 0 i:: ('j 0 +-> ..... 40 ::l 0 bJ) (I.) -!! crjCl) 30 B .S (I.) tn 20 ,:;S (I.) (;...,0 .....E ;(I.) 10 i::: Q) <(; Q) e o cil ·;;; .E ~ u "'(' Figure 3.10 Relative distribution of general proteolytic activity, trypsin, chymotrypsin, elastase, carboxypeptidase A and B and leucine aminopeptidase activity of the lower \ intestinal tract. General proteolytic activity was measured using BSA as the substrate (Section 3 .3 .2.1) remaining proteolytic activities were determined using synthetic substrates (Sections 3.3.2.3-3.3.2.5). Anterior C• ), middle (D) and posterior (fl) '/ intestinal sections. The data represents the mean of three experiments. 'I ,, 3.4.2 Mucosal proteolytic activity 3.4.2.1. Collection and storage of mucosal homogenates Methods for the extraction and storage of mucosal homogenates were investigated based on proteolytic activity of posterior mucosal homogenate towards human LHRH. Firstly, the mucosa was prepared from whole frozen intestine of salmon (234 ± 15 g, n=5) (Section 3.3.1.4). The homogenised mucosae were prepared from whole frozen intestine immediately assayed for hLHRH metabolism (Section 3.3.2.2) and the remaining material stored at -80°C to be assayed for hLHRH metabolic activity at a later date. Human LHRH (0.2 mM) was incubated with mucosal homogenates (5 µg of protein) in 101 50 mM sodium phosphate buffer pH 7.4 at 15°C. The degradation ofhLHRH followed pseudo-zero order kinetics. Fresh posterior mucosa prepared from whole intestine hydrolysed hLHRH with a half-life of 51 ± 2.5 min (mean± s.d., n=3), while the mucosa stored at -80°C hydrolysed hLHRH with a half-life of93 ± 14 min (mean± s.d., n=3). Secondly, mucosa was prepared from fresh intestine and snap frozen in liquid nitrogen at the Portobello Marine Laboratory and stored at -80°C on arrival to Dunedin (Section 3.3.1.4). The half-life ofhLHRH in this posterior mucosal homogenate was 41 ± 4.0 min ,, (mean± s.d., n=3). Since the mucosa extracted, homogenised and snap frozen at Portobello Marine Laboratory gave the highest metabolic activity towards hLHRH, ., ' mucosa prepared this way were used for all subsequent studies. 3.4.2.2 Stability of LHRH in mucosal homogenates The metabolic stability of human and salmon LHRH in middle and posterior mucosal homogenates were determined. Both human and salmonLHRH (0.2 mM) were metabolically stable in the middle mucosa (50 µg/ml of protein) when incubated at 15°C, pH 7.4 for 4 h. However under the same conditions both LHRH analogues were degraded -', in the posterior mucosal homogenate (Figure 3 .11 ). The degradation of both LHRH analogues followed pseudo-zero order kinetics. The half-lives of the human and salmon LHRH analogues in the posterior mucosal homogenate were 111 ± 10 min and 101 ± 8 t min (mean± s.d., n=3), respectively. ( ir 102 0.0161 (A) I 0.012 0.008 0.004 0.000 j 0 5 10 15 20 25 0.0041 (B) 0.003 ., ' ,-._ 0.002 ~ $ 0.001 II) en ~ .') 0 0.000 0..en II) '\ 0 5 10 15 20 25 ~ l-< 0 0.0121 0 (C) II) ~ Q 0.008 .,I 0.004 0.000 0 5 10 15 20 25 ( 0.0061 (D) 0.004 :·r 0.002 0.000 0 5 10 15 20 25 Time (min) Figure 3.11 Capillary electrophoretic separation ofhLHRH (A), and hLHRH posterior .l. mucosal digest (B), sLHRH (C) and sLHRH posterior mucosal digest (D) (Section 3.4.2.3). Samples were analysed 180 min after incubation at pH 7.4, 15°C (Section 3.4.1.6). 103 3.4.2.3 General proteolytic activity The proteolytic digestion of BSA in the middle and posterior mucosal homogenates (Section 3.3.2.1) is shown in Figure 3.12. Mucosal homogenates (20 µg of protein) were incubated with BSA (1mg, ·1 ml total volume) in 50 mM sodium phosphate buffer pH 7.4 at 15°C (Section 3 .3 .2.1 ). The formation of acid-soluble peptides from BSA followed pseudo-zero order kinetics. The general proteolytic activities of the middle and posterior j\ mucosa! homogenates were 442.4 ± 28 µg / h per mg of protein and 664 ± 36 µg / h per / mg ofmucosal protein, respectively (mean± s.d., n = 3) and were significantly different (p < 0.005). ~ 25 • I ...... \ .g ao 20 • f/l s '[)"O -~ 15 - ro '-' C+-t f/l 0 i:1 ...... ~ 10 .8 a .._. ~ 5 \ 0.. . R--- \ § & 0 I-~~~~~.--~~~~--.~~~~~-, 0 20 40 60 Time (min) \ ,I Figure 3.12 Time-course of the formation of acid-soluble peptides from BSA after -( incubation with the middle (open circles) and posterior (closed circles) mucosal ,.. homogenates (Section 3.4.2.3). 104 3.4.2.4 Distribution of proteolytic enzymes The distribution of proteolytic enzymes was determined in the middle and posterior mucosa. Mucosal homogenates (12.5 µg of protein) were assayed for specific proteolytic activities towards synthetic substrates at pH 7.4, 25°C (Sections 3.3.2.3 & 3.3.2.4). Specific activities towards the synthetic substrates for trypsin, chymotrypsin, elastase and ·; leucine aminopeptidases are shown in Table 3.10. Posterior tryptic-like and chymotryptic-like activity was 2.5- and 6-fold higher than in the middle mucosa (p < ·( 0.001), respectively. Leucine aminopeptidase activity in the middle and posterior mucosa .,, were similar (p > 0.05) and no elastase proteolytic activity was found in the middle or posterior mucosa. Table 3 .10 Specific activities of the middle and posterior mucosal homogenates towards the synthetic substrates for trypsin, chymotrypsin, elastase and leucine aminopeptidase. Specific activitiesa Intestinal Trypsin-like Chymotrypsin Elastase-like Leucine section -like aminopeptidase ( \ Middle 2.6 ± 0.06 80 ± 21 0 53 ± 1.5 i Posterior 6.2 ± 0.5 470 ± 26 0 57 ± 1.1 a Units of activity µmol /min/ g of protein (mean± s.d., n=3). ) ,' ·r 105 3.4.3 Inhibition of proteolytic activity 3.4.3.1 Inhibition of proteolysis by bile salts The effect of bile salts on the proteolytic activity of intestinal lumenal extracts were I tested at bile salt concentrations of 2 mM and 5 mM. Bile salts at concentrations above the critical micelle concentration (CMC) have the potential to trap lipophilic enzyme substrates inside the micelles, resulting in reduced enzyme hydrolysis. The CMC of I deoxycholic acid (DOC) in 50 mM sodium phosphate buffer pH 8.0 was determined using a surface tension method (Section 3.3.1.7). Figure 3.13 shows the plot of surface ( tension against the log concentration of DOC. The CMC of DOC was determined graphically from the intersection of the extrapolated linear portions of the data. The CMC of DOC in 50 mM sodium phosphate buffer pH 8.0 was 3 mM. 62 1 • 60 (_" i I • .\ 58 i-._; s:: .....0 56 I Cl) • s:: 54 .....Q) Q) (.) 52 I ~ • ~ 50 Cl) ~ •• l 48 • • • • -r 46 -1 -0.5 0 0.5 1 1.5 2 ~ Log concentration (mM) Figure 3.13 Change in surface tension with concentration of sodium deoxycholate in . , 50 mM sodium phosphate buffer pH 8.0 (Section 3.3.1.7) . DOC at concentrations above and below the CMC was effective in inhibiting salmon J,, lumenal proteolytic activities of trypsin, chymotrypsin and aminopeptidase as shown in Table 3.11 (p < 0.005). Samples of lumenal homogenates (15-20 µg of protein) were incubated in the presence and absence of DOC in 50 mM sodium phosphate, pH 8.0 (50 106 µl total volume) for 30 min at 25°C and were assayed for specific proteolytic activities (Sections 3.3.2.3 & 3.3.2.4). The rank order of DOC inhibition of these enzymes followed: leucine aminopeptidase > trypsin > chymotrypsin. DOC at 2 and 5 mM was ( I,, ineffective in inhibiting elastase proteolytic activity. Table 3 .11 Effect of sodium deoxycholate on proteolytic activities of the intestinal lumenal contents. f Inhibition of proteolytic activity in intestinal lumenal sections (% inhibitiont Proteolytic DOC(mM) Anterior Middle Posterior , ' activity Tryptic 2 43 ±2.0 21 ± 1.7 10 ± 0.5 5b - - - Chymotryptic 2 11 ± 1.0 8 ± 0.8 11 ± 1.2 5 10 ± 0.8 8 ± 1.2 10 ± 0.9 \ Elastase 2 0 0 0 ~ 5 0 0 0 ) ,) Leucine 2 100 100 100 aminopeptidase 5 100 100 100 a data are means± s.d., n=3. I , b substrate precipitated in the presence of bile salt ·t Table 3.12 shows that DOC at 2 mM and 5 mM significantly inhibited the intestinal 'r lumenal activity towards BSA in all intestinal sections (p < 0.005). Samples of the anterior (13.77 µg of protein), middle (16.5 µg of protein) and posterior (17 µg of protein) intestinal lumenal homogenates were assayed for general proteolytic activity (Section 3.3.2.1) in 50 mM sodium phosphate buffer pH 8.0 in the absence and presence of DOC at 2 mM and 5 mM. The inhibition of proteolytic activity at 5 mM was significantly higher than at 2 mM in all intestinal sections (p < 0.01). 107 Table 3 .12 Percentage inhibition of general proteolytic activity of lumenal homogenates in the presence of 2 mM and 5 mM sodium deoxycholate. Intestinal section Inhibition of proteolytic activity in the intestinal lumenal sections (% inhibitiont Control0 DOC (2mM) DOC (5 mM) Anterior 2599 ± 241 40 ± 2.5 45 ±2.6 { Middle 2226 ± 262 46 ± 1.9 54 ± 3.5 -/ Posterior 1088 ± 73 · 29 ± 1.5 40 ±4.6 a data are means± s.d., n=3. b Specific activity µg / h / mg of protein (mean± s.d., n=3). ~ 3.4.3.2 Inhibition of proteolysis by chelators The divalent ion chelators, Carbopol934P (CP934P), citric acid and EDTA had variable effects on the inhibition of the proteolytic activities of the intestinal lumen (Table 3.13). Samples of the anterior (13.77 µg of protein), middle (16.5 µg of protein) and posterior \ ( (17 µg of protein) intestinal lumenal homogenates were assayed for general proteolytic activity (Section 3.3.2.1) in 50 mM sodium phosphate buffer pH 8.0 in the absence and presence of CP934P 0.35 % (w/v), citric acid (10 mM) and EDTA (2 mM). The rank order of the effectiveness of the chelators followed EDTA > CP93P > citric acid (Table 3.13). The formation of acid-soluble peptides in the presence of CP934P followed pseudo-zero order kinetics, in contrast to the autolytic profile described by LueBen et al. T ; (1995). In comparison to CP934P, EDTA was several-fold more effective as an inhibitor .( of general proteolytic activity and most effective in the posterior intestinal region (Table 3.13). When lumenal extracts (15-20 µg of protein) were assayed for trypsin, ) chymotrypsin and elastase activity at pH 8.0, 25°C in the presence of 2 mM EDTA no ·l inhibition of activity was observed (Sections 3.3.2.3 & 3.3.2.4). Under the same conditions the leucine aminopeptidase activity of the lumen was completely inhibited. ' j 108 Table 3 .13 Percentage inhibition of general proteolytic activity of the intestinal lumen in the presence of CP934P, citric acid and EDTA. Intestinal Controt Percentage inhibitionu section CP934P Citric acid EDTA Anterior 2599 ± 241 7±2 7 ± 1.5 15 ±2 Middle 2226 ± 262 14± 3 2 ± 0.5 16 ± 1.5 Posterior 1088 ± 73 11 ±2 3 ± 1.2 30 ± 2.5 I a Specific activity µg / h / mg of protein (mean± s.d., n=3) b data are means± s.d., n=3 . ./ l, 3.4.3.3 Effect of soybean trypsin inhibitor The effect of the SBTI on proteolytic activities of the intestinal lumen was studied. The effect of SBTI on the general proteolytic activity could not be determined as the U ·1 formation of acid-soluble peptides was masked by the polypeptide inhibitor itself. The inhibition of lumenal trypsin, chymotrypsin and elastase by soybean trypsin inhibitor was \ determined using their respective synthetic substrates (Sections 3.3.2.3 & 3.3.2.4) . . !, Anterior, middle and posterior sections (10 µg of protein) were incubated with 0.5 and I I_ i ;'1 1.0 mg soybean trypsin inhibitor (200 µl total volume) at 15°C for 10 min. Aliquots (50 µ1 sample) were assayed for trypsin, chymotrypsin and elastase activity at pH 8.0, 25°C (Sections 3.3.2.3 & 3.3.2.4). Trypsin and chymotrypsin activity was completely inhibited I in all intestinal sections at both concentrations of the inhibitor; however SBTI was only - partially effective in inhibiting elastase activity (Table 3.14). > ;,; Table 3.14 Percentage inhibition oflumenal elastase proteolytic activity by soybean \ trypsin inhibitor at 1 :50 and 1: 100 intestinal extract : soybean trypsin inhibitor. Intestinal section Percentage inhibitiona 1:50 1:100 Anterior 61 ± 2.0 65 ± 2.5 L \ .) Middle 55 ± 2.4 62 ± 2.0 Posterior 66 ±2.2 56 ± 1.5 a Percentage activity inhibited (mean± s.d., n=3). 109 3.4.3.4 Inhibition of lumenal LHRH proteolysis The effects of various inhibitors on the lumenal proteolytic activity towards human and salmon LHRH (0.2 mM) were determined. The inhibitors EDTA (2 mM), SBTI (2.0 mg/ml), DOC (5 mM) and CP934P (0.35 % w/v) were incubated with the lumenal homogenates (1.25 µg of protein) in 50 mM sodium phosphate buffer pH 8.0 at 15°C (Section 3.3.2.2). Figures 3.14 & 3.15 show the half-lives of human and salmon LHRH in the intestinal sections in the absence and presence of various protease inhibitors. >i .\ ,\ ,\ ;v ;( \ 110 Control (A) EDTA 28.6 N.S. SBTI 327 ***** DOC 54 ***** i CP934P ).2 ** CP934P/SBTI 287 ***** CP934P/DOC 38 *** 0 100 200 300 400 500 Control 26 (B) EDTA 29.4 N.S. SBTI 330 ***** :\ ·DOC 30 N.S. CP934P 70 ** <· CP934P/SBTI 240 ***** CP934P/DOC 26 N.S. ,.,I 0 100 200 300 400 500 ·I' Control '--- 60 (C) EDTA 65 N.S. ,! SBTI ***** 450 :{ DOC 82 ** CP934P 107 **** CP934P/SBTI 250 ***** CP934P/DOC 76 * ' I --, 0 100 200 300 400 500 Half.life (min) Figure 3.14 Half-lives ofhLHRH in anterior (A), middle (B) and posterior (C) lumenal ' ( :,";- homogenates in the absence (control) and presence ofEDTA, SBTI, DOC and CP934P. Half-life represents the mean time (min± s.d., n = 3) for the LHRH concentration to drop to half its initial concentration. N.S. not significantly different from control,(*) p < 0.01, (**) p < 0.005, (***) p < 0.0025, (****) p < 0.001. (*****) p < 0.0005, compared to control by Student's t test. 111 Control (A) EDTA SBTI 301 ***** DOC 39 ***** CP934P 38.5 ***** CP934P/SBTI 268 ***** CP934P/DOC + 44 ***** 0 100 200 300 400 500 1- Control (B) EDTA N.S. SBTI 330***** DOC CP934P 47.5 ***** CP934P/SBTI 266 ***** CP934P/DOC + 0 100 200 300 400 500 ., --Z Control 50.5 (C) EDTA 64 * :, SBTI 380 ***** DOC 78 *** CP934P 98 ***** CP934P/SBTI 200 ***** CP934P/DOC 76 **** ;, + 'I 0 100 200 300 400 500 Half:. life (min) ' ,i Figure 3.15 Half-lives of sLHRH in anterior (A), middle (B) and posterior (C) lumenal homogenates in the absence ( control) and presence of EDTA, SBTI, DOC and CP934P. Half-life represents the mean time (min± s.d., n = 3) for the LHRH concentration to drop to half its initial concentration. N.S. not significantly different from control,(*) p < 0.01, (**) p < 0.005, (***) p < 0.0025, (****) p < 0.001, compared to control by Student's t test. 112 Human and salmon LHRH analogues were most stable in the presence of soybean trypsin inhibitor, which increased the half-lives of the LHRH analogues between 7.5- to 15-fold in the anterior, middle and posterior intestinal sections. CP934P increased the half-lives of the LHRH analogues in the anterior, middle and posterior sections between 1. 7 and 2. 7-fold. Figure 3 .16 shows a typical peak height versus time profile of human LHRH incubated in the anterior lumenal homogenate in the presence and absence of CP934P. The rank order of effectiveness of protease inhibitors at the concentrations tested to ·1 protect human and salmon LHRH analogues in the lumen of the anterior, middle and posterior intestinal sections followed SBTI > CP934P >DOC> EDTA. 1500 I 1:l I oJ) ' ] 1000 ~ ., Cl) "\ p.. 500 ,\ ( 0 0 50 100 150 200 250 Time (min) _.,. Figure 3.16 Time-course ofhLHRH proteolysis when incubated with the anterior lumenal homogenate in the presence ( open circles) and absence of CP934P (0.35% w/v) ( closed ,, circles) at pH 8.0,15°C (Section 3.4.3.4). Each symbol represents the peak height (mean± s.d., n=3). 113 3.4.3.5 Inhibition of mucosal proteolytic activity Since both human and salmon LHRH analogues were stable in the middle intestinal mucosa homogenates (Section 3.4.2.2) the effects of various protease inhibitors on stabilising the LHRH analogues were only tested in the posterior mucosa! homogenate. The effects of SBTI (2 mg/ml), DOC (5 mM), bestatin (3.5 mM) and EDTA (2.0 mM) on the rates of hydrolysis of human and salmon LHRH analogues in the posterior mucosal homogenates (50 µg/ml) are shown in Figure 3.17 (Section 3.3.2.2). The human and salmon LHRH analogues were markedly stabilised by EDTA, DOC and the aminopeptidase inhibitor bestatin but not by the serine protease inhibitor SBTI. '/ \ \ \ 114 (A) Control SBTI DOC Bestatin ** { 210** EDTA + I 0 50 100 150 200 250 300 l Control (B) ,I SBTI :1 DOC 185 ** I ", Bestatin 170 ** EDTA '-, ! 200 ** 0 50 100 150 200 250 300 Half-life (min) ., Figure 3.17 Half-lives of hLHRH (A) and sLHRH (B) in posterior mucosal homogenates ,> in the absence (control) and presence ofEDTA, DOC, bestatin and SBTI. Half-life represents the mean time for the concentration of LHRHs to fall to half its original concentration (min± s.d., n = 3). N.S. not significantly different from control, (**) p < 0.005, compared to control by Student's t test. 115 3.5 Discussion 3.5.1 Lumenal proteolytic activity The intestine of the salmon was divided into three anatomically distinct sections: anterior, middle and posterior as described in (Section 1.4.1). There was no significant difference in the lumenal pH of the anterior, middle and posterior intestinal sections (Section 3.4.1.1 ). Consequently the lumenal proteolytic activities of the three intestinal sections were determined at pH 8.0 for consistency. Previously, Ash (1980) measured the pH of the lumenal contents of trout and reported lower pH values for the anterior section of7.5- f:- ,'( 7.6 with similar pH values for the middle (7.8-8;0) and posterior of (7.8-8.4) sections. Supematants of lumenal homogenates that were immediately snap frozen in liquid nitrogen and stored at -80°C had the highest lumenal proteolytic activity towards BSA (Section 3.4.1.2). Therefore all subsequent homogenates were stored in this way for metabolic analysis. There was no significant difference in the general proteolytic activity or protein content ,•\' between salmon which had no food released into their tank either 12 h or 24 h prior to .;, collection (p > 0.1) (Sections 3.4.1.3 & 3.4.1.4). Previous studies by Einarsson et al. (1996) reported that Atlantic salmon starved from 2 days onwards resulted in significantly lower levels of trypsin and chymotrypsin activity in the intestinal lumen. It would appear that the extra 12 h period without release of feed was not long enough to a ., cause significant decrease in the lumenal proteolytic activity. The relative standard deviation of the general proteolytic activities of combined intestinal sections from the 12 and 24 h fasted groups were 57 and 54%, respectively. Therefore the longer period without feeding did not improve the reproducibility of proteolytic activities among individual fish. In both fasted groups of fish there was a significant difference between both the specific proteolytic activities and total proteolytic activities of the anterior, middle and posterior intestinal sections (p < 0.001). The proteolytic activity and the total proteolytic activity towards BSA in the posterior intestinal section was significantly lower than the anterior 116 (p < 0.05) and middle (p < 0.1) intestinal sections (Section 3.4.1.4). A large biological variation in proteolytic activities was observed among individual fish. The RSD for the anterior, middle and posterior general proteolytic activity were 62, 43 and 59%, l respectively. Potentially, this high biological variability in proteolytic activity among fish will effect the oral bioavailability of protein and peptide drugs among salmon. The observation that proteolytic activity decreases distally along the intestinal tract was I further supported by ranking the proteolytic activities and total activities of individual salmon (Section 3.4.1.4). The rank order anterior> middle> posterior was observed in 5 I I out 10 salmon for proteolytic activities and for 7 out of 10 for total activities. Since the ,l probabilities these frequencies (or greater) occurring randomly are 1 in 67 and 1 in 3750 then we can conclude that the posterior intestine has the lowest proteolytic activity towards BSA. ,, :1 The progressive decrease in lumenal proteolytic activity towards the anus has been observed in a number of other fish species (Hofer and Schiemer, 1981, Hofer, 1982), \ including trout (Krogdahl et al., 1994). In trout the lumenal tryptic activity towards the synthetic substrate benzoyl-L- Arg- p-nitroanilide of the middle intestine was 3-fold \ greater than in the posterior intestine (Krogdahl et al., 1994). The lower levels of l > activated pancreatic endopeptidases in the lower intestinal lumen have been investigated in a number of mammals, with conflicting conclusions. Investigators believe that the / lower levels of activated pancreatic enzymes are due to, autolytic degradation (Roy et al., F ·r 1967, Bohe et al., 1983) enteropancreatic circulation (Beyon and Kay, 1976, Diamond, 1978) inactivation by the intestinal microflora (Genell et al., 1977) or binding to I, 'I intestinal particles (Goldberg et al., 1969). In vitro experiments with the intestinal lumenal fluid of roach showed only small losses of proteolytic activity during the time corresponding to the natural gut passage times, suggesting that the progressive decrease in proteolytic activity along the fish intestine is not due to autolysis (Hofer and Schiemer, 1981). The pancreatic endopeptidases, trypsin, chymotrypsin and elastase were expected to be the major proteolytic enzymes involved in the metabolism of the LHRH analogues in the 117 intestinal lumen. Human and salmon LHRH analogues are C-terminally and N terminally blocked and therefore were not expected to be hydrolysed by the exopeptidases carboxypeptidases A and B or the aminopeptidases (MacCann, 1977). The possible cleavage sites for the pancreatic endopeptidases are summarized in Figure 3 .18. Trypsin activity is specific towards lysine and arginine residues (Keil-Dlouha et al., 1971 ), chymotrypsin activity is specific towards phenylalanine, tryptophan and tyrosine (Baumann et al., 1970), whereas elastase has a broad specificity hydrolysing adjacent to alanine, serine, glycine and valine amino acid residues (Atlas et al., 1970). ( C E E T Human LHRH pGlu-His-Trp-!.--serl.·-Tyrt-G1y-!.--Leu-Arg-!.--Pro-Gly-NH2 C C E E Salmon LHRH pGlu-His-Trp-!.--ser-!.--Tyrt-G1y-!.--Trp-Leu-Pro-Gly-NH2 C Figure 3.18 Potential cleavage sites of human and salmon LHRH by trypsin (T), '., chymotrypsin (C) and elastase (E). ·( In order to determine the relative contribution of each of the pancreatic endopeptidases to ;- the lumenal degradation of the LHRH analogues, bovine trypsin (TPCK-treated), bovine ·r a-chymotrypsin (TLCK-treated) and porcine elastase were incubated at 15°C at pH 8.0 >" with 0.1 mM of human and salmon LHRH (Section 3.4.1.5). Under these conditions only .', chymotrypsin and elastase hydrolysed the LHRH analogues. Chymotrypsin rapidly degraded both human and salmon LHRH analogues with half-lives (min) of 2.5 ± 0.3 and 2.7 ± 0.4 min (mean± s.d., n = 3), respectively. Human and salmon LHRH were :( considerably more stable in the presence of elastase with half-lives (min) of 90 ± 15 and 114 ± 21 (mean± s.d., n = 3), respectively. The metabolic stability of salmon LHRH analogue to trypsin was expected as there are no potential cleavage sites for trypsin; however, for human LHRH the Arg8 residue is a potential cleavage site for trypsin (Figure 3.18) and yet the hLHRH was stable to trypsin. The active sites of the 118 endopeptidases can be divided into sul;,sites which bind residues of the peptide substrate on either side of the scissile bond. Schechter and Berger (1967) introduced a binding site notation to indicate the position of the amino acid residues in relation to the enzyme cleavage site described by Figure 3.19. Using a series of peptides Schellenberger et al. (1993) showed that the presence of a Pro residue in the P' 1 site of the substrates is not tolerated in the S' 1 subsite of trypsin and thereby prevents the catalytic acyl transfer reaction. This supports the stability ofhLHRH to trypsin. acyl enzyme j. C S3 S2 S1 I S1' Si S3' COOH Figure 3 .19 Schematic representation of enzyme-substrate complex. The active subsites (S) of the enzyme are located on both sides of the catalytic site (C) of the substrate (P). \ The primary chymotryptic and elastase metabolites generated from the respective human and salmon LHRH analogues had similar electrophoretic mobilities (Section 3.4.1.5). Figure 3 .18 shows that the predicted :fragments generated by elastase cleavage would differ by only one small unionised amino acid to that predicted for chymotrypsin. This small increase in molecular weight may not alter the electrophoretic mobility sufficiently to resolve a different cleavage pattern from elastase digestion compared to chymotrypsin. ·( Elastase proteolytic activity towards the LHRH analogues was dramatically lower than the proteolytic activity observed by chymotrypsin. Using synthetic substrates it was shown that the specificity of elastase is reduced greatly by large amino acids at the P' 1 residue subsite (Schellenberger et ai., 1989). According to Scheme 3.1 the P'l side chains of the predicted elastase cleavage sites of human LHRH include tyrosine and leucine while the corresponding P' 1 side chains of salmon LHRH include tyrosine and tryptophan. Both tyrosine and tryptophan amino acids have large bulky side chains and 119 therefore are expected to reduce the subsite specificity of the porcine elastase for those sites. The low proteolytic activity observed by the porcine elastase could alternatively be attributed to contaminant chymotrypsin activity, as the elastase preparation has no specific inhibitor of chymotrypsin. These results indicate that chymotrypsin is the major metabolic barrier of the intestinal lumen to LHRH. The rates of proteolysis of human and salmon LHRH by the intestinal lumen were similar and showed regional differences within the lower intestinal tract of salmon. The rate of -( degradation of human and salmon LHRH analogues in the anterior and middle intestinal J, sections were approximately 2-fold higher than in the posterior intestinal section (p < 0.01). However the electrophoretic profiles of human and salmon LHRH digests were the same in each intestinal section. The generation of the same metabolic fragments from human and salmon LHRH in each intestinal section suggests that the same proteolytic enzymes are involved in the proteolysis of LHRH. The similar degradation rates of human and salmon LHRH in the different intestinal sections suggests that both LHRH analogues are hydrolysed by the same protease. The primary metabolites of the hLHRH lumenal digests (Figure 3.7, B-D) had a similar electrophoretic pattern to the a-chymotrypsin digest ofhLHRH (Figure 3.2, B). This suggests that a-chymotrypsin is ,) the major intestinal enzyme for the metabolism of both LHRH analogues. The electrophoretic pattern of the sLHRH digest by lumenal enzymes (Figure 3.8) was much more complex than for a-chymotrypsin digest of sLHRH (Figure 3.3). This suggests that lumenal proteases other than a-chymotrypsin are involved in the hydrolysis of sLHRH. ;> However this observation does not preclude a-chymotrypsin as being the primary protease involved in the hydrolysis of sLHRH. The lumenal proteases involved in generating the new metabolic fragments may only be active on the primary metabolites generated by a-chymotrypsin. By the use of specific synthetic substrat.es the distribution of the major proteases of the intestinal lumen was surveyed (Section 3 .4.1. 7). Because each enzyme has different catalytic activities towards its synthetic substrate, specific activities of the different proteases are not comparable. However this work showed that a wide range of proteolytic 120 enzyme activities exist in the intestinal lumen of salmon and serves as evidence of the ability of the lumen to digest a wide range of protein and peptide drugs. The proteolytic activities observed include trypsin, chymotrypsin, elastase, carboxypeptidase A, carboxypeptidase Band leucine aminopeptidase. Most of the proteolytic activities were shown to reside in the anterior and middle intestinal sections (Figure 3.10). The higher proteolytic activities in the anterior and middle sections correlate with the higher proteolytic activities towards BSA and the LHRH analogues observed in these regions. The effects of inhibitors in stabilising the LHRH analogues in the lumenal homogenates ·, were variable (Section 3.4.3.4). The relative potency of each inhibitor was similar in each intestinal section, supporting the previous observation that the same proteases are involved in the hydrolysis of the LHRH analogues in each intestinal section. Soybean trypsin inhibitor (SBTI), a known inhibitor of mammalian trypsin and a-chymotrypsin activity was shown to completely inhibit lumenal chymotrypsin and l trypsin activity of salmon (Section 3.4.3.3). SBTI also inhibited lumenal elastase activity by 55-65% of control activity. SBTI was the most effective inhibitor for stabilising the •,\' LHRH analogues against lumenal proteolytic attack. This indicates that the serine \ ·I endopeptidases are the principle proteases involved in LHRH hydrolysis in the salmon intestinal lumen. / Using synthetic substrates, it was observed that chymotrypsin and trypsin activities of the salmon intestinal lumen are also inhibited by DOC. Metabolic studies with purified pancreatic enzymes (Section 3.4.1.5) suggest that the inhibition ofLHRH hydrolysis by DOC is due to the inhibition of chymotrypsin activity. The divalent ion chelator EDTA, has no effect on the trypsin, chymotrypsin and elastase activities of the salmon intestinal lumen (Section 3.4.3.2). Likewise EDTA was ineffective in stabilising the LHRH analogues in the intestinal lumen of salmon (Figures 3.14 & 3.15). The divalent ion chelators EDTA and CP934P were effective in inhibiting lumenal proteolytic activity towards BSA (Section 3.4.3.2), indicating the involvement ofmetalloproteinases (Lanzillo et al., 1986, Han et al., 1995). EDTA was a more effective inhibitor than 121 CP934P. EDTA inhibition in the posterior intestine was 2-fold greater than the anterior and posterior sections (Table 3 .13), suggesting a higher concentration of metalloproteinases reside in the posterior intestine. CP934P was effective in reducing the intestinal lumenal proteolytic activity towards the LHRH analogues (Section 3.4.3.4). The half-lives of the LHRH lumenal hydrolysis were increased between 1. 7 and 2. 7-fold in the presence of CP934P. The rates of hydrolysis in the presence of CP934P followed pseudo-zero order kinetics (Figure 3.16). This is in "I, contrast to work by Junginger and co-workers (LueBen et al., 1995 and 1996c) who reported the rates of cx-chymotrypsin and trypsin hydrolysis in the presence of CP934P followed second-order kinetics, indicative of enhanced enzyme autolysis. They postulated that CP934P chelates essential divalent ions from the proteases, resulting in the opening the protease structure to autolytic attack. The comparatively lower inhibition of salmon lumenal proteases by CP934P suggests that the lumenal proteolytic enzymes of the salmon intestine may either have a lower requirement for divalent ions or a higher L affinity for their divalent ions compared to the mammalian proteases (Section 1.4.2.1). " i 4.5.2 Mucosal proteolytic activity The posterior mucosal homogenates prepared from fresh intestines and snap frozen in liquid nitrogen gave the higher hLHRH proteolytic activity compared to mucosa prepared from frozen intestine and assayed immediately after homogenisation (Section 3 .4.2.1 ). \ Subsequently mucosal homogenates were prepared from fresh intestine and snap frozen in liquid nitrogen for later storage at -80°C. The degradation of BSA and human and salmon LHRH by intestinal mucosa homogenates was section dependent (Sections 3.4.2.2 & 3.4.2.3). In contrast to lumenal proteolytic activities the mucosal activities were found to be higher in the posterior intestinal section. Interestingly, both of the LHRH analogues were metabolised in the posterior mucosa! homogenate but not by the middle mucosa. Previous studies have shown that due to presence ofN terminal pyrogulatamate residues and C terminal amide 122 bonds in LHRH analogues, these peptides are resistant to degradation by most exopeptidases such as amiriopeptidases and carboxypeptidases (MacCann, 1977). The proteolytic activity of the lysosomal enzymes of the enterocytes should be negligible at pH 7.4, as they are most active in an acidic environment (Barrett and Heath, 1977). Capillary electrophoretograms of the posterior mucosal digests of human and salmon LHRH analogues showed the formation of multiple metabolic fragments (Figure 3.17). The large number of metabolites suggests that there are a large number ofLHRH bonds susceptible to the proteases of the intestinal mucosa. Much of this activity could be attributed to the exopeptidases hydrolysing the fragments generated by the principal cleavage by the endopeptidases. Although it is possible that amidases and pyroglutamidases are also present in the mucosa! homogenates . . The proteo.lytic activity of the posterior mucosa towards LHRH corresponded to the higher activity towards the ·synthetic substrate of chymotrypsin, N-benzoyl-L-Tyr ethyl ester (Section 3.4.2.4). This hydrolytic activity is most likely due to the activity ofbrush border membrane endopeptidase 24.18, which has previously been shown to hydrolyse )• N-benzoyl-L-tyrosyl-p-amino benzoic acid, in the presence of the specific chymotrypsin inhibitor TPCK (Sterchi et al., 1982). In vitro, endopeptidase 24.18, has been shown to hydrolyse hLHRH at three bonds: Trp3-Ser4, Ser4-Tyr5 and Tyr5-Gly6, with Trp3-Ser4 being the preferred site (Stephenson and Kenny, 1988). The LHRH hydrolytic activity of the posterior mucosa may also be due to the brush-border membrane peptidases endopeptidase 24.11 and angiotensin converting enzyme (ACE) which have been shown to cleave hLHRH at the Gly6-Leu7 bond (Lasdun et al., 1989) and the Trp3-Ser4 site (Han . et al., 1995), respectively. Amino acid analysis ofLHRH fragments generated from rabbit rectal mucosa in vitro, indicated the involvement of endopeptidase 24.11, ACE and endopeptidase 24.15 in the hydrolysis ofhLHRH (Han et al., 1995). The HPLC profile showed that endopeptidase· 24.15 is the primary degrading enzyme in rabbit rectal mucosa (Han et al., 1995). 123 Digestion of hum.an and salmon LHRH in mucosal homogenates was significantly inhibited by DOC, EDTA and the aminopeptidase specific inhibitor bestatin (Section 3.4.3.5). SBTI a serine endopeptidase inhibitor, did not inhibit the hydrolysis ofLHRH analogues in the posterior mucosal homogenate, indicating the absence of the serine i ' proteases in the mucosal preparations. Contrary to previous results by MacCann (1977) inhibition by bestatin indicates that aminopeptidases are involved in the hydrolysis of the LHRH analogues. "., II' EDTA, a metal chelator was the most effective inhibitor ofmucosal proteolytic activity towards LHRH analogues (Section 3.4.3.5). EDTA is known to inhibit metalloproteases such as carboxypeptidases, ACE, aminopeptidases and endopeptidase 24.15 (Lanzillo et al., 1986, Han et al., 1995). Of these ACE and endopeptidase 24.15 have been shown C\ previously to hydrolyse hLHRH (Han et al., 1995). DOC was reported to be an inhibitor ofleucine aminopeptidases (Hirai et al., 1981) and endopeptidase 24.11 (Bai, 1994). As endopeptidase 24.11 has been reported to hydrolyse hum.an LHRH (Han et al., 1995), the effect of DOC is therefore most likely due to the inhibition of this enzyme. The inhibitory .effect of the divalent ion chelator CP934P was not tested on the mucosal \ ( homogenates, since CP934P was previously shown to be ineffective in inhibiting the -.. ·'\ proteolytic activity of the rat brush-border membrane vesicles (BBMV) towards the LHRH analogue buserelin (LeuJ3en et al., 1996a). It was suggested that the lack of proteolytic inhibition by CP934P was due to the proteases being embedded or attached to the epithelial membrane, thereby reducing the accessibility of the protease divalent ions "'. ,;i- for chelation (LeuJ3en et al., 1996a). Alternatively, the lack of inhibition ofBBMV proteolytic activity by CP934P maybe unrelated to chelation. This possibility is explained in Chapter 4. Since EDTA and DOC when used as proteolytic inhibitors have a broad inhibitory specificity, it is not possible to determine which specific enzymes are involved in the hydrolysis of the LHRH analogues. Amino acid analysis of metabolites and the use of 124 s ' specific inhibitors should make it possible to identify the cleavage sites of the LHRH and identify the proteolytic enzymes involved. i !; ') '+ i ~ "/ ':), l\ > 125 Chapter 4: Polyacrylic Acid Inhibition Studies ·! ) '\ ,) \s . .., .'\ 1 ,} .\ ·, :, - -..... ,,- 126 4.1 Introduction ' , In vitro metabolic studies showed that the polyacrylic acid Carbopol934P® (CP934P) increased the enzymatic stability ofLHRH analogues in the intestinal lumenal homogenates from salmon (Section 3.4.3.4). -( t Two mechanisms have been proposed by which acrylic acid polymers inhibit the proteolytic activity of enzymes (Section 1.3 .1.5). Bai et al (1995b) demonstrated that -..{ ~· the degree of proteolytic inhibition by polyacrylic acids correlated with their ability to ~ :, reduce the pH of incubates below the optimum pH of the proteolytic enzymes. Bai et al. (1995b) however speculated that with the constant secretion of bicarbonate and carbonate ions into the intestinal lumen the polyacrylic acids would have limited ability to reduce the local pH in vivo. Under buffered conditions at pH 6.7, dispersions ofpolyacrylic acids were shown to inhibit the activity of trypsin, a-chymotrypsin and carboxypeptidase A and Band the cytosolic leucine aminopeptidase (Lue.Ben et al., 1996c). The rate of proteolysis in the presence of polymer dispersions followed second-order kinetics and the authors .\ speculated that this was due to autolysis (LueBen et al., 1996c). The proposed ~ l mechanism of inhibition was that the polyacrylic acids bind the essential enzyme co factors calcium and zinc from the enzyme, causing a conformational change resulting in enhanced enzyme autolysis and loss of enzyme activity (Lue.Ben et al., 1995). In ."'r tr contrast, initial studies with salmon lumenal homogenates showed that the reduced rate ofLHRH hydrolysis in the presence of CP934P followed pseudo-zero kinetics (Section -/ 3.4.3.4). This chapter reports that the CP934P dispersions at pH 7.0 inhibited the proteolytic activity of bovine trypsin and bovine a-chymotrypsin. In the presence of CP934P, the rates of hydrolysis of the proteolytic substrates followed pseudo-zero order kinetics as observed in Section 3.4.3.4. It was found that CP934P reduces proteolytic activity by interacting with the proteolytic enzyme, thereby identifying another mechanism by which polyacrylic acids may improve the metabolic stability of peptides in the intestinal 127 lumen. A kinetic model is proposed to describe the possible routes for protease inactivation by polyacrylic acids. ' 0 ·i ,,, ·\ \ ·\ •( ;r J.; .-\ ). >, \ 128 4.2 Materials ' > Carbopol ® 934P (C934P) was gifted from BF Goodrich (Cleveland, OH, USA). Bovine trypsin (TPCK treated, Type XIII), bovine a-chymotrypsin (TLCK treated, Type XII), porcine pepsin (A, P-7125), N-a-benzoyl-L-Arg ethyl ester (BABE), N-a ·' ) benzyol-L-Tyr ethyl ester (BTEE), soybean trypsin-chymotrypsin inhibitor (Bowman Birk inhibitor), human luteinizing hormone-releasing hormone (hLHRH) and mineral oil (heavy white oil) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). -~ •)l Bovine serum albumin fraction V (BSA) was purchased from Serva (Feinbiochemica GmbH & Co., Heidelberg, Germany). Sample tubes of 0.5 ml and 1.5 ml were purchased from Eppendorf (Eppendorf-Netheler-Hinz-GmbH Hamburg, Germany). Ultrafree® - MC (0.45 µm, 0.2 cm2 membrane area) centrifugal filtration units were purchased from Millipore Corporation (Bedford, MA, USA). All other chemicals were of analytical grade and were purchased from Ajax Chemicals (Auburn, NSW, Australia). 4.3 Methods '1 4.3.1 Analytical methods 4.3.1.1 Protein estimations " Protein was determined by the method of Lowry as modified by Peterson (1977); TCA " 1~ precipitation was not required. Assays were performed in microtitre plate wells in a total volume of 250 µ1, and the absorbance (750 nm) was read using a Microplate Spectrophotometer (Spectra MAX 340) controlled by an external computer with SOFT max PRO software (Molecular Devices Corporation, CA, USA). Protein was determined with standards_ of the protein covering the range 1-10 µg. 129 4.3.1.2 BSA hydrolytic activity Bovine trypsin (1 mg) was prepared in 50 mM sodium phosphate buffer pH 7.0 (1 ml total volume). Trypsin solution was transferred to a 1.5 ml sample tube containing 50 mM sodium phosphate buffer pH 7.0 with or without CP934P (total volume 0.5 ml), pre-equilibrated at 37°C. After 30 min BSA (1 mg) in 50 mM sodium phosphate buffer f ,'- pH 7.0 pre-equilibrated at 37°C was added (1 ml total volume). Samples (0.1 ml) were removed immediately after mixing and after 30, 60 and 90 min and transferred to 0.5 ml · sample tubes containing 5·µ1 of 30% (v/v) TCA to stop the reaction and precipitate \ >' intact protein. The samples were centrifuged at 20 000 g for 10 min and the supernatant was analysed for acid-soluble peptides (Section 4.3.1.1). 4.3.1.3 BAEE hydrolytic activity BABE was prepared in a solution of 50 mMsodium phosphate buffer pH 7.0 to give a \ final concentration of 1.0 mM. The enzyme solution (0.1 ml) was transferred to a 1 ml quartz cuvette of 1 cm path length containing 0.1 ml of phosphate buffer. To initiate the reaction, 0.8 ml of the BABE stock was added. Buffer and substrate stock solution were both pre-incubated at 37°C for at least 10 min. The reaction was monitored by a Hewlett Packard 8452A diode array spectrophotometer at 253 nm, with a jacketed cell maintained at 37°C by recirculating water. \ 4.3.1.4 BTEE hydrolytic activity "( The chymotrypsin substrate BTEE (0.059 mmoles) was initially dissolved in 31.7 ml of .· ~ r, methanol then diluted with 50 mM sodium phosphate buffer pH 7 .0 to give a final ' volume of 50 ml and substrate concentration of 1.18 mM (Wirt, 197 4). The enzyme solution (0.05 ml) was transferred to a 1 ml quartz cuvette of 1 cm path length, \. > containing 0.36 ml of phosphate buffer. To i11itiate the reaction 0.59 ml of the synthetic substrate solution was added. Buffer and substrate stock were both pre-incubated at 37°C for at least 10 min. The reaction was monitored by a Hewlett Packard 8452A diode array spectrophotometer at 256 nm, with a jacketed cell maintained at 37°C by recirculating water, using a molar extinction coefficient of 964. The unit of activity is defined as 1 µmole of BABE hydrolysed per min at 37°C. 130 4.3.1.5 Capillary electrophoresis Capillary electrophoresis was carried out in a Dionex Model CES-1 capillary electrophoresis instrument (Dionex, Sunnyvale, CA, USA) under conditions described in Section 2.3.1. Samples were electrophoresed in 10 mM sodium phosphate buffer at pH 2.5 containing 75 mM NaCL Rates ofLHRH proteolysis were reported as changes in peak height of the principal metabolite. 4.3.1.6 HPLC of LHRH The HPLC system consisted of a Shimadzu pump (LC - 6A), a Rheodyne injector with a 50 µl loading loop and a Shimadzu spectrophotometric UV visible detector (SPD-6AV). Data were collected via a Dionex advanced computer interface at 5 Hz and integrated .1 using the Dionex A-450 Chromatography Automation Software v 3.3.2. The column, Phenomenex C18 (250 x 4.60 mm, 5 µm) was equilibrated in a mobile phase of 21 % acetonitrile in 0.1 % trifluoroacetic acid at 1.0 ml/min. The eluate was monitored at 215 nm. All samples were diluted 10-fold in mobile phase before analysis. 4.3.1.7 Gel electrophoresis Sodium dodecylsufate-polyacrylamide gel electrophoresis (SDS PAGE) of trypsin was done using the method ofLaemmli and Farve (1973) using 10% acrylamide separating gels and 5% stacking gels. Samples (0.1 ml) were lyophilized, resuspended in 15 µl of ,, dissociation buffer, boiled for 1 min, centrifuged at 5 000 g for 5 min then loaded onto the gel. Electrophoresis was at constant voltage of 150 V (25 - 30 µA). Gels were ,_ stained by a reductive silver stain (Morrisey, 1981 ). The molecular weight markers were soybean trypsin inhibitor (MW 20 000), trypsinogen (MW 24 000), carbonic anhydrase (MW 29 000), glyceralderhyde-3-phosphate dehydrogenase (MW 36 000) and ovalbumin (MW 45 000). 4.3.1.8 Gel filtration The instrumentation consisted of a Spectra-Physics (SP88 l 0) HPLC pump, a Rheodyne injector with a 100 µl loading loop, Shimadzu SPD-6A UV detector set at 215 nm and a Hewlett Packard (3390A) integrator. Samples (50 µ1) were applied to the gel filtration 131 column (Waters Protein-Pak TM SW, 7.8 x 300 mm) and separated using a mobile phase of 10 mM phosphate buffer pH 6.5 containing 150 mM sodium chloride at 1 mVmin. 4.3.1.9 HPLC of trypsin '( Analysis of trypsin by reverse phase HPLC were conducted using a Spectra-Physics ternary pump (SP8800), Phenomenex on-line mobile phase degasser (DS-4000), Rheodyne injector with a 50 µl loop and Spectra-Physics Spectra 100 variable wavelength detector while data were recorded on Spectra-Physics ChromJet integrator. -~' ) The column, Jupiter C4 (150 x 4.6 mm, 5 µm) were equilibrated in 5% acetonitrile in 0.08% trifluoroacetic acid at 1.0 mVmin. Samples were eluted from the column with a linear gradient of acetonitrile (5-65%, 30 min) in 0.08% trifluoroacetic acid at a flow ,, '\ rate of 1 mVmin. The eluate was monitored at 215 nm. 4.3.2 Inhibition studies The enzyme stability studies were determined in 50 mM phosphate buffer adjusted to ,I pH 7.0 with sodium hydroxide. CP934P 0.5% (w/v) was dispersed in water by low speed stirring, disodium hydrogen phosphate was added to give a final concentration of \ 50 mM and the dispersion was adjusted to pH 7.0 with sodium hydroxide. The I proteolytic enzyme was dissolved in phosphate buffer, pH 7.0 to give a stock solution of \ ,' 1 mg/ml. The enzyme stock solution (10 µI) was mixed with either 0.7 ml of the 0.5% (w/v) CP934P or 0.7 ml of phosphate buffer and incubated at 37°C. After 30 min C( the substrate dissolved in the phosphate buffer pH 7.0 was added to give a final volume ' ~ of 1 ml. The stability of BABE and LHRH to proteolytic digestion was determined at , :,- 20 mM and 0.4 mM, respectively. At various time intervals 50 µl samples were removed and added to 50 µl of 0.2 M HCl to stop the reaction and to precipitate the ' 'r CP934P. Samples were centrifuged (20 000 g for 10 min) and the supematants were analysed. LHRH was analysed by capillary electrophoresis (Section 4.3.1.5). BABE was determined by measuring the formation of N-a-benzoyl-L-Arg from N-a-benzoyl-L Arg ethyl ester, based on the method of Schwert and Takenaka (1955): supematants (50 µl) were transferred to a 1 ml quartz cuvette of 1 cm path length containing 950 µl of 50 mM sodium phosphate buffer pH 7 and the absorbances read at 252 nm, using a Hewlett Packard 8452A diode array spectrophotometer. 132 4.3.3 Binding studies 4.3.3.1 Centrifugal filtration CP934P (0.5% w/v) dispersed in 50 mM phosphate buffer pH 7 .0, was mixed with the proteolytic enzyme (12.5 µg) and LHRH (0.1 µmol) or BABE (5 µmol) to give a final CP934P concentration of 0.35% (w/v) and a total volume of 0.25 ml. In the control experiment, the CP934P was replaced with phosphate buffer pH 7.0. Following incubation at 370c for 30 min, 0.2 ml samples of the suspensions were removed, placed in a Ultrafree® - MC centrifugal filtration units and centrifuged at 5 000 g for 2 h at room temperature. The filtrates were analysed for unbound material. BABE was .1 measured by the change in absorbance at 252 nm and LHRH was measured by capillary electrophoresis (Section 4.3.1.5). The enzyme filtrates were assayed for protein content by the modified Lowry assay (Section 4.3 .1.1) and trypsin or chymotrypsin activity using their respective synthetic substrates, BABE (Section 4.3.1.3) and BTEE (Section 4.3.1.4). The trypsin filtrates were also analysed by SDS-PAGE (Section 4.3.1.7), gel :\ filtration (Section 4.3.1.8) and reverse-phase HPLC (Section 4.3 .1.9). 4.3.2.2 Langmuir isotherm The effect of protein concentration on the binding of trypsin to 0.35% (w/v) CP934P was evaluated over a range of trypsin concentrations (35, 50, 70, 80, 150, 285, and 525 µg/ml) in 50 mM phosphate buffer pH 7.0. CP934P 0.5 % (w/v) in 50 mM phosphate buffer pH 7.0 (0.7 ml) was pre-incubated at 37°C for 10 min, then the appropriate volume of trypsin (2 mg/ml) was added with phosphate buffer pH 7.0 to give a total volume of 1 ml. The dispersion was vortexed and incubated at 37°C for 30 min. The incubates (0.4 ml) were transferred to the centrifugal filtration unit and centrifuged at 5 000 g for 2 h at room temperature. The filtrates were assayed for protein content (Section 4.3.1.1 ). 133 The data obtained were fitted to the linearised form of the Langmuir equation Langmuir !, (1918): c/(x/m) = 1/ab + c/a Equation 4.1 where xis the mass of trypsin abound to CP934P, mis the mass of the CP934P, a is the capacity of CP934P to bind trypsin, b is the affinity constant of trypsin for CP934P, c is ,j the concentration of free trypsin. ·y 4.3.2.3 Effect of ionic strength and pH on the adsorption of trypsin to 1. ' CP934P CP934P 0.5% (w/v) was dispersed in deionised water by low speed stirring. Sodium chloride was added to give final concentrations of 70, 130 and 215 mM and the pH was adjusted with sodium hydroxide. For control experiments the buffer was prepared the :( same but 5 mM disodium phosphate was added instead ofCP934P. Trypsin was dissolved in deionised water to give a stock solution of lmg/ml. The trypsin stock was mixed with either 0.7 ml of CP934P dispersion or 0.7 ml of the control buffer to give a l. final volume of 1.0 ml and incubated at 37°C. After 30 min incubation 0.4 ml samples "'.' of the suspensions were removed and placed in Ultrafree® - MC centrifugal filtration units and centrifuged at 5 000 g for 2 h at room temperature. The filtrates were assayed ,,,. for protein content by the modified Lowry assay (Section 4.3.1.1 ). ( :; 4.3.2.4 Adsorption of proteins to CP934P 1-,. I> .• CP934P 0.5% (w/v) was dispersed in deionised water by low speed stirring. Sodium chloride was added to give final concentrations of 70 mM and the pH was adjusted to pH 7.0 with sodium hydroxide. For control experiments the buffer was prepared the 1·· same but 5 mM disodium phosphate was added instead ofCP934P. Proteins were l, dissolved in deionised water to give a stock solution of lmg/ml. The protein stock was '> mixed with either 0.7 ml of CP934P dispersion or 0.7 ml of the control buffer to give a final volume of 1.0 ml and incubated at 37°C. After 30 min incubation 0.4 ml samples of the suspensions were removed and placed in Ultrafree® - MC centrifugal filtration 134 units and centrifuged at 5 000 g for 2 h at room temperature. The filtrates were assayed for protein content by the modified Lowry assay (Section 4.3.1.1). ',, ', ./ > ·' 135 4.4 Results 4.4.1 Inhibition of trypsin 4.4.1.1 Inhibition studies Initial experiments using BSA as the substrate showed that CP934P 0.35 % (w/v) dispersed in phosphate buffer pH 7.0 had no inhibitory effect on trypsin activity (Figure 4.1). ,_...... _g bJ) 40 2, :E0 rn "O..... 10 c.) Figure 4.1 Time-course of the formation of acid-soluble peptides from BSA after incubation with bovine trypsin (10 µg/ml) in the presence ( open circles) and in the } absence (closed circles) ofCP934P (0.35% w/v) at pH 7.0 (Section 4.3.1.3). Each point represents the mean ± s.d. of three experiments. However under the same conditions, CP934P inhibited the tryptic digestion of both BABE and LHRH (Figures 4.2 & 4.3). The rate of the formation of BA from BABE in the presence and absence of CP934P was constant over the period assayed. In the presence of CP934P the rate of BABE hydrolysis was 34% of the control (Figure 4.2). Using capillary electrophoresis the proteolysis ofLHRH was monitored by measuring the peak height during formation of a principal metabolite (mobility 12.5 ± 0.1 min, mean± s.d., n = 12) from LHRH (mobility 18.34 ± 0.34 min, n = 15, Figure 4.4). 136 0.8 ,-._ § N 0.6 l(') N "-' 25 0.4 § -B '_) ~ 0.2 ~ l i 0 0 10 20 30 40 50 Time (min) t~;_ •• Figure 4.2 Formation of BA from BABE (20 m.M) by trypsin (10 µg/ml) in the presence of CP934P (0.35% w/v) (closed circles) and in the absence of CP934P (open circles) at pH 7 .0 (Section 4.3.2). Each point represents the mean± s.d. of three experiments. '"" .I\ ·( { 250 $ 200 \ ( $ \ -I .:El 150 on ] 100 ~ ~ 50 I 'j- 0 ...====--..::'.._- 0 10 20 30 ,, Time (hours) '' :, Figure 4.3 Formation of the principal metabolite ofhLHRH following incubation of hLHRH (0.4 m.M) with trypsin (10 µg/ml) in the presence of CP934P (0.35% w/v) (closed circles) and in the absence of CP934P (open circles) at pH 7.0 (Section 4.3.2). Each point represents the mean± s.d. of three experiments. ·, 137 Figure 4.4 shows capillary electrophoretograrns of LHRH extracts after 24 h digestion with trypsin in the presence and absence of CP934P. The rate of metabolite formation was constant over 24 h in the presence and absence of CP934P (Figure 4.3). In the presence of CP934P the rate ofLHRH metabolite formation was 28% of the control. ' '; t', ~ I 1-f '< ,· ,~ '., I' -r ,\ _,. ,\ 138 0.030 l (A) 4 > ( 1 0.020 ) :;y 0.010 0.000 0 5 10 15 20 } '', ,--. . 0.030 l ~ (B) .._,--< 0 Ul 0.020 i::: ). 0 0.. Ul 0 I-< I-< 0.010 0 ) .\ t 0 ) ' t \ 0 0.000 '1 0 5 10 15 20 .0.030 .\ , I (C) > / 0.020 ../ ) J 0.010 ~,_ '> ) ,;, 0.000 'f' ~ ~ ( 0 5 10 15 20 Time (min) Figure 4.4 C_apillary electrophoretic separation of incubations ofhLHRH with trypsin (10 µg/ml): in the presence of CP934P (0.35% w/v) at pH 7.0 immediately after addition of trypsin (A), after 24 h incubation in the presence of CP934P (B) and after 24 h incubation in the absence of CP934P (C) at 37°C (Section 4.4.1.1 ). 139 4.4.1.2 Binding studies The spectrum of CP934P (0.005 % w/v) from 190 to 300 nm is shown in Figure 4.5. The absorption maximum of CP934P in 50 mM sodium phosphate buffer pH 7 .0 was 204 nm. The standard curve for absorbance at 204 nm versus concentration of CP934P was linear (r > 0.99) within the concentration range of 0.00125 to 0.01 % (w/v). ,. I 0.345 ) .r (I) § 0.245 "8 0 \ 0.045 -t·'----,----,------,,------r---~- 190 210 230 250 270 290 Wavelength (mn) Figure 4.5 Absorbance spectrum of CP934P 0.005% (w/v) in 50 mM sodium phosphate 1' ' I buffer pH 7.0 (Section 4.4.1.2) Dispersions of CP934P (0.35% w/v) in 50 mM sodium phosphate buffer pH 7 .0 (Section 4.3.2) were centrifuged at 800, 1 500 and 20 000 g for 30 or 60 min. Samples y ' ' ' ,'r (50 µ1) were taken from the top of the supernatant, diluted to 1 ml with 50 mM sodium phosphate buffer pH 7.0, transferred to quartz cuvette and the absorbances at 204 nm was determined using a Hewlett Packard 8452 A diode array spectrophotometer (Table 4.1). Only after CP934P dispersions were centrifuged at 20 000 g for 30 and 60 min were two phases observed. The CP934P concentration of the upper phase of the dispersions centrifuged at 20 000 g for 30 and 60 min were 0.052 and 0.047% (w/v), respectively, which equated to approximately 14% of the original CP934P concentration. 140 Table 4.1 Sedimentation of CP934P 0.35 % (w/v) in 50 mM sodium phosphate buffer pH 7. 0. Data are absorbances of samples taken from the upper region of the dispersion. Centrifugation Absorbance at 204 nm ,,, Control 1.05 ± 0.01 a 800 g for 30 min 0.923 ± 0.11 1 500 g for 30 min 0.92 ± 0.16 20 000 g for 30 min 0.168 ± 0.011 : 1 20 000 g for 60 min 0.156 ± 0.0046 i ) { a each value represents the mean ± s.d., n=3. ) ·~ '<:·, As we were unable to pellet the CP934P by centrifugation as described by LueBen et al. (1995) we investigated centrifugal filtration as a method for studying the binding of proteolytic substrates and proteins to CP934P (0.35% w/v) at pH 7.0. Samples (0.4 ml) of CP934P (0.35% w/v) at pH 7 .0 were transferred into Ultrafree® - MC centrifugal filtration units and centrifuged at 5 000 g for 2 hat room temperature (Section 4.3.3.1). j \ The filtrates were analysed for CP934P by diluting the filtrate (50 µl) in 50 mM sodium ''\ phosphate buffer to give 1 ml final volume and measuring the absorbance at 204 nm as J \ described above. Analysis showed that the CP934P was completely retained by the membrane. Consequently the bindingofLHRH (0.4 mM), BABE (20 mM) and trypsin ) (50 µg/ml) to CP934P 0.35% (w/v) at pH 7.0 was investigated using centrifugal filtration (Section 4.3.3.1), Non-specific binding to the membrane was not observed ; ' 'j .] during filtration of either BABE or LHRH, and less than 5% of the trypsin protein was bound to the membrane during filtration. There was no significant binding ofLHRH to CP934P and only 10 ± 2% of BABE was bound to CP934P at pH 7.0. Analysis of the trypsin filtrate for protein content showed that 32 ± 5% of the protein was filtered implying 68% binding to the CP934P (Table 4.2). Using BABE the synthetic substrate of trypsin (Section 4.3.1.3) only 11± 2% of the tryptic activity was filtered after incubation with CP934P and this had a specific activity of only 34% of the control activity. 141 Table 4.2 The protein concentration, activity and specific activity of trypsin solutions (50 µg/ml) filtered after incubation with pH 7 buffer alone (control) and after incubation with CP934P (mean± s.d., n = 3). Control CP934P % Filtered Protein concentration 50.0 ± 2.0 16.0 ± 1.9 32 (µg/ml) Trypsin activitl 0.118 ± 0.02 0.0129 ± 0.001 11 (µmol/min/ml) Trypsin specific 2.36 ± 0.26 0.81 ± 0.07 34b a activity (µmol/min/mg) a " Tryptic activity for BABE \ b Relative specific activity 4.4.1.3 Gel electrophoresis SDS-PAGE was used to identify trypsin and investigate lower molecular weight autolytic forms of trypsin in the centrifugal filtrates (Section 4.3.1.7). Figure 4.6 shows the SDS-P AGE analysis of the filtrates after reductive silver staining (Section 4.3 .1. 7). The filtrate of trypsin incubated in buffer alone gave a single band that migrated identically with unfiltered trypsin and the migration of the protein corresponded to a molecular weight of23 500. No protein bands were observed when the filtrates of the CP934P treated trypsin were analysed and stained with Coomassie Blue ( data not presented). However a faint band of23 500 molecular weight was observed in filtrates of the CP934P treated trypsin when the gel was stained with the more sensitive reductive silver stain (Section 4.3.1.7). No autolytic protein fragments were observed for any trypsin filtrates either in the presence or absence of CP934P. 142 !>- t r .. ~ •t 1 2 ., MW (X 10-3) -~.·.a~, · f ~~ ,;,- \ t>- "'- 45- '* 36- "" •• 29- " .... 24- 20- I A, J. Figure 4.6 SDS-PAGE of trypsin filtrates after silver staining. Trypsin filtrates after incubation in the presence of CP934P (0.35% w/v) (lane 1) in the absence of CP934P r ,, ,., (lane 2) in 50 mM sodium phosphate buffer pH 7.0. Migration and molecular weights of protein standards indicated on figure. ~ ;. L Po t>-- t,.. ' I>- t,.. 143 4.4.1.4 Gel filtration Trypsin had a retention time of 10.2 ± 0.5 min (mean± s.d., n=3) when analysed by gel filtration (Section 4.3.1.8). Trypsin (50 µg/ml) was autolysed by heating at 60°C for 45 min. Samples of the trypsin before and after heating were analysed by gel filtration and assayed for protein content by the modified Lowry assay (Section 4.3.1.1). Gel :filtration (Figure 4. 7) showed a dramatic reduction in the peak area and height for the autolysed ') trypsin with respect to the control. In addition a shoulder of lower molecular weight material eluting after 12 min was also observed for the autolysed trypsin. Protein > '.> determinations of control and autolysed trypsin showed no significant difference (p > 0.05), presumably because the modified Lowry assay (Section 4.3.1.1) measures the tyrosine content of peptides and proteins and is not influenced by autolysis of the protein. 180 ·,\ l A 120 ,,-., 60 '._/~ Q) \ Cl'.l ~ 0 0 0.. Cl'.l Q) I-< 0 5 10 15 20 25 I-< 180 -, 0 B ,'1 .:· t5 Q) t5 Q) 120 Q 60 0 0 5 10 15 20 25 Time (min) .'-.- '.:.. Figure 4.7 Gel filtration of trypsin (50 µg/ml) stored on ice (A) and after incubation at 60°C for 45 min (B) in 50 mM sodium phosphate buffer pH 7.0 (Section 4.3.1.8). 144 CP934P (0.35% w/v) and trypsin (50µg/ml) in 50 mM sodium phosphate buffer pH 7.0 were incubated alone and together at 37°C for 30 min and the centrifugal filtrates were analysed by gel filtration (Section 4.3.1.8). In the absence of CP934P, trypsin eluted as a single peak with a retention time 11. 7 ± 0.8 min (Figure 4.8 A). No peaks were observed for phosphate buffer pH 7 .0 only (chromatograph not shown) suggesting that ·) the peak observed for CP934P only (Figure 4.8 C) is a low molecular weight fraction of the CP934P. CP934P treated trypsin eluted after 10.6 ± 0.6 min (Figure 4.8 B) and the peak area was corrected by subtracting the peak area of the CP934P contaminant. The corrected peak area of the trypsin filtrate in the presence of CP934P was 32 ± 4% of the area of the control trypsin. There was no low molecular weight shoulder on the CP934P •:( treated trypsin filtrate peak. The protein contents of the control filtrate and CP934P treated trypsin filtrate determined by the modified Lowry assay (Section 4.3.1.1) were 48 ± 3 µg/ml and 14 ± 2 µg/ml, respectively, implying that 29% only of the protein was filtered in the presence of CP934P. \ l ·\ I '· ,-, > ... ;... 145 180 A 120 60 ·) 0 -4-----.----,---,-----,----, 5 10 15 20 25 ,--.., 0 f 180 g B (I.) ~ 0 120 ~ ~ i-.. ~ 60 ' 0* 0 -1------,---.---...... ----r----, 0 5 10 15 20 25 180 '., C 120 '60 /·- -~ 0 t::::=::': ' ;y 0 5 10 15 20 25 Time (min) ',. Figure 4.8 Gel filtration of centrifugal filtrates of trypsin (50 µg/ml) in the absence of "' "- CP934P (0.35% w/v) (A) and in the presence of CP934P (0.35% w/v) (B) and of CP934P (0.35% w/v) in the absence of trypsin (C) in sodium phosphate buffer pH 7.0 after incubation at 37°C for 30 min (Section 4.3.1.8). "- :,-, 146 4.4.1.5 HPLC of trypsin The centrifugal filtrates of trypsin in 50 mM sodium phosphate buffer pH 7.0 in the presence and absence of CP934P (Section 4.3.3.1) were analysed by reverse phase HPLC (Section 4.3 .1.9). The chromatograms of the filtrates of trypsin in phosphate buffer only show that trypsin after heating at 37°C for 30 min has a greater number of peaks (Figure 4.9 B) than trypsin that was maintained at 4°C (Figure 4.9 A). Filtrates . I from trypsin incubated at 37°C for 30 min in the presence of CP934P had a greater number of peaks than trypsin incubated in the absence of CP934P and the dominant peak at 20 min was not present in the filtrate (Figure 4.9 C) . .\ \ \ J, > "r >- 147 ;,''! :Ji A - ' . \ ~~ ~ 1 B 0 § ~ ~ ...t ...g 1l,} ,\ 0 ) ' \ _, ,\ C ! ', I' !lL-"----...... ---""'""""7i-- ~----- I I m 0 5 10 15 20 25 Time (min) --" > Figure 4.9 Reverse-phase HPLC of centrifugal filtrates of trypsin (50 µg/m1) maintained at 4°C (A), trypsin incubated at 37°C in the absence of CP934P (0.35% w/v) (B), and incubated in the presence of CP934P (0.35% w/v) at 37°C (C) in 50 mM sodium phosphate buffer pH 7.0 for 30 min (Section 4.4.1.5). 148 4.4.1.6 Langmuir isotherm The binding of trypsin to CP934P (0.35% w/v) at pH 7.0 over the concentration range 35-525 µg/ml was determined (Section 4.3.3.2). The data fitted the linearised form of the Langmuir isotherm with/= 0.997 (Figure 4.10). The binding capacity of trypsin for CP934P was 30 mg/g with an affinity constant of 0.028 ml/µg. I ., 18000 Q _§ 12000 e,-... ~ 6000 ~ 0 -1-----,.----.-----.----r----, 0 100 200 300 400 500 Unbound trypsin (ug/ml) .\ Figure 4.10 Langmuir absorption isotherm of trypsin onto CP934P (0.35% w/v) incubated in 50 mM sodium phosphate buffer pH 7.0 at 37°C for 30 min (Section 4.3.3.2). Each point represents the mean of three experiments. > t 4.4.1.7 Effect of ionic strength and pH on the adsorption of trypsin to ., CP934P ,)'- The effect of ionic strength and pH on the binding of trypsin to CP934P was investigated (Section 4.3.2.3). Figure 4.11 shows that the amount of unbound trypsin decreased significantly (p < 0.05) from 50 to 90 mM sodium chloride concentration, while from 90 to 150 mM sodium chloride concentration there was no significant change (p > 0.05) in binding of trypsin to CP934P ...... ',>.. ,' 149 Q 180 2,t I 1 .s 120 ...... ., I I 0 50 100 150 200 Sodium chloride (mM) Figure 4.11 Effect of sodium chloride concentration on binding of trypsin (300 µg/ml) to CP934P (0.35% w/v) at pH 7.0 (Section 4.4.1.6). Each point represents the mean± s.d. of three experiments. The effect of pH on trypsin binding to CP934P was tested similarly with CP934P dispersed in 50 mM NaCl at pH 6.0, 7.0 and 8.0 (Section 4.3.2.3). Figure 4.12 shows there was no significant difference (p > 0.05) in binding of trypsin to CP934P at pH 6.0, 7.0 and 8.0. ,.-.., 250 -s bl} 2, 200 ...... i:::: Figure 4.12 Effect of pH on binding of trypsin (300 µg/ml) to CP934P (0.35% w/v) at pH 6.0, 7.0 and 8.0 (Section 4.4.1.6). Each point represents the mean± s.d. of three experiments. 150 4.4.1.8 Protein binding The amount of BSA, SBTI, trypsin, chymotrypsin and pepsin (300 µg/ml) bound to CP934P (0.35% w/v, 50 mM NaCl at pH 7) was determined (Section 4.3.2.4). Table 4.3 shows the protein content of the filtrates after incubation in pH 7 buffer alone (control) and after incubation with CP934P. Under these conditions the rank order of protein binding to CP934P was BSA > SBTI >trypsin - chymotrypsin - pepsin. Table 4.3 Concentration of protein solutions (300 µg/ml) filtered after incubation with .,t pH 7 buffer alone (control) and after incubation with CP934P at 37°C for 30 min in phosphate buffer pH 7.0 (mean± s.d., n = 3). Protein Control CP934P % Filtered BSA 288 ± 26a 0 0 SBTI 283 ± 7.0 139 ± 11 49 "' Trypsin 298 ± 15 244 ± 18 82 Chymotrypsin 290 ±20 238 ± 12.3 82 Pepsin 290± 30 231 ±29 80 a Protein content (µg/ml) .'.\ :;, 4.4.2 Inhibition of a-chymotrypsin activity \, 4.4.2.1 Inhibition studies The hydrolysis ofhLHRH (0.4 mM) by bovine a-chymotrypsin (10 µg/ml) in the presence and absence of CP934P (0.35% w/v) was determined (Section 4.3.2). The ,r stopped reaction samples were analysed by reverse phase HPLC (Section 4.3 .1.6). Figure 4.13 shows that hLHRH was rapidly hydrolysed by chymotrypsin at pH 7. 0 at 37°C. One-way analysis of variance showed that in the presence of CP934P (0.35 % w/v) the hLHRH peak heights of the reaction samples did not significantly change over the 2 min period of incubation (p > 0.2). However regression analysis shows there is a ; significant relationship between peak heights and time (p < 0.05), indicating that hLHRH is being slowly degraded by chymotrypsin. HPLC chromatograms ofhLHRH and hLHRH after 1.5 min incubation with chymotrypsin (10 µg/ml) in the presence and absence of CP934P are shown in Figure 4.14. 151 0.08 5' 0.06 s ~ • • • i ~ ...... OJ) 0.04 (1) ~ ' ., 1 0.02 0-i 0 0 0.5 1 1.5 2 2.5 ( ' Time (min) Figure 4.13 Time-course of the hydrolysis ofhLHRH by chymotrypsin in the presence (closed circles) and absence (open circles) of CP934P (0.35% w/v) at pH 7.0 (Section 4.4.2.1). :~ \. ) ' ;. _., ·) ":, -'.. ;.\ :, } 152 0.06 (A) 0.04 0.02 ( ' 0 0 5 10 15 0.06 (B) ,-._ ( > ~ $ ~ 0.04 C/l o·~ 0.. C/l ',__ ~ 0.02 I-< I-< ....,0 () ~ t> 0 Q ) 0 5 10 15 \_ 0.06 (C) , .'> 0.04 0.02 'I I"'/ 0 0 5 10 15 Time (min) ' ' Figure 4.14 Reverse phase_ HPLC chromatograms ofhLHRH bovine chymotrypsin digests in the presence of CP934P (0.35% w/v) at pH 7 after 0.5 min (A) and 2.0 min (B) incubation at 37°C and after 2 min incubation in the absence of CP934P at 37°C . } (Section 4.4.2.1 ). 153 4.4.2.2 Binding studies The binding of chymotrypsin (50 µg/ml) to 0.35% (w/v) CP934P at pH 7.0 was ,, investigated using Ultrafree® - MC centrifugal filtration units (Section 4.3.2.1). No chymotrypsin protein was bound to the membrane during filtration. Analysis of the chymotrypsin filtrate for protein content (Section 4.3 .1.1) showed that 78% of the protein was filtered implying 22% binding to the CP934P (Table 4.4). Using BTEE, the synthetic substrate for chymotrypsin (Section 4.3.1.4) no significant chymotryptic activity was detected in the filtrate after incubation of chymotrypsin with CP934P (Table 4.4). ·( Table 4.4 Protein concentration, activity and specific activity of chymotrypsin (50 µg/ml) solutions filtered after incubation with pH 7 buffer alone ( control) and after incubation with CP934P (mean± s.d., n = 3). '·t, I ;, Control CP934P % Filtered Protein concentration 51.3 ± 4.9 40 ± 3.8 78 :~' (µg/ml) I ' b Chymotrypsin 1.61 ± 0.13 0.1 ± 0.03 0 . . a activity '., b Chymotrypsin 31 ± 2.6 2.6 ± 0.72 0 (• specific activityc (µmol/min/mg) a Chymotryptic activity for BTEE, µmol/min/ml of filtrate (', b ).. not significantly different from blank C µmol/rnin/mg of protein ' ~ 154 4.5 Discussion 4.5.1 Inhibitor studies Initial results showed that a CP934P dispersion at pH 7.0 did not inhibit trypsin activity when BSA was used as the substrate (Figure 4.1 ). The lack inhibition of proteolytic activity towards BSA by CP934P was in agreement with previous studies with salmon 1 intestinal lumen homogenates (Section 3.4.3.2). Using the low molecular weight ( ' synthetic substrate BABE, CP934P was shown to inhibit trypsin activity (LueBen et al., 1995). When the inhibition experiment was repeated with BABE or LHRH as the -/ substrate, reduced hydrolytic activity by trypsin in the presence of CP934P was -,, ~· observed (Section 4.4.1.1). The degree of proteolytic inhibition by CP934P was similar for both BABE and LHRH with hydrolysis rates 34 and 28% of the controls, respectively (Figures 4.2 & 4.3). Under the conditions of this study the rate of ~ metabolite formation from BABE and LHRH was constant both in the presence and absence of CP934P. The constant rate of metabolite formation in the presence of ·' CP934P suggests that enhanced enzyme autolysis does not play a role in the reduced proteolytic activity of trypsin. This is in contrast to the enhanced autolysis due to chelation postulated by Junginger and co-workers (LueBen et al., 1995). ,} CP934P could reduce the proteolytic activity by interacting with either the substrate or the enzyme or both. LueBen et al. (1995) previously investigated protein binding to -, '1 0.35% (w/v) CP934P at pH 6.7; however they measured the binding of 125I-BSA to r, CP934P and not trypsin. They reported that the binding of 125I-BSA to CP934P at pH ·,- 6.7 was not significant after separating the unbound 125I-BSA from CP934P by ' > centrifugation· at 800 g for 30 min. However, under similar conditions the CP934P was not completely pelleted, even at higher centrifugal force (Section 4.4.1.1 ). This suggests that the 125I-BSA measured after centrifugation at 800 g by LueBen et al. (1995) could still have been bound to CP934P. Centrifugal filtration was used to assess the binding of > trypsin, BABE and LHRH to CP934P (Section 4.4.1.2). The centrifugal filtration units (Ultrafree® - MC) completely retained CP934P (Section 4.4.1.2). No LHRH and only 10% of the BABE was bound to the CP934P suggesting that it is the enzyme that binds to the CP934P. 155 Protein analysis of the trypsin filtrate showed that 68% of the trypsin bound to the CP934P under the conditions of the study (Table 4.2). These results suggest that CP934P inhibits trypsin activity by binding the enzyme thereby decreasing the ·, concentration of free enzyme to hydrolyse substrate. However, CP934P treated trypsin filtrates had only 11 % of the control hydrolytic activity towards BABE, which corresponded to a relative specific activity of 34% (Table 4.2). This indicates that in the presence of CP934P trypsin does undergo inactivation. ,, In comparison to trypsin, CP934P had a greater inhibitory effect on chymotrypsin activity. Human LHRH was completely hydrolysed after 2 min at pH 7.0 at 37°C whereas in the presence of 0.35 % (w/v) CP934P there was no significant difference (p > 0.2) in the peak heights of hLHRH reaction samples after 2 min (Figure 4.13). The L significant regression analysis (p < 0.05) between peak height and time suggest that there is some residual chymotrypsin activity towards hLHRH. Binding studies showed that 22% of chymotrypsin (50 µg/ml) protein was bound to CP93P compared to 68% of trypsin under the same conditions (Tables 4.2 & 4.4). However it was expected that trypsin and chymotrypsin would bind similarly since they are both very similar in j primary and teritary structure (Stryer, 1988, Hedstrom et al., 1994). Analysis of the CP934P treated chymotrypsin :filtrates for BTEE hydrolytic activity showed that the unbound chymotrypsin was completely inactive (Table 4.4). The lower binding of ·, chymotrypsin to CP934P and higher proteolytic inhibition suggests that adsorbed chymotrypsin is more susceptible to denaturation. Previous studies by Junginger and co-workers used trypsin as a model protease to characterise the mechanism of inhibition by polyacrylic acids (LueBen et al. 1995). They suggested that CP934P inhibits trypsin by chelating essential calcium ions from its structure resulting in a conformational change in the protein opening it to autolysis. i> Autolysis is associated with the loss of activity of a proteolytic enzyme in solution incubated at optimum pH. In the case of trypsin, the trypsin molecules can cleave each other at lysyl and arginyl bonds. Three autolytic cleavage sites have been identified in 145 146 117 118 61 62 trypsin, Lys -Ser , Arg -Val , and Lys -Ser ( conventional chymotrypsin 156 numbering is used, Maroux and Desnuelle, 1969). It was shown that only cleavage at Lys61 -Ser62 resulted in lower tryptic activity (Maroux and Desnuelle, 1969). Maroux and Desnuelle (1969) also suggested that no fragment of the trypsin molecule will be released after cleavage at these sites because they will remain interconnected by 117 disulfide bonds. Varallyay et al. (1998) more recently showed that if either the Arg - Val118, or Lys61 -Ser62 sites are cleaved, the multiple cleavage sites within the Arg117- Val118, Lys61 -Ser62 segment will be subject to autolytic attack resulting in loss of activity. The 57 residue segment between the Arg117-Val118 and Lys61 -Ser62 (i.e. Ser62-Arg117) < cleavage sites contains 10 potential cleavage sites and two members of the catalytic triad. It is believed that the loss of activity is due to the multiple cleavage of this segment resulting in either the disruption of the trypsin structure or the displacement of ( 'l the members of the catalytic triad. The Ser62-Arg117segment is not interconnected by disulfide linkages and therefore multiple cleavages within this segment will result in the '- displacement of small fragments. Therefore it is expected that the displacement of these small autolytic fragments will not significantly change the molecular weight of the protein in relation to the native trypsin. SDS-PAGE is not expected to resolve autolysed from intact forms of trypsin. However, the lower staining intensity of the CP934P treated trypsin (Figure 4.6, Lane 1) supports the observation that trypsin is bound to CP934P (Section 4.4.1.2). The faint protein band of the CP934P treated trypsin had the same molecular weight as the trypsin only f filtrate (Figure 4.6, lane 2). No lower molecular weight peptides or protein were I 'I identified in the CP934P treated trypsin filtrate. The gel filtration system also had insufficient power to resolve the autolysed trypsin from the intact trypsin. However, gel filtration (Section 4.4.1.4) was used to monitor autoiysis by reiating peak area of the trypsin species to the amount of tyrosine reactive groups by the modified method of Lowry. This was demonstrated by gel filtration . ;. (Figure 4.7) and protein estimations of trypsin that had undergone accelerated autolysis by heating at 60°C for 45 min (Section 4.4.1.4). For the autolysed trypsin sample there was a decrease in the peak area with no change in the protein content, estimated by the Lowry assay. This suggests that the decrease in the trypsin peak area is due to less 157 absorbing material as a result of the displacement oflow molecular weight fragments which could not be identified by the gel filtration system. The chromatogram of the CP934P treated trypsin filtrates showed a decrease in the area of the trypsin peak compared with the untreated trypsin (Figure 4.8). The CP934P only filtrate contained a low molecular weight contaminant which migrated similarly to trypsin. The peak area of the CP934P treated trypsin was subtracted from the CP934P contaminant peak. The peak area of the CP934P treated trypsin equated to 28% of the control. The decrease in peak area of the CP934P treated trypsin corresponded to the lower protein content of the i ' CP934P treated trypsin filtrate (Table 4.2). The corresponding decrease in peak area and protein content (reactive tyrosines) indicates that binding of the enzyme plays an important role in the inhibition of trypsin by CP934P. This observation suggests that the binding of trypsin to CP934P is major mechanism of proteolytic inhibition. The CP934P treated trypsin eluted from the gel filtration column after 10.6 ± 0.6 min _, (mean± s.d., n=3) while trypsin alone eluted after 11. 7 ± 0.8 min. The slightly shorter ·~ elution time for the CP934P treated trypsin suggested that the treated trypsin might be of higher molecular mass compared to native trypsin. The apparent higher molecular mass of the treated trypsin could be due to a conformational change in the protease structure. It is well lmown that protein surface interaction can result in conformational change and loss of activity (Ghose and Bull, 1962, Tzannis et al., 1997, Norde and Zoungrana, 1998). The specific activity of the unbound CP934P treated trypsin being 34% of the untreated trypsin also suggests that surface inactivation is occurring (Table ' 1 'c 4.2). Reverse phase HPLC of trypsin filtrates suggested a number of different structural conformations of trypsin present in the different treatments (Figure 4.9). Trypsin that was incubated at 37°C for 30 min had a greater number of peaks than the trypsin maintained and filtered at 4°C. The greater number of peaks could be due to limited '- > autolysis of trypsin at the elevated temperature. A greater number of peaks were observed in the filtrate from the CP934P treated trypsin that was incubated at 37°C. The largest peak (20 min), presumably native trypsin, was not present in the filtrate from CP934P treated trypsin, again suggesting that trypsin was bound to the CP934P. The 158 greater number of peaks in the filtrate from CP934P treated trypsin suggests that there are a greater number of conformational forms of trypsin. A number of kinetic models have been proposed for the adsorption and desorption of proteins in solution (Andrade and Hlady, 1986). The Sodequist and Walton model (Soderquist and Walton, 1980) is the most commonly accepted kinetic model for protein adsorption to surfaces (Figure 4.15). The adsorption of proteins is described by three l stages: (1) an initial brief period where adsorption is reversible, (2) with time, adsorbed proteins undergo a slow conformational change and protein desorption is slow and (3) the more denatured adsorbed protein undergoes very slow desorption and is less likely ., to readhere. .~'- ., ~ ~ Ka Kd Kd' Ka' Kd" ~ • ·t Cs "" 2 3 Time Figure 4.15 Proposed kinetic mechanism for protein adsorption and denaturation (Soderquist and Watson, 1980). ~ For proteolytic enzymes, conformational changes resulting from surface interaction could make them substrates for the native enzyme and result in enhanced autolysis. Maste et al. (1995) investigated the inactivation of the proteolytic enzyme subtilisin at different solid-liquid interfaces. They proposed several models for proteolytic 159 inactivation by adsorption: (1) denatured protein on the surface is attacked by active enzyme on the surface, (2) denatured protein on the surface is attacked by active enzyme in solution (3) denatured protein at the surface is displaced by active enzyme. Using HPLC they measured the degree of subtilisin autolysis by determining the relative amount of unbound intact protein and acid soluble peptide fragments with time. Their results suggest that surface enhanced inactivation of subtilisin was due to the attack on adsorbed denatured subtilisin by subtilisin in solution (mechanism 2). ·\ ,\ The results of this work suggest that the lower proteolytic activity of proteases in the presence of polyacrylic acid is due to two processes. Firstly, direct adsorption of . -~- enzyme to the polymer thereby reducing the free concentration of active enzyme and secondly surface enhanced inactivation (Maste et al., 1995). The proposed kinetic .\ ·-1,, model of protease inhibition by polyacrylic acid is shown in Figure 4.16. l~\I; 4 ____ .,. - (s; ---• 4 ~¥-_, // i_ I " "--:/ / J, j. ' I I \ / ' \ 3 / 2 3 / { ' I I / ' / ' / I>, ' / ~ / ' -t ,, ~------..., (s; I ~ Time(min) ~ ·r > :t' Figure 4.16 Proposed routes for the inhibition of proteolytic activity in the presence of --... ;;..- polyacrylic acid. A straight arrow denotes a change in state, a dashed arrow denotes proteolytic attack (autolysis). 160 (1) the protease is adsorbed to the polymer surface and denatured, (2) the adsorbed denatured protein is desorbed back into solution, (3) the denatured adsorbed protein is attacked by active protease, (4) the desorbed denatured protein is attacked by active protease in solution (Figure 4.16). Gel filtration, SDS-PAGE, and protein analysis of the filtrates from CP934P treated trypsin indicate that the loss oftryptic activity can largely be attributed to adsorption to the polymer and not autolysis (mechanism 1). Analysis of the filtrates for proteolytic I \ activity and conformational stability (HPLC) indicated that the unbound trypsin has lower residual activity and present in different conformational forms. These ., observations suggest that mechanism 2 is also involved in the inhibition of proteolytic activity, however to a lesser extent. Inhibition studies showed that CP934P completely inhibited chymotrypsin activity towards hLHRH (Section 4.4.2.1) yet had lower binding to CP934P than trypsin (Section 4.4.2.2). The strong inhibitory activity of CP934P towards hLHRH was further supported by the unbound CP934P treated chymotrypsin having no activity towards \ BTEE (Table 4.4.). These observations suggest that mechanism 1 does not play a major ~ role in the inhibition of chymotrypsin by CP934P whereas the denaturation mechanisms 2, 3, and 4 play a greater role in inhibition. Further work is required determine whether this lower specific activity is solely due to conformational denaturation of the enzyme (mechanism 2), autolysis (mechanism 3 and 4) or a combination. • i ,. ,. The identification of autolysis products of trypsin in the presence of CP934P does not necessarily support the hypothesis of enhanced autolysis by bivalent ion chelation '., proposed by Junginger and co-workers (LueBen et al., 1996c). They based their ,'.( hypothesis on the observed decline in proteolytic hydrolysis rate with time (LueBen et al., 1996c). It would appear from their control experiments that the decline in hydrolysis . ;, rate over time was likely to be due to deviation from substrate saturation conditions. For example, LueBen et al. (1995 and 1996c) measured the trypsin hydrolytic activity towards BABE at 1.5 mM over a 250 min period. Under similar conditions we were only able to obtain pseudo-zero order kinetics (substrate saturation) for the hydrolysis of 161 BABE by trypsin at 20 mM BABE. (Figure 4.2). During their inhibitions studies Junginger and co-workers demonstrated that the proteolytic activity of the CP934P inhibited protease could be partially elevated with the addition of calcium (LueBen et al., 1995). They postulated that this increase in activity was due to the replacement of calcium which was removed from protease by the CP934P. However, this observation could be due to the calcium binding to the CP934P and displacing the adsorbed protease. Junginger's group tested their proposed chelation mechanism by incubating i I trypsin with the strong divalent ion chelator EDTA (7.5% w/v) at pH 6.7. EDTA however did not effect the hydrolytic activity of trypsin towards BABE (LueBen et al., ( 1996c). Junginger and co-workers suggested that the CP934P binding affinity for calcium is much higher than EDTA at pH 6.7. However they did not report any experimental evidence to support this statement. Previous work by Russin et al. (1974) ', showed that EDTA deprives trypsin of calcium only in the presence of 8 M urea, .l indicating that calcium is only accessible to chelation by EDTA after relaxation of the protein structure. 4.5.2 Binding studies Typically, Michaelis-Menton kinetic parameters (Km, kcat, KJ are useful for determining , J the type of inhibition (non-competitive and competitive inhibition) and relative potency of inhibitors. These parameters however cannot be determined as the interaction / between trypsin and CP934P is unlikely to be specific to a small number ofloci on the .Y 1 1 trypsin protein and the interaction will result in significant disruption of the three '< dimensional structure of trypsin. For this type of inhibition the binding capacity and affinity of trypsin for CP934P are more useful parameters for determining the potency of the CP934P to inhibit trypsin. The binding of trypsin to CP934P (0.35% w/v) at pH '- '.'( 7 .0 (Section 4.4.1.6) showed a good fit to the Langmuir model (r = 0.997), with a binding capacity of trypsin for CP934P of30 mg/g and an affinity constant of 0.028 ml/µg (Figure 4.10). (., ;~ The non-specific adsorption of proteins to interfaces have generally been attributed to electrostatic (Barroug et al., 1989), hydrophobic interactions (Elwing et al., 1987), and conformational stability of the protein (Norde, 1986). Given that the acrylic acid 162 polymer has a hydrophobic backbone and negatively charged carboxylic acid groups, the protein-polymer interactions can be expected to be hydrophobic, ionic or hydrogen bonding or a combination of these. The trypsin binding experiments showed that > increasing the sodium chloride concentration from 50 mM to 90 mM significantly increased (p < 0.05) the amount of trypsin bound to CP934P whereas from 90 mM to150 mM sodium chloride concentration there was no significant change (p > 0.05) in \ protein binding (Figure 4.11). The enhanced adsorption of trypsin to CP934P from 50 ( \ mM to 90 mM sodium chloride could be attributed to the reduced charge repulsions within and between trypsin molecules. Norde (1986) postulated that the existence of ( ions provides a less desirable environment for hydrophobic regions of the protein. Under these conditions, the proteins hydrophobic region is buried further into the interior of the protein to be away from the medium. In the presence ions the hydrophilic '\ regions of the protein becomes less repulsive to each other (if oppositely charged) due • to shielding effects of the medium. The reduced charge and the more compact trypsin structure allows the trypsin molecules to pack more closely on the polymer. Similarly, Tsai et al. (1996) postulated that the enhanced binding of the peptide calcitonin to ,)' poly(D,L-lactide-co-glycolide) with increasing sodium chloride concentration was due 1" to the peptide being more compact with lower repulsive forces. Similar binding of ~ i trypsin to CP934P at 90 mM and 150 mM sodium chloride suggests that the binding / I capacity of CP934P for trypsin has been reached. However in these studies increasing I'' the sodium chloride concentration will also reduce the size of the CP934P particles I-~ I. ( 1 thereby reducing the surface area for protein binding (BFGoodrich, 1994b). The reduced CP934P particle size may also explain why there was no significant change in trypsin binding from 90 to 150 mM sodium chloride. '·.,, The effect of pH on protein adsorption has been studied and generally the maximum adsorption occurs at or ne~ the pl of the protein (Norde, 1986, Luey et al., 1991). At its pl the protein will have less repulsive forces, resulting in compact packing of the protein l > and less repulsive forces between adsorbed proteins; again allowing for the adsorbing proteins to pack more tightly (Norde, 1986, Luey et al., 1991). As trypsin has a pl of approximately 10.5 (Cunningham, 1954) and CP934P has an apparent pKa value of 6. 7 (Charman et al., 1997) it can therefore be postulated that as the pH increases from pH 6 163 to 8 the amount of trypsin bound to CP934P will increase. However between pH 6.0 and 8.0 there was no significant change (p > 0.05) in the amount of trypsin bound to CP934P. This suggests that at pH 8 the trypsin molecules are in the loose form to overcome intramolecular charge repulsions, rather than a compact form which promotes surface adsorption. BSA, trypsin, chymotrypsin, pepsin and SBTI (300 µg/ml) showed variable binding to CP934P 0.35 % (w/v) at pH 7.0 (Section 4.4.1.8). The rank order of protein binding to CP934P was BSA > SBTI > trypsin, chymotrypsin and pepsin. The pl values of the individual proteins did not correlate with their affinity for CP934P (Table 4.5). However work by Tsai et al. (1996) indicates that significant changes in protein adsorption only occurs within a range of 1-2 pH units of the protein pl. ,,< Table 4.5 CP934P binding, molecular weight and pl of proteins and peptides interacting \ with CP934P. Protein % filtereda MW pl Reference BSA 0 66 700 4.7 Giacomelli et al. (1997) SBTI 49 8 000 4.3 Birk (1985) Trypsin 82 23 800 10.5 Cunningham (1954) ' ' a-Chymotrypsin 82 23 500 9.1 Desuelle (1960) Pepsin 80 35 000 1.0 Borey and Yanari (1960) 1 a calculate from data presented in Table 4.3 The high binding of BSA to CP934P can be attributed to its low conformational stability (soft protein) (Giacomelli et al., 1997). Norde (1986) observed that the proteins with a relatively lower degree of conformational stability bind to variety of surfaces (hydrophobic or hydrophilic) even under adverse electrostatic conditions. In contrast, high stability proteins (hard protein) adsorb mainly on hydrophobic surfaces, whereas on hydrophilic surfaces only under conditions favorable to electrostatic interaction (Norde, 1986). In order to determine whether the conformational stability of the proteins is a driving force for protein binding to CP934P the relative conformation stability of the remaining proteins have to be determined. Young et al. (1988) detennined the 164 relative conformational stability of proteins by their tendency to precipitate in varying amounts of ethanol in an aqueous solution. The higher binding of BSA to CP934P eludes to why CP934P did not inhibit trypsin activity when BSA was used as the substrate (Figure 4.1). Under these conditions the greater amount of BSA bound to CP934P will prevent trypsin from interacting with the polymer thereby remaining active in solution to hydrolyse the excess BSA in solution. l·t " ., ~ !' 165 Chapter 5: General Disscussion and Future Perspectives > 1 \ ,, ,/ ( i ," I' 166 5.1 General discussion Oral delivery of peptide and protein drugs is the most convenient route and least intrusive for delivery to fish. Although the delivery of proteins and peptides across the intestine is difficult with respect to low mucosa! permeability and metabolic barriers (Section 1.2), it remains very attractive for highly potent peptide hormones. A number of investigators have demonstrated that the intestine of teleosts does absorb small amounts of peptides and proteins (Section 1.4.3). Despite the attempts to improve the oral delivery of peptides and proteins in teleosts (Section 1.4.4) there has been no systematic studies of the intestinal metabolic barriers to the oral delivery of peptides and proteins in teleosts. " Previous studies characterising the intestinal proteases of teleosts show that they differ from mammalian proteases in their activity, stability, requirement for divalent ions and susceptibility inhibitors (Section 1.4.2). This suggests that the protease inhibitors used to stabilise peptides in the intestine of mammalian species may have different activities in .\ teleosts. The aim of this dissertation was to study the metabolic barriers of the lower intestinal tract of salmon and investigate strategies to overcome this barrier to improve the oral bioavailability of peptide and protein drugs. The highly potent LHRH analogues have been used in the aquaculture industry to synchronise reproduction in fish (Donaldson and Hunter, 1983). To study the metabolic stability of human and salmon LHRH analogues in the lower intestinal tract of salmon a 'I CE assay was developed (Section 2.0). Bovine a-chymotrypsin digests of human and ,, salmon LHRH analogues were used to study separation and reproducibility of the CE assay. Sample stacking and run buffer pH were important variables in controlling separation and reproducibility of the assay (Sections 2.4.2 & 2.4.3). The optimum conditions for reproducibility and separation ofLHRH analogues from their metabolites were as follows: dilution of the LHRH digest 4-fold in deionised water, electrophoresis in 10 mM sodium acetate buffer at pH 4.0 containing 75 mM sodium chloride at constant current of 60 µA. Under these conditions hLHRH and sLHRH eluted after 12.5 and _18.5 min, respectively (Section 2.4.3). The CE assay offered a simple cheap assay that could be used to analyse the metabolic stability of both LHRH analogues. However the Dionex 167 instrument was unreliable with many instrument problems. Subsequently HPLC was also used to monitor the metabolic stability ofhLHRH. The lower intestinal tract of the salmon is divided into three anatomically distinct sections: the anterior, middle and posterior. The lumenal and mucosal homogenates were assayed for proteolytic activity towards BSA and LHRH analogues. BSA and the LHRH analogues were most stable in the posterior section (Sections 3.4.1.4 & 3.4.1.6). Both LHRH analogues had similar metabolic stabilities in each intestinal section, suggesting that both analogues are metabolised by the same protease (Section 3.4.1.7). The higher !1- metabolic stability of BSA and LHRH analogues in the posterior intestine correlated with the lower hydrolytic rates towards the synthetic substrates of chymotrypsin, trypsin and elastase. CE oflumenal LHRH digests showed that the LHRH analogues had identical hydrolytic patterns in each intestinal section (Section 3.4.1.6). This suggests that the \ LHRH analogues are metabolised by the same protease(s) along the entire lower intestinal tract of salmon. " The pancreatic endopeptidases, chymotrypsin, trypsin and elastase were expected to be the major enzymes involved in the metabolism ofLHRH analogues in the intestinal lumen. Incubations ofLHRH analogues with purified commercial preparations of , r a-chymotrypsin, trypsin and elastase showed that the LHRH analogues were extensively metabolised by bovine a-chymotrypsin, whereas the LHRH analogues were slowly hydrolysed by porcine elastase and bovine trypsin (Section 3.4.1.5). This suggests that ,, the chymotryptic activity of the salmon intestinal lumen is the major lumenal metabolic barrier to the oral delivery LHRH analogues to salmon. Protease inhibitors are commonly used to increase the metabolic stability of peptides and :, proteins in the intestinal lumen to increase their opportunity for absorption across the intestinal epithelium. Different classes of protease inhibitors, including EDTA, C934P, DOC and SBTI, were tested for their ability to stablise the LHRH analogues in the intestinal lumen in salmon (Section 3.4.3.4). SBTI was shown to be most the effective 168 inhibitor in stabilising the LHRH analogues against the lumenal proteolytic activity of the salmon intestine. The high inhibitory activity of SBTI (Seidl and Liener, 1972), a well known inhibitor of chymotrypsin, supports the hypothesis that chymotrypsin is the major protease involved in the metabolism of LHRH analogues in the intestinal lumen. To understand how the mucosal membrane peptidases may influence bioavailability of peptide and proteins, the stability of BSA and LHRH analogues were compared in mucosal homogenates prepared from the middle and posterior intestinal sections. ( '.• Proteolytic activity of the posterior mucosal homogenate was 4-fold higher than that ':/ observed for the middle mucosal homogenate (Section 3.4.2.3). Mucosal enzymes of the posterior section hydrolysed the LHRH analogues, whereas for the middle mucosal ,, homogenate had no hydrolytic activity towards the LHRH analogues (Section 3.4.2.2). The LHRH analogues had similar metabolic stabilities in the posterior mucosal \ homogenates, suggesting the same protease(s) are involved in their metabolism. The non-specific proteolytic inhibitors DOC, EDTA and the aminopeptidase specific inhibitor bestatin significantly inhibited (p < 0.005) the mucosal hydrolytic activity towards the LHRH analogues (Section 3.4.3.5). Despite the LHRH analogues being N '( and C- terminally blocked, the inhibitory activity ofbestatin, suggests that ( \ :-r aminopeptidases are involved in the mucosal metabolism of the LHRH analogues ' ~ (Section 3.4.3.5). However this anomaly was not investigated in this work. ii The higher proteolytic activity of the posterior mucosal homogenate towards BSA and the LHRH analogues correlated with higher activity towards the chymotryptic substrate ·' N-benzoyl-L-Tyr ethyl ester. The lack of inhibition of mucosal proteolytic activity '; towards the LHRH analogues by the SBTI indicates that this activity is not due to pancreatic chymotrypsin activity. Previous studies suggests that this activity is due to the membrane bound protease endopeptidase 24.18 (Sterchi et al., 1982) which has also been shown to hydrolyse hLHRH (Stephenson and Kenny, 1988). 169 In comparison to the middle intestinal LHRH proteolytic activity, the posterior intestinal section has lower lumenal activity but higher mucosa! activity. In order to determine which section is most metabolically favorable for the oral delivery of the LHRH analogues, the combined activities of the lumen and mucosa in each section must be considered. However the in vitro kinetics of proteolysis is likely to be quite different to in vivo. The activity towards the substrate will depend on its concentration. In this study the in vitro proteolysis ofLHRH was determined at 0.2 mM, a concentration is considerably higher than that expected in vivo after oral delivery of peptide. To estimate the catalytic " activity towards a substrate at different substrate concentrations, the K.i1 (Michaelis Menton constant) and kcat (catalytic turnover) values of the proteolytic enzyme must be determined. However these values could not be determined for crude homogenates as they can only be determined by using purified enzyme preparations in vitro. Obviously, the concentrations of the lumenal proteases in vivo are also going to determine proteolytic ',\ activity towards the LHRH analogues. The in vitro lumenal activity towards the LHRH was determined at a protein concentration of 5 µg/ml (Section 3 .4.1.6), whereas the in vivo concentrations of proteins in the lumen are expected to be in the range of 5-10 mg/ml I " (Section 3.4.1.3). In addition, homogenisation of intestinal extracts may alter the activity of the enzymes, particularly intestinal membrane-bound proteases, which have been dislodged from their membrane. The true metabolic barrier of the intestinal tract can only be determined in vivo, however Schep et al. (1997) has shown that there are also regional ~ differences in permeability of the salmonid lower intestine to the hydrophilic marker mannitol. To account for these regional differences in intestinal permeability the adsorption of the LHRH analogues could be related to the permeability of the intestine to mannitol. Mannitol is believed to be absorbed across membranes by the same route as peptides (Nellans, 1991, Ma et al., 1993, Sakai et al., 1997) and is not influenced by the metabolic barrier. Dispersions of CP934P (0.35 % w/v, pH 8.0 at 15°C) were not as effective at inhibiting proteolytic activity of the salmon intestinal lumen compared with purified pancreatic enzymes as described by LueBen et al. (1995) (Section 3.4.3.4). Furthermore, the rates of 170 LHRH hydrolysis remained constant in the presence of CP934P, indicating no autolysis in contrast to that postulated by Lue13en et al. (1995). Commercial preparations of bovine trypsin and a-chymotrypsin were used to investigate the effect of proteolytic inhibition by enhanced autolysis in the presence of CP934P. Dispersions of CP934P (0.35 % w/v, pH 7.0 at 37°C) inhibited the proteolytic activity of bovine trypsin and chymotrypsin. In '( the presence ofCP934P, hydrolysis ofBAEE and hLHRH by trypsin was reduced to 34% and 28% of control activity, respectively (Section 4.4.1.1 ). The rates of substrate hydrolysis followed pseudo-zero kinetics, again indicating that inhibitions were not due ,, to elevated autolysis of the enzyme. Under the same conditions, hLHRH was stablised Y, :, against the activity of bovine a-chymotrypsin (Section 4.4.2.1 ). Binding studies using centrifugal filtration, showed binding of trypsin ( 68% bound), chymotrypsin (22%) at 50 µg/ml and BABE (10%) at 20 mM to CP934P (0.35% w/v, pH ,' 7.0), whereas no LHRH (0.4 mM) was bound (Tables 4.2 & 4.4). This suggests that CP934P reduces proteolytic activity by binding the proteases, thereby reducing the free concentration of protease available to hydrolyse the proteolytic substrates. However the lower relative specific activities of the filtrates suggests that surface inactivation of the proteases is also occurring (Tables 4.2 & 4.4). Chymotrypsin appeared to be more susceptible than trypsin to inactivation; the unbound CP934P treated chymotrypsin had no hydrolytic activity, whereas the unbound trypsin had a relative specific activity of 34% '( (Tables 4.2 & 4.4). Analysis of the unbound trypsin by gel filtration indicated that inhibition was due to trypsin binding to the polymer rather than elevated autolysis (Section 4.4.1.4). SDS-PAGE (Section 4.4.1.3) and reverse phase HPLC (Section 4.4.1.5) further supported lack of autolysis so suggesting in inhibition as being due to the binding of trypsin to CP934P. Reverse phase chromatography of the filtrates also showed that trypsin exists in a greater number of conformational forms in the presence of CP934P .~ (Section .4.4.1.5). This suggests that the lower relative specific activity of the unbound trypsin is due to surface inactivation, resulting in irreversible confo1mational changes of the trypsin protein. 171 A kinetic model was proposed for the inhibition of proteolytic activity by polyacrylic acids (Section 4.5). This model consists of four mechanisms: (1) the protease is adsorbed to the polymer surface and denatured (2) the adsorbed denatured protein is desorbed back into solution (3) the denatured desorbed protein is attacked by active protease and ( 4) the adsorbed denatured protein is attacked by active proteases in solution. '\ Since Michaelis-Menton kinetic parameters cannot not be used for this type of inhibition, the binding capacity and affinity of trypsin for CP934P were determined. The trypsin ( ,, adsorption data obtained fitted well to a Langmuir isotherm (Section 4.4.1.6). The " :r equilibrium binding capacity of trypsin to CP934P (0.35% w/v) at pH 7.0 was 30 mg/g with an affinity constant of 0.028 ml/µg (Figure 4.10). The effect of sodium chloride concentration and pH on the binding of trypsin to CP934P was investigated to shed some understanding of the mechanism of binding and to determine what effect these variables might have on CP934P inhibition of trypsin activity ~ in the salmon intestine (Section 4.4.1. 7). Preliminary studies showed that from 50 mM to ,_ ' 90 mM sodium chloride concentration at pH 7.0 the amount of trypsin bound to CP934P significantly increased (p < 0.05) whereas from 90 mM to 150 mM sodium chloride there ( was no significant change in trypsin binding (p > 0.05). The higher binding of trypsin to CP934P from 50 mM to 90 mM sodium chloride concentration may be due to the trypsin molecule being more compact with lower repulsive forces, allowing the adsorbed proteins to pack more closely on the polymer (Norde 1986, Tsai et al., 1996). The similar binding ~ of trypsin to CP934P at 90 mM and 150 mM suggests that maximum binding capacity of CP934P for trypsin has been reached. However increasing the sodium chloride concentration may also influence binding by reducing the size of the CP934P particles '.\ (BFGoodrich, 1994b). To overcome the effect of particle size, future binding studies should use the non-crosslinked water soluble Carbopol907 (BFGoodrich, 1994a). 172 It has been postulated that near the pl of the protein the repulsive forces within and between the protein are minimised, thereby reducing the interspace to overcome the repulsive force on adsorption (Norde, 1986). Since trypsin has a pl of approximately 10.5 (Cunningham, 1954) it was therefore expected that trypsin binding to CP934P would increase from pH 6 to 8. However binding studies showed that there was no significant I change (p > 0.2) in trypsin binding to CP934P over the pH range 6 to 8. This suggests that there is little change in the repulsive forces within and between trypsin molecules between pH 6 and 8. This work suggests that the pH range tested is not sufficiently close < ,~ to the pl of trypsin to enhance adsorption of trypsin to CP934P. Calcitonin which has a pl ( around 9.4, was shown to have maximum adsorption to poly(D,L-lactide-co-glycolide in the 7.4 to 10 pH range (Tsai et al., 1996). Trypsin binding to CP934P should be test in the 8-10 pH range to determine whether maximum adsorption of trypsin occurs near its i -\ pl. It is expected that the size of the CP934P particles would not change over the 6-10 pH range and therefore will not effect trypsin binding to CP934P (BFGoodrich, 1994b). :,, BSA, SBTI, trypsin, chymotrypsin and pepsin (300 µg/ml) showed variable binding to CP934P (0.35 % w/v, 50 rnM NaCl, at pH 7). CP934P had the highest binding for BSA, followed by soybean trypsin inhibitor and then trypsin, chymotrypsin and pepsin, with ( similar binding. The binding of the proteins did not correlate with the protein pl or 'i molecular weight (Table 4.5). The reason for the poor relationship between pl and protein adsorption at pH 7 could be due to the pis of the proteins tested being more than 2 pH units away from pH 7. Previous work by Tsai et al. (1996) indicates that for significant "' changes in protein adsorption near the protein pl, the pH of the medium has to be within 1-2 pH units of the protein pl. The high binding for BSA, may be due to its low !, conformational stability (BSA), since proteins with low conformational stability have \ been shown to have higher surface absorption properties (Norde, 1986). ;, 173 5.2 Future perspectives The next stage of this work should assess the use of proteolytic inhibitors in improving the oral absorption ofLHRH analogues in vivo. This could be determined by the in vivo model developed by Schep (1999); in this model the salmon posterior intestine would be - '! perfused enabling direct delivery ofLHRH and inhibitors to the posterior intestine. f ) I A number of parameters have to be considered in order to relate the in vitro inhibitory performance of CP934P dispersions to their in vivo performance. The components of the ~ ,, intestinal fluid, which could affect protease-polymer interaction, include: bile salts, -( ' y proteins, sodium, potassium and calcium ions and their associated ions. The variable affinities of different proteins for CP934P suggest that lumenal proteases will have to compete with other lumenal proteins for binding to CP934P (Section 4.4.1.8). This effect was observed when CP934P was unable to inhibit trypsin activity in the presence of BSA (Figure 4.1). BSA was added to the incubate as the proteolytic substrate, however the higher CP934P binding of BSA compared to trypsin (Table 4.3) suggests that BSA " displaced the CP934P bound trypsin to hydrolyse the excess unbound BSA. Given this it is expected that the composition and concentration of lumenal proteins will have a major .,, influence on the ability of CP934P to protect peptides from metabolic attack in vivo. The ; in vitro CP934P inhibitory assays described could be used to suggest what effects these ( components would have on the in vivo inhibitory activity of CP934P. Individual t t components could be assayed at concentrations similar to those of the intestinal lumen and the inhibitory activity of CP934P towards various purified lumenal proteases would ' "' be tested. The inhibitory activities would be tested against the major lumenal proteases including, trypsin, chymotrypsin and elastase. ' In addition to using protease inhibitors to improve the oral absorption of peptides, a number of other strategies need to be investigated (Section 1.3). Intestinal regions with low proteolytic activity maybe targeted for peptide drug delivery (Section 3 .4.1 ). This work indicates that the posterior lumen is most favorable for LHRH delivery however the higher mucosal activity of the posterior section must also be considered. Since most 174 delivery systems use lumenal pH, gut enzymes or gut transit times to target specific regions of the intestine (Section 1.3.4) these parameters need to be investigated in salmon in order to delivery peptide drugs to specific regions of the intestine. Recent experiments by Schep (1998) demonstrate that the permeation of the salmonid posterior intestine to ;, the hydrophilic marker mannitol was enhanced in the presence of DOC, suggesting that inclusion of DOC in oral formulations could also improve the oral absorption of peptides in salmon. Future investigations should also study analogues that are more resistant to metabolic degradation. McLean et al. (1991) showed that the oral delivery of the LHRH 10 6 analogue des-Gly [Ala ] LHRH ethylamide to salmon resulted in a 100-fold increase in peak plasma concentrations of the peptide compared to native LHRH. In order to achieve higher levels of orally absorbed peptide in salmon it is likely that a combination of all these strategies will be required. In conclusion, this thesis has demonstrated in vitro that there are regional differences in proteolytic activities of the lower intestinal tract of salmon towards BSA and LHRH analogues. The in vitro proteolytic stabilities of BSA and LHRH analogues can be "' improve by protease inhibitors. Therefore the addition of protease inhibitors in oral formulations could be useful in improving the oral bioavailabilities of peptides and proteins in salmon. In vitro metabolic studies show that polyacrylic acids inhibit proteases by protease-polymer interaction. A better understanding of the underlying i mechanisms of proteolytic inhibition by acrylic acid polymers could lead to improved formulations enabling greater oral bioavailabilities of peptide drugs. ;;; 175 References Abrahamson, D.R. and Rodewald, R. (1981). Evidence for the sorting of endocytic vesicle contents during the receptor-mediated transport IgG across the newborn rat intestine. Journal of Cell Biology 91, 270-280. ·t Altria, K.D. (1993). Essential peak area normalization for quantitative \( impurity content determination by capillary electrophoresis. Chromatographia 35, I ' 177-182. _}.. : '/ Altria, K.D., Frake, P., Gill, I., Hadgett, T., Kelly, M.A. and Rudd, D.R. (1995). Validated capillary electrophoresis method for the assay of a range of basic drugs. Journal ofPharmaceutical and Biomedical Analysis 13, 951-957. Andrade, J.D. and Hlady, V. (1986). Protein adsorption and materials biocompatibility: A tutorial review and suggested hypothese. In: Advances in Polymer Science. Eds: P.J . . t Kuhn and B.B. Byers, Berlin, Springer-Verlag. pp 1-58. Anik, S.T. and Hwang, l-Y. (1983). Adsorption of [D-Nal(2)6)]LHRH, a decapeptide, t onto glass and other surfaces. International Journal ofPharmaceutics 16, 795-801. t Aoyagi, T., Tobe, H., Kojima, F., Hamada, M., Takeuchi, T. and Umezawa, H. (1978) . . " Amastatin, an inhibitor of aminopeptidase A, produced by actinomycetes Journal of Antibiotics 31, 636-638. Aungst, B.J. (1993). Novel formulation strategies for improving oral bioavailability of drugs with poor membrane permeation or presystemic metabolism. Journal of Pharmaceutical Sciences 82, 979-87. 176 ,. Aronson, N.N. and Barrett, A.G. (1978). The specificity of cathepsin B. Hydrolysis of glucagon at the C-terminus by a peptidyldipeptidase mechanism. Biochemistry Journal 1971, 759-761. Artursson, P. and Karlsson, J. (1991). Correlation between oral drug absorption in ( humans and apparent drug permeability coefficients in human intestinal epithelial (Caco-2) cells. Journal ofBiochemistry Biophysics Research Communications 175, \. I ~ 880-885. }..- '( Artursson, P., Lind.mark, T., Davis, S.S. and Illum; I. (1994). Effect of chitosan on the permeabilty of monolayers of intestinal epithelial cells (Caco-2). Pharmaceutical ·" Research 11, 1358-1361. ·( i Artursson, P. (1990). Epithelial transport of drugs in cell culture. I: A model for studying the passive diffusion of drugs over intestinal absorptive (Caco-2) cells. Journal of Pharmaceutical Sciences 79, 476-482 . . t Asgeirsson, B. and Bjamason, J.B. (1991). Structural and kinetic properties of chymotrypsin from Atlantic cod ( Gadus morhua ). Comparison with bovine chymotrypsin. Comparative Biochemistry and Physiology B 99, 327-335. ( ,. Asgeirsson, B., Fox, J.W. and Bjamason, J.B. (1989). Purification and characterization of trypsin from the poikilotherm Gadus morhua. European Journal ofBiochemistry 180, '\ 85-94. > ~ A :,-. Ash, R. (1980). Hydrolytic capacity of the trout (Salmo gairdneri) intestinal mucosa with respect to three specific dipeptides. Comparative Biochemistry and Physiology B 65, 173-176. 177 Ash, R. (1985). Protein Digestion and Absorption. In: Nutrition and Feeding in Fish. Eds: C.B. Cowey, A.M. Mackie, and G.R. Bell, Academic Press: Orlando. Ashford, M. and Fell, J.T. (1994). Targeting drugs to the colon: delivery systems for oral administration. Journal ofDrug Targeting 2, 241-257. '1 Ashford, M., Fell, J.T. and Woodhead, P.J. (1993). An in vitro investigation into the -; suitability of pH-dependent polymers for colonic targeting. International Journal of Pharmaceutics 91, 241-245. \ Atisook, K. and Madara, J.L. (1991 ). An oligopeptide permeates intestinal tight junctions _,, at glucose-elicited dilations. Implications for oligopeptide adsorption. Gastroenterology 100, 719-724. l Atlas, D., Levit, S., Schetchter, I. and Berger, A. (1970). On the active site of elastase, partial mapping by means of specific peptide substrates. FEBS Letters 11, 281-283. j_ Austin, B. (1984). The future of bacterial fish vaccines. Vaccine 2, 249-254. Bai, J.P.F. (1994a). Effects of bile salts on brush-border and cytosolic proteolytic / activities of intestinal enterocytes. International Journal ofPharmaceutics 111, 147-152. ,. Bai, J.P.F. (1994b). Subcellular distribution of proteolytic activities degrading bioactive .\ peptides and analogues in mucosal cells of the small intestine and colon. Journal of Pharmacy and Pharmacology 46, 671-675. .. , Bai, J.P.F. (1995). Colonic delivery of peptide and protein drugs: consideration of :, intracellular proteolytic enzymes. s.t.p; Pharmaceutical Sciences 5, 30. 178 ~ Bai, J.P.F. and Amidon, G.L. (1992). Structure specificity ofmucosal-cell transport and metabolism of amino acids peptide drugs: implications for oral peptide delivery. Pharmaceutical Research 9, 969-978. Bai, J.P.F., Chang, L.-L. and Guo, J.-H. (1995a). Targeting of peptide and protein drugs to specific sites in the oral route. Critical Reviews in Therapeutic Drug Carrier Systems i' 'i 12, 339-371. { -f· I ., Bai, J.P.F., Chang, L.L. and Guo, J.H. (1995b). Effects of polyacrylic polymers on the ' .., lurnenal proteolysis of peptide drugs in the colon. Journal ofPharmaceutical Sciences 84, 1291-1294. \ Baluom, M., Friedman, D.I. and Rubinstein, A. (1997). Absorption enhancement of calcitonin in the rat intestine by carbopol-containing submicron emulsions. International I -' Journal ofPharmaceutics 154, 235-243. > J Banga, A.K. (1995). Therapeutic peptides and proteins: formulation, processing, and "' delivery systems, Technomic Publishing, Lancaster. I ,} -, Banks, W.A. and Kastin, A.J. (1985). Peptides and the blood-brain barrier: lipophilicity as a predictor of permeability. Brain Research Bulletin 15, 287-292. Barrett, A.J. and Heath, M.F.(1977). Lysosomal enzymes. In: Lysosomes: A Laboratory ·'t Handbook. Ed: J.T. Dingle, Elsevier, Amsterdam. p. 19. I ., j-. Barroug, A., Lemaitre, J. and Rouxhet, P.G. (1989). Lysozyme and apatites- A model of protein adsorption controlled by electrostatic interactions. Colloids and Surfaces 37, 339- ,., 355. 179 Barth, R. and Afting, E.G. (1984). Cathepsin D from pig myometrium. Biochemistry Journal 219, 899-904. Baumann, W., Bizzozero, S. and Dutler, H. (1970). Specificity of a-chymotrypsin dipeptide substrates. FEES Letters 8, 257-260. Bergeron, T. and Woodward, B. (1982). The development of the stratum compactum of ·7 the small intestine of the rainbow trout (Samo gairdneri). Canadian Journal ofZoology 60, 1513-1516. '{ .\ Bergman, T., Agerberth, B. and Jomvall, H. (1991). Direct analysis of peptides and amino acids from capillary electrophoresis. FEES Letters 283, 100-103. Bergot, P., Solari, A. and Luquent, P. (1975). Surfaces of pyloric caeca and intestine in rainbow trout (Salmo gairdneri, Richard). Annuals ofHydrobiology 6, 27-43 . .\ Bernkop-Schniirch, A. and Apprich, I. (1997). Synthesis and evaluation of a modified mucoadhesive polymer protecting from a-chymotrypsinic degradation. 146, 247-254. Bernkop-Schniirch, A. and Krajicek, M.E. (1998). Mucoadhesive polymers for peroral peptide delivery and absorption: synthesis and evaluation of different chitosan-EDTA conjugates. Journal of Controlled Release 50, 215-223. ,, Bernkop~Schniirch, A. (1998). The use of inhibitory gents to overcome the enzymatic barrier to perorally administered therapeutic peptides and proteins. Journal of Controlled Release 52, 1-16. Beyon, R. and Kay, I. (1976). Enteropancreatic circulation of digestive enzymes. Nature 260, 78-79. 180 BFGoodrich. (1994a). Carbopol product information guide. BFGoodrich, Specialty Polymers and Chemicals Division, Cleveland. Bulletin 2. BFGoodrich. (1994b ). Carbopol product information guide. BFGoodrich, Specialty Polymers and Chemicals Division, Cleveland. Bulletin 11. ; l~ ( ''( Bieth, J., Spiess, B. and Wermuth, C.G. (1974). The synthesis and analytical use of a highly sensitive and convenient substrate of elastase. Biochemical Medicine 11, "' 350-357. -~ I "( .\ Birk, Y. (1985). The Bowman-Birk inhibitor. International Journal ofPeptide and ! ' Protein Research 25, 113-131. ''\ Bloor, J.R., Morrison, J.C. and Rhodes, C.T. (1970). Effect of pH on micellar properties of a nonionic surfactant. Journal ofPharmaceutical Sciences 59, 387-391. l Blow, D.M. (1971). The structure of chymotrypsin, In: The Enzymes. Ed: P.D. Boyer, Academic Press, New York. p. 185-212. Bockman, D.E. and Cooper, M.D. (1973). Pinocytosis by epithelium associated with r lymphoid follicles in the Bursa of Fabricus, appendix and Peyer's patches. An electron ' ' ' microscopy study. American Journal ofAnatomy 136, 455-478. ~- Bode, W. and Schwager, P. (1975). The refined crystal structure of bovine beta-trypsin at ,, 1.8 A resolution. IL Crystallographic refinement, calcium binding site, benzamidine binding site and active site at pH 7 .0. Journal ofMolecular Biology 98, 693-717. I ~, I . Bohe, M., Borgstrom, A., Genell, S. and Ohlsson, K. (1983). Determination of imrnunoreactive trypsin, pancreatic elastase and chymotrypsin in extracts of human feces and ileostomy drainage. Digestion 27, 8-15. 181 " 131 Bohe, M., Borgstrom, A., Genell, S. and Ohlsson, K. (1986). Metabolism of I-labelled human pancreatic cationic trypsin after intraduodenal adminictration. Digestion 34, 127- 135. Bohely, P. and Seglen, J. (1992). Proteases and proteolysis in the lysosome. Experimentia ( 48, 151-153. 1 x Borchard, G., LueBen, H.L., de Boer, A.G., Verhoef, J.C., Lehr, C.-M. and Junginger, H.E. (1996). The potential ofmucoadhesive polymers in enhancing intestinal peptide ~ drug absorption. III: Effects of chitosan-glutamate and carbomer on epithelial tight junctions in vitro. Journal of Controlled Release 39, 131-138. Borey, G. and Yanari. P. (1960). Pepsin. In: The Enzymes. Eds: P.D. Boyers, H. Lardy, and K. Myrback, Academic Press, New York. p. 63. -' Breton, B., Roelants, I., Mikolajczyk, T., Epler, P. and Ollevier, F. (1995). Induced )_ spawning in teleost fish after oral administration of GnRH-A. In: Repoductive Physiology ofFish, 5th international symposium, Eds: Austin, Texas. p. 37. Burgi, D.S. and Chein, R.L. (1991). Optimization in sample stacking for high performance capillary electrophoresis. Analytical Chemistry 63, 2042-2047. ~ Burton, P.S., Conradi, R.A. and Hilgers, A.R. (1991). (B) Mechanism of peptide and protein absorption: (2) transcellular mechanism of peptide and protein absorption: passive aspects. Advanced Drug Delivery Reviews 7, 365-386. Burton, P.S., Conradi, R.A., Hilgers, A.R. and Ho, N.F.H. (1993). Evidence for a polarized efflux system for peptides in the apical membrane of Caco-2 cells. Biochemistry and Biophysical Research Communications 190, 760-766. 182 Camilleri, P., Okafo, G.N., Southan, C. and Brown, R. (1991). Analytical and micropreparative capillary electrophoresis of the peptides from calcitonin. Analytical Biochemistry 198, 36-42. ( Chadwick, V.S., Ganginella, T.S., Carlson, G.L., Debongnie, J.C., Philips, S.F. and I Hofmann, A.F. (1979). Effects of molecular structure on bile acid induced alterations in 1 adsorptive function, permeability, and morphology in the perfused rabbit colon. Journal ofLaboratory Clinical Medicine 94, 661-674. ' Chan, K.M., Davies, P.L., Childs, S., Veinot, L. and Ling, V. (1992). P-glycoprotein genes in the winter flounder, Pleuronectes americanus: isolation of two types of genomic ' clones carrying 3' terminal exons. Biochimica et Biophysica Acta 1171, 65-72. i \ Charman, W.N., Porter, C.J.H., Mithani, S. and Dressman, J.B. (1997). Physicochemical ', and physiological mechanisms of the effects of food on drug adsorption: The role of J, lipids and pH. Journal ofPharmaceutical Sciences 86, 269-283. ~ } Cheeseman, C.I. and Parsons, D.S. (1974). Intestinal absorption of peptides. Peptide ( uptake by small intestine of Rana pipiens. Biochimica et Biophysica Acta 373, 523-526. A Cho, M.J., Scieszka, J.F. and Burton, P.S. (1989). Citric acid as an adjuvant for ~ transepithelial transport. International Journal ofPharmaceutics 52, 79-81. ;., Choulis, N.H. and Loh, L.H. (1971). Critical micelle concentration determination of some nonionic surfactants using interface refractometry. Canadian Journal ofPharmaceutical Science 6, 93-96. 183 Chu, S.W. and Walker, W.A. (1991). Growth factor signal transduction in human intestinal cells. In: Immunology of milk and the neonate. Ed: J. Mestecky, Plenum, New York. p. 107-112. Clarke, J., MacDonald, N.L. and Stark, J.R. (1985). Metabolism in marine flatfish-II. Protein digestion in Dover sole (Solea solea L.). Comparative Biochemistry and Physiology B 81, 217-222. -I " Cohen, T., Gertler, A. and Birk, Y. (1981a). Pancreatic proteolytic enzymes from carp ( Cyprinus carpio )-I. Purification and physical properties of trypsin, chymotrypsin, .> elastase and carboxypeptidase B. Comparative Biochemistry Physiology B 69, 639-646 . Cohen, T., Gertler, A. and Birk, Y. (1981b). Pancreatic proteolytic enzymes from carp ( Cyprinus carpio )-II. Kinetic properties and inhibition studies of trypsin, chymotrypsin ~ and elastase. Comparative Biochemistry Physiology B 69, 647-653. .\ Conradi, R.A., Hilgers, A.R., Ho, N.F.H. and Burton, P.S. (1991). The influence of .\ peptide structure on transport across Caco-2 cells. Pharmaceutical Research 8, 1453-1460. Constantinides, P.P., Scarlet, J.-P., Lancaster, C., Marcello, J., Marks, G., Ellens, H. and Smith, P.L. (1994). Formulation and intestinal absorption enhancement evaluation of water-in-oil microemulsions incorporating medium-chain glyceride. Pharmaceutical ,\_ Research 11, 1385-1390. :.:,, Cornell, R., Walker, W.A. and Isselbacher, K.J. (1971). Small intestinal adsorption of horseraq.ish peroxidase. Acytochemical study. Laboratory Investigations 25, 42-48. Croston, C.B. (1960). Tryptic enzymes of chinook salmon. Archives ofBiochemistry and Biophysics 89, 202-206. 184 Cunningham, L. (1954). Molecular-kinetics properties of crystalline diisopropyl phosphoryl trypsin. Journal ofBiological Chemistry 211, 13-16. Davina, J.H.M., Rijkers, G.T., Rombout, J.H.W.M., Timmermans, L.P.M. and Van Muiswinkel, W.B. (1980). Lymphoid and non-lymphoid cells in the intestine of cyprinid fish, In: Development and Differentiation ofthe Vertebrate Lymphocytes. Ed: J.D. 1 Horton, Elsevier: Amsterdam. p. 129-140. I, Desuelle, P. (1960). Chymotrypsin. In: The Enzymes. Eds: P.D. Boyers, H. Lardy, and K. > Myrback, Academic Press: New York. p. 93. Di Constanzo, G., Florentz, A., Leray, C. and Nonnotte, L. (1983). Structural and functional organization of the brush border membrane in the rainbow trout intestine. S, Molecular Physiology 4, 111-123. )- Diamond, J.M. (1978). Reabsorption of digestive enzymes: playing with poison. Nature \. 271, 111-112. Dionex (1989). Capillary electrophoresis System I Operator's Manual, D.C., Document No. 034213, Revision 03. Donaldson, E.M. and Hunter, G.A. (1983). Induced final maturation, ovulation and spermiation in cultured fish. In: Fish physiology. Eds: W.S. Hoar, D.J. Randall, and E.M. s Donaldson. Vol. IXB. New York: Academic press. p. 351-403. Dorin, D., Martin, P., Sire, M.F., Smal, J. and Vernier, J.M. (1993). Protein uptake by intestinal macrophages and eosinophilic granulocytes in trout: An in vitro study. Biology ofthe Cell 19, 37-44. 185 Dorin, D., Sire, M.F. and Vernier, J.M. (1993). Endocytosis and intracellular degradation ofheterologous protein by eosinophilic granulocytes isolated from rainbow trout (Oncorhynchus mykiss) posterior intestine. Biology of the Cell. 79, 219-224. Dose, E.V. and Guiochon, G.A. (1991). Internal standardization technique for capillary zone electrophoresis. Analytical Chemistry 61, 292-303. r Einarsson, S., Davies, P.S. and Talbot, C. (1996). The effect of feeding on the secretion 'y of pepsin, trypsin and chymotrypsin in the Atlantic salmon, Salmo salar L. Fish Physiology and Biochemistry 15, 439-446. Elias, A.N., Gordon, I., Vaziri, N.D., Chune, G., Pandian, M.R., Gwinup, G. and Wesley, R. (1992). Effective portal insulin delivery with enzyme-protected capsules in pancreatectomized pigs. General Pharmacology 23, 55-59 . 1 . Elwing, H., Askendal, A. and Lundstrom, A. (1987). Competition between adsorbed fibrinogen and high-molecular-weight kininogen on solid surfaces incubated in human I, plasma (the Vroman effect): influence of solid surface wetability. Journal ofBiomedical Material Research 21, 1023-1028. Erlanger, B.F., Kokowsky, N. and Cohen, W. (1961). The preparation and properties of two new chromogenic substrates of trypsin. Archives ofBiochemistry and Biophysics 95, ,.. 271-278. Ezeasor, D.N. and Stoke, W.M. (1980). Scanning electron microscope study of the gut :>\ mucosa ofrainbow trout Salmo gairdneri Richardson. Journal ofFish Biology 17, 529-539. 186 Ezeasor, D.N. and Stokoe, W.M. (1980). A cytochemical, light and electron microscopic study of the eosinophilic granuale cells in the gut of the rainbow trout, Salmo gairdneri t,,, Richardson. Journal ofFish Biology 17, 619-634. Ezeasor, D.N. and Stokoe, W.M. (1981). Light and electron microscopic studies on the ,( absorptive cells of the intestine, caeca and rectum of the adult rainbow trout, Salmo gairdneri, Rich. Journal ofFish Biology 18, 527-544. ~ ly Fange, R. and Grove, D. (1979). Digestion, In: Bioenergetics and Growth. Eds: W.S. ·,\ < Hoar, D.J. Randall, and E.M. Donaldson, Academic Press: New York. P 34-56. I.;:· Fletcher, T.S., Alhadeff, M., Craik, C.S. and Largman, C. (1987). Isolation and i characterization of a cDNA encoding rat cationic trypsinogen. Biochemistry 26, 3081-3086. .\ Florence, A.T. and Attwood, D. (1998). Factors affecting the critical micelle 'j concentration and micelle size. In: Physiochemical Principles ofPharmacy. Macmillan Press Ltd: London. p. 230-233. r ; I, ¥ I ·, Folk, J.E., Piez, K.A., Carroll, W.R. and Gladner, J.A. (1960). Carboxypeptidase B IV. Purification and characterization of porcine enzyme. Journal ofBiological Chemistry t • 235, 2272-2277. "' ·\ Food and Agriculture Organization Fisheries Department. (1997). The state of the world ' . fisheries and aquaculture. Food and Agriculture Organization of the United Nations. ·\ Fraisse, M., Woo, N.Y.S., Noaillac-Depeyre, J. and Murat, J.C. (1981). Distribution ,\ pattern of digestive enzyme activities in the intestine of the catfish (Ameiurus nebulous L.) and of the carp (Cyrinus carpio L.). Comparative Biochemistry and Physiology A 70, 443-446. 187 Fricker, G., Drewe, J., Huwyle, J., Gutman, H. and Beglinger, C. (1996). Relevance of P-glycoprotein for the entemal absorption of cyclosporin A: in vitro and in vivo correlation. British Journal ofPharmacology 118, 1841-1847. Fujita, T., Fujita, T., Morikawa, K., Tanaka, H., Iemura, 0., Yamamoto, A. and Muranishi, S. (1996). Improvement of intestinal absorption of human calcitonin by '; chemical modification with fatty acids: synergistic effects of acylation and absorption y enhancers. International Journal ofPharmaceutics 134, 47-57. ,, Fukunaga, M., Miller, M.M. and Deftos, L.J. (1991). Liposomes entrapped calcitonin and parathyroid hormone are orally effective in rats. Hormone Metabolism Research 23, 166-167. & Gabel, D. and Kasche, V. (1973). Autolysis of p-trypsin Influence of calcium ions and heat. Acta Chemica Scandinavica 27, 1971-1981. "' Ganapathy, V. and Leibach, F.H. (1983). Role of pH gradient and membrane potential in dipeptide transport in intestinal and renal brush-border membrane vesicles from the 1 rabbit. Studies with L-camosine and glycyl-L-proline. Journal ofBiological Chemistry 258, 14189-14192. Ganapathy, V., Mendicino, J. and Leibach, F.H. (1981). Evidence for a dipeptide transport system in renal brush border membranes from rabbit. Biochimica et Biophysica />- Acta 642, 381-339. >- Genell, S., Gustafsson, B.E. and Ohlsson, K. (1977). Immunochemical quantitation of .\ pancreatic endopeptidase in the intestinal contents of germfree and convential rats. Scandinavian Journal ofGastroenterology 12, 811-820. 188 Georgopoulou, U., Sire, M.-F. and Vernier, J.-M. (1985). Macromolecular absorption of proteins by epithelial cells of the posterior intestinal segment and their intracellular digestion in the rainbow trout. Ultrastructural and biochemical study. Biology of the Cell 53, 269-282. Georgopoulou, U., Sire, M.F. and Vernier, J.M. (1986). Immunological demonstration of intestinal absorption and digestion of protein molecules in the trout (Salmo gairdneri). )' Cell Tissue Research 245, 387-395. )' Georgopoulou, U. and Vernier, J.M. (1986). Local immunological response in the posterior intestinal segment of the rainbow trout after oral administration of ·, > macromolecules. Development and Comparative Immunology 10, 529-537. Georgopoulu, U., Dabrowski, K. and Vernier, J.M. (1988). Absorption of intact proteins f ', by the intestinal epithelium of trout, Salmo gairdneri. A luminesence enzyme I ., immunoassay and cytochemical study. Cell Tissue Research 251, 145-52. .\ Ghose, S. and Bull, H.B. (1962). Adsorption and enzymatic activity of chymotrypsin at :; an n-octadecane-water interface. Archives ofBiochimica and Biophysics 99, 121-125. I i ·i, Giacomelli, C.E., Avena, M.J. and De Pauli, C.P. (1997). Adsorption of bovine albumin •' onto Ti02 particles. Journal of Colloid and Interface Science 188, 387-395. \ Gibson, R.S., McFarlane, C., Hay, S. and G.T., M. (1989). Significance ofmicroflora in > proteolysis in the colon. Applied and Enviromental Microbiology 55, 679-683. ..>. Goldberg, D.M., Campell, R. and Roy, A.D. (1969). Studies on the binding of trypsin and chymotrypsin by human intestinal mucosa. Scandinavian Journal of Gastroenterology 4, 217-226. 189 Gonnella, P.A. and Neutra, M.R. (1984). Membrane-bound and fluid-phase macromolecules enter separate prelysosomal compartments in adsorptive cells of suckling rat ileum. Journal of Cell Biology 99, 909-917. Goren, A., Zohar, Y., Fridkin, M. and Elhanati, E. (1990). Degradation of gonadotropin I releasing hormones in the Gilthead Seabream, Sparus aurata 1. Cleavage of native ''i salmon GnRH and mammalian LHRH in the pituitary. General and Comparative \ r Endocrinology 291-305. ' 19, ; y ., Green, J.S. and Jorgenson, J.W. (1989). Minimizing adsorption of proteins on fused silica / in capillary zone electrophoresis by the addition of alkali metal salts to the buffers. > Journal of Chromatography 478, 63-70. ; \, ',1. Grossman, P.D., Colburn, J.C., Lauer, H.H., Nielsen, R.G., Riggin, R.M., Sittampalam, G.S. and Rickard, E.C. (1989). Application of free-solution capillary electrophoresis to " \ the analytical scale separation of proteins and peptides. Analytical Chemistry 61, 1186- 1194. t 1 Grossman, P.D., Kenneth, J., Petrie, G. and Lauer, H.H. (1988). Effect of buffer pH and peptide composition on the selectivity of peptide separations by capillary zone f electrophoresis. Analytical Biochemistry 173, 265-270. ;,- Gu, J.-M., Robinson, J.R. and Leung, S.-H.S. (1988). Binding of acrylic polymers to \ mucin/epithelial surface: structure-property relationships. Critical Reviews in Therapeutic ,:r Drug Carrier Systems 5, 21-67. ,, Gumbiner, B. (1987). Structure, biochemistry, and assembly of epithelial tight junctions. ' American Journal ofPhysiology 253, C749-758. 190 ,., Hajjou, M. and Le Gal, Y. (1994). Purification and characterization of an aminopeptidase from tuna (Thunnus albacures) pyloric caeca. Biochimica et Biophysics Acta 1204, 1-13. Han, K., Park, J.S., Chung, Y.B., Lee, M.J., Moon, D.C. and Robinson, J.R. (1995). Identification of enzymatic degradation products of luteinzing hormone releasing ,, hormone (LHRH)/[D-Ala6] LHRH in rabbit mucosa! homogenates. Pharmaceutical f './ Research 12, 1539-1544. I , . Harris, D., Fell, J.T., Taylor, D.C., Lynch, J. and Sharma, H.L. (1990). GI transit of / potential bioadhesive formulations in the rat. Journal of Controlled Release 12, 55-65. I 1 Hashimoto, M., Takada, K., Kiso, Y. and Muranishi, S. (1989). Synthesis of palmitoyl " derivatives of insulin and their biological activities. Pharmaceutical Research 6, 171-176. \ Hashizume, M., Douen, T., Murakami, M., Yamamoto, A., Takada, K. and Muranishi, S. '""' \ (1992). Improvement oflarge intestinal absorption of insulin by chemical modification with palmitic acids in rats. Journal Pharmacy and Pharmacology 44, 555-559. l- II '( Raviv, F., Fitzpatrick, T.D., Nichols, C.J., Swenson, RE., Bush, E.N., Diaz, G., Nguyen, A., Nellans, H.N., Hoffman, D.J., Ghanbari, H., Johnson, E.S., Love, S., Cybulski, V. and I Greer, J. (1992). Stabilization of the N-terminal residues ofluteinizing hormone-releasing hormone agonists and the effect on pharmacokinetics. Journal ofMedicinal Chemistry I ~ 35, 3890-3894. \ >· ~ Hedstrom, L., Perona, J.J. and Rutter, W.J. (1994). Converting trypsin to chymotrypsin: I , residue 172 is a substrate specificity determinant. Biochemistry 33, 8757-8763. ; .\ Reiger, D.N. (1992). High performance capillary electrophoresis- An introduction. France: Hewlett-Parkard Company. 191 Hemmer, M.J., Courtney, L.A. and Ortego, L.S. (1995). Immunohistochemical detection of P-glycoprotein in teleost tissues using mammalian polyclonal and monoclonal antibodies. Journal ofExperimental Zoology 272, 69-77. Hertz, Y., Tchelet, A., Madar, Z. and Gertler, A. (1991). Absorption ofbioactive human ,1 growth hormone after oral administration in the common carp ( Cyrinus carpio) and its '{ enhancement by deoxycholate. Journal of Comparative Physiology B 161, 159-163. ,, 'l . Hirai, S., Yashiki, T. and Mirna, H. (1981). Mechanism for enhancement of the nasal absorption of insulin by surfactants. International Journal Pharmacy 9, 173-184. "' .', Hirji, K.N. and Courtney, W.A. (1982). Leucine aminopeptidase activity in the digestive tract of perch, Percafluviatils L. Journal ofFish Biology 21, 615-622. Hjelmeland, K. and Raa, J. (1982). Characteristics of two trypsin type isoenzymes \ isolated from the artic fish capeilin (Ma/lotus vill). Comparative Biochemistry and > Physiology B 71, 557-562. Hjerten, S. (1985). (High-performance electrophoresis) Elimination of electroendosmosis and solute adsorption. Journal of Chromatography 347, 191-198. \ "' '! ,. Hofer, R. (1982). Protein digestion and proteolytic activity in the digestive tract of an I> omnivorous cyprinid. Comparative Biochemistry and Physiology A 72, 55-63. ,, ~ Hofer, R. and Schiemer, F. (1981). Proteolytic activity in the digestive tract of several species of fish with different habits. Oecologia 48, 342-345. i> Holmes, B. (1996). Blue revolutionaries. New Scientist 152, 32-36. 192 Huang, X., Coleman, W.F .. and Zare, R.N. (1989). Analysis of factors causing peak broadening in capillary zone electrophoresis. Journal of Chromatography 480, 95-100. Huang, Y.Y. and Wu, Y.Y. (1996). Stability of peptides during iontophoretic transdermal delivery. International Journal ofPharmaceutics 131, 19-23. Iida, H. and Yamamoto, T. (1985). futracellular transport of horseradish peroxidase in absorptive cells of goldfish hindgut in vitro, with special reference to the cytoplasmic ·r tubules. Cell Tissue Research 240, 553-560. ,, Ikesue, K., Kopeckova, P. and Kopecek, J. (1993). Degradation of proteins by guinea pig ·' intestinal enzymes. International Journal ofPharmaceutics 95, 171-179. Ilan, E., Amselem, S., Weisspapir, M., Schwartz, J., Yogev, A., Zawoznik, E. and '.- Friedman, D. (1996). Improved oral delivery of desmopressin via a novel vehicle: .~ mucoadhesive submicron emulsion. Pharmaceutical Research 13, 1083-1087 . \. ~' } Illum, L., Farraj, N.F. and Davis, S.S. (1994). Chitosan as a novel nasal delivery system for peptide drugs. Pharmaceutical Research 11, 1186-1189. > \ V 11 Ishizawa, T., Hayashi, M. and Awazu, S. (1987). Enhancement of jejunal and colonic ! ~ ,.. absorption of fosfomycin by promotors in the rat. Journal ofPharmacy and ,> Pharmacology 39, 892-895. \_ Ito, K., Irie, Y., Hagiwara, T., Sakai, Y., Hayashi, M., Sakakibara, T., Kishi, H., Okada, :, M., Sakamoto, M. and Suzuki, M. (1983). Toxicological studies on bestatin. II. Subacute ~ toxicity test and recovery study in beagle dogs. Japanese Journal ofAntibiotics 36, 2985-3052. 193 Jorgenson, J.W. and Lukacs, K. (1983). Capillary zone electrophoresis. Science 222, 266-272. Junginger, H.E., Lehr, C.-M., Bouwstra, J.A., Tukker, J.J. and Verhoef, J.C. (1990). Specific site intestinal absorption using bioadhesives: improved oral delivery of peptide .1 drugs by means of bioadhesive polymers. In: Bioadhesion-Possibilities and Future '( Trends. Eds: R. Gurny and H.E. Junginger, Paperback APV 25: Wissenschaftliche Verlagsgesellschaft, Stuttgrat. p. 177-190. 7 . Karls, M.S., Rush, B.D., Wilkinson, K.F., Vidmar, T.J., Burton, P.S. and Ruwart, M.J. l (' '.\ (1991). Desolvation Energy: A major determinant of absorption, but not clearance, of ·' peptides in rats. Pharmaceutical Research 8, 1477-1481. !, Kassell, B., Radicevic, M., Ansfield, M. and Lasowski, M. (1965). The basic trypsin inhibitor of bovine pancreas. IV. The linear sequence of the 58 amino acids. Biochemistry Biophysics Research Communications 18, 255-259. ' J Keil-Dlouha, V., Zylber, N_., Tong, N. and Keil, B. (1971). Cleavage of glucagon by a and ~-trypsin. FEBS Letters 16, 287-289. -~ ., Kidron, M., Bar-On, H., Berry, E.M. and Ziv, E. (1982). The absorption of insulin from .. various regions of the rat intestine. Life Sciences 31, 2837-2842 . /> > King, S.W., Lum, V.R. and Fife, T.H. (1987). Interaction of carboxypeptidase A with \. carbamate and carbonate esters. Biochemistry 26, 2294-2300. > Koldovsky, 0., Britton, J., Davies, D., Davies, T., Grimes, J., King, W., Roa, R. and \ Schaudies, R. (1991). The developing gastrointestinal tract and milkborne epidermal growth factor. In: Immunology ofmilk and the neonate. Ed: J. Mestecky, Plenum, New York. p. 99-106. 194 ~ Kotze, A.F., de Leeuw, B.J., LueBen, H.L., de Boer, A.G., Verhoef, J.C. and Junginger, H.B. (1997). Chitosans for enhanced delivery of therapeutic peptides across the intestinal epithelia: in vitro evaluation in Caco-2 cell monolayers. International Journal of Pharmaceutics 159, 243-253. ,, '( Kotze, A.F., LueBen, H.L., de Leeuw, H.L., de Boer, A.G., Verhoef, J.C. and Junginger, H.B. (1997). N-trimethyl chitosan chloride as a potential absorption enhancer across " . mucosal surfac:es: In vitro evaluation in intestinal epithelial cells (Caco-2) . Pharmaceutical Research 14, 1197-1202. I " ;~ Kramer, W., Wess, G., Neckermann, G., Schubert, G., Fink, J., Girbig, F., Gutjahr, U., Kowalewski, S., Baringhaus, K.-H., Boger, G., Bnhsen, A., Falk, B., Friedrich, M., !, Glombik, H., Hoffinan, A., Pittius, C. and Urmann, M. (1994). Intestinal absorption of .,__ peptides by coupling to bile acids. Journal ofBiologica!Chemistry 269, 10621-10627 . :;,, ,, Krementz, A.B. and Chapman, G.B. (1975). Ultrastructure of the posterior half of the _J intestine of the channel catfish, Ictalurus punctatus. Journal ofMorphology 145, ~ 441-449. ,; j Kristjansson, M.M. and Nielsen, H.H. (1992). Purification and characterization of two ~ chymotrypsin-like proteases from the pyloric caeca of rainbow trout (Oncorhychus t>- mykiss). Comparative Biochemistry and Physiology B 101, 247-253. ;, .> Krogdahl, A. and Holm, H. (1983). Pancreatic proteinases from man, trout, rat, pig, cow ·>- chicken, mink and fox. Enzyme activities and inhibition by soybean and lima bean proteinase inhibitors. Comparative Biochemistry and Physiology B 74, 403-409. 195 Krogdahl, A., Lea, T.B. and Olli, J.J. (1994). Soybean proteinase inhibitors affect intestinal trypsin activities and amino acid digestabilities in rainbow trout ( Oncorhynchus mykiss). Comparative Biochemistry Physiology A 107, 215-219. Kruntiz, M. (1947). Crystalline soybean trypsin inhibitor. Journal of General Physiology 30,291-311. Laemmli, U.K. and Farve, M. (1973). Maturation of the head of bacteriophage T4. r Journal ofMolecular Biology 80, 3647-3649. ,' \ Lambert, W.J. and Middleton, D.L. (1990). pH hysteresis effect with silica capillaries in .~ capillary zone electrophoresis. Analytical Chemistry 62, 1585-1587 . Landers, J.P. (1993). Capillary electrophoresis: pioneering new approaches for 18, .,-_' biomolecular analysis. Trends in Biochemical Sciences 409-414. ~ Lang, V.B., Langguth, P., Ottiger, C., Wunderli-Allenspach, H., Rogan, D., Rothen Rutishauser, B., Perriard, J.-C., Lang, S., Biber, J. and Meerkle, H.P. (1997). Structure ~· permeation relations ofMet-enkephalin peptide analogues on absorption and secretion t· • mechanisms in Caco-2 monolayers. Journal ofPharmaceutical Sciences 86, 846-853 . Langmuir, I. (1918). The adsorption of gases on plane surfaces of glass, mica and ,> platnium. Journal ofAmerican Chemical Society 40, 1361-1368. ~ Lanzillo, J.J., Dasaarthy, Y., Stevens, J. and Fanburg, B.L. (1986). Conversion of angiotensin"'.1 to angiotensin-2 by a latent endothelial cell peptidyl dipeptidase that is ~ · not angiotensin-converting enzyme. Biochemistry Biophysics Research Communications 134, 770-776. 196 Lasdun, A., Renzink, S., Molineaux, C.J. and Orlowski, M. (1989). Inhibition of endopeptidase 24.15 slows the site in vivo degradation of luteinizing hormone releasing hormone. American Pharmacology Experimental Therapeutics 251, 439-447. Layer, P., Go, V.L.W. and DiMagno, E.P. (1986). Fate of pancreatic enzymes during small intestinal aboral transit in humans. American Journal ofPhysiology 14, G475-480. ' { " Layer, P., Jansen, J.B.M., Cherian, L., Lamers, C.B. and Goebell, H. (1990). Feedback r regulation of human pancreatic secretion. Gastroenterology 98, 1311-1319. ~ LeBail, P.Y., Sire, M.F. and Vernier, J.M. (1989). Intestinal transfer of growth hormone :> into the circulatory system of the rainbow trout, Sa/mo gairdneri: interference of granule cells. Journal ofExperimental Zoology 251, 101-107. \ Lee, V.H.L., Traver, R.D. and Taub, M.E. (1991). Enzymatic barriers to peptide and ·"- protein drug delivery. In: Peptide and protein drug delivery. Ed: J.R. Robinson, Marcel t, Dekker Inc: New York. p. 303-358. . Lehr, C.-M., Bouwstra, J.A., Kok, W., de Boer, A.G., Tukker, J.J., Verhoef, J.C., ·•· Breimer, D.D. and Junginger, H.E. (1992a). Effects of the mucoadhesive polymer ' -, polycarbophil on the intestinal absorption of a peptide drug in the rat. Journal of .. Pharmacy and Pharmacology 44, 402-407 . .• Lehr, C.-M., Bouwstra, J.A., Schacht, E.H. and Junginger, H.E. (1992b). In vitro evaluation of mucoadhesive properties of chitosan and some other natural polymers. :>, International Journal ofPharmaceutics 78, 43-48. •\ Leo, A., Hansch, C. and Elkins, D. (1977). Partition coefficients and their uses. Chemical Reviews 71, 525-616. 197 Lester, R., Pyrek, J., Little, J.M. and Adcock, E.W. (1983). What is meant by the term "bile acid"? American Journal ofPhysiology 249, Gl07-l 10. Lin, S. and Van Wart, H. (1982). Efffect of cryosolvents and subzero temperatures on the hydrolysis of L-leucine-p-nitroanilide by porcine kidney leucine aminopeptidase. ·.I Biochemistry 21, 5528. ' . \ i Lindahl, A., Ungell, A. L., Knutson, L. and Lennema:s, H. (1997). Characterization of r fluids from the stomach and proximal jejunum in men and women. Pharmaceutical ~ .. .. ,, Research 14, 497-502 . Lue.Ben, H.L., Bohner, V., Perard, D., Langguth, P., Verhoef, J.C., de Boer, A.G., Merkle, "' H.P. and Junginger, H.E. (1996a). Mucoadhesive polymers in peroral peptide drug delivery. V. Effect ofpoly(acrylates) on the enzymatic degradation of peptide drugs by .j intestinal brush border membrane vehicles. International Journal ofPharmaceutics 141, .'>. 39-52 . ;. Lue.Ben, H.L., de Leeuw, B.J., Langemeyer, M.W.E., de Boer, A.G., Verhoef, J.C. and ', ., Junginger, H.E. (1996b). Mucoadhesive polymers in peroral peptide drug delivery. VI. Carbomer and chitosan improve the intestinal absorption of the peptide drug buserelin ' F( in vivo. Pharmaceutical Research 13, 1668-1672. I I • Lue.Ben, H.L., de Leeuw, B.J., Perard, D., Lehr, C.-M., de Boer, A.G., Verhoef, J.C. and I), {.>-. Junginger, H.E. (1996c). Mucoadhesive polymers in peroral peptide drug delivery. I. i\ I\ ~ Influence ofmucoahesive excipients on the proteolytic activity of intestinal enzymes. ! I > European Journal ofPharmaceutical Sciences 4, 117-128. I •\ Lue.Ben, H.L., Lehr, C.-M., Rentel, C.-0., Noach, A.B.J., de Boer, A.G., Verhoef, J.C. and Junginger, H.E. (1994). Bioadhesive polymers for the peroral delivery of peptide drugs. Journal of Controlled Release 29, 329-338. 198 J LueBen, H.L., Rentel, C.-0., Kotze, A.F., de Boer, A.G., Verhoef, J.C. and Junginger, H.B. (1997). Mucoadhesive polymers in peroral peptide drug delivery. IV. Polycarbophil ,, and chitosan are potent enhancers of peptide transport across intestinal mucosa in vitro. ,\ Journal of Controlled Release 45, 15-23. ' ·i LueBen, H.L., Verhoef, J.C., Borchard, G., Lehr, C.-M., de Boer, A.G. and Junginger, r H.B. (1995). Mucoadhesive polymers in preoral peptide drug delivery. II. Carbomer and r . polycarbophil are potent inhibitors of the intestinal proteolytic enzyme trypsin . -, Pharmaceutical Research 12, 1293-1298. ,~ Luey, J.-K., Mc Guire, J. and Sproull, R.D. (1991). The effect of pH and NaCl "' concentration on adsorption of P-lactoglobulin at hydrophilic and hydrophobic silicone .., surfaces. Journal of Colloid and Interface Science 143, 4898-499 . .~ >. Luniden, S., Pantzar, N., Broeders, A., Ohilin, M. and Westrom, B.R. (1991). Differences " in transport rate of oxytocin and vasopressin analogues across the proximal and distal isolated segments of the small intestine of the rat. Pharmaceutical Research 8, i' 1274-1280. 'I ,, Ma, T.Y. Hollander, D. Riga, R. Bhalla, D. (1993). Autoradiographic determination of /> permeation pathway of permeability probes across intestinal and tracheal epithelia. Journal ofLaboratory and Clinical Medicine 122, 590-683. ',,, MacCann, S.M. (1977). Luteinizing-hormone-releasing-hormone. Physiology in Medicine ,>, 296, 797-802. Macfarlane, G.T., Allison, C., Gibson, S.A. and Cummings, J.H. (1988). Contribution of the microflora to proteolysis in the human large intestine. Journal ofApplied Bacteriology 64, 37-46 199 ;' Marcon-Genty, D., Tome, D., Kheroua, 0., Dumontier, A.M., Heyman, M. and Desjeux, ' J.F. (1989). Transport of P-lactoglobulin across rabbit ileum in vivo. American Journal of ,, Physiology 256, G943-943 Maroux, S. and Desnuelle, P. (1969). On some autolyzed derivatives of bovine trypsin. Biochimica et Biophysica Acta 181, 59-72. I "f Martinez, A., Olsen, R.L. and Serra, J.L. (1988). Purification and characterization of two ; ,. '""\ trypsin-like enzymes from the digestive tract of anchovy Engraulis encracicholus. Comparative Biochemistry Physiology B 91, 677-684. ·"-· Martinez, A. and Serra, J.L. (1989). Proteolytic activities in the digestive tract of anchovy (Engraulis encrasicholus). Comparative Biochemistry Physiology B 93, 61-66. "' ,\ Maste, M.C.L., Rinina, H.A., Brands, C.M.J., Egmond, M.R. and Norde, W. (1995). ·>, Inactivation of a subtilisin in colloid systems. Biochimica et Biophysica Acta 1252, 261-268. ?- ,-, .,. McCormick, R.M. (1988). Capillary zone electrophoretic separation of peptides and I r;.( proteins using low pH buffers in modified silica capillaries. Analytical Chemistry 60, I> 2322-2328. _.. McLaughlin, G.M., Nolan, J.A., Lindahl, J.L., Palmieri, R.H., Anderson, K.W., Morris, ·.,, S.C., Morrison, J.A. and Bronzert, T.J. (1992). Pharmaceutical drng separations by >, HPCE: Practical guidelines. Journal ofLiquid Chromatography 15, 961-1021. McLean, E. and Ash, R. (1986). The time-course of appearance and net accumulation of horseradish peroxidase (HRP) presented orally to juvenile carp Cyprinus carpio (L). Comparative Biochemistry and Physiology A 88, 687-690. 200 ,I McLean, E., Parker, D.B., Warby, C.M., Sherwood, N.M. and Donaldson, E.M. (1991). Gonadotropin release following oral delivery of luteinizing hormone-releasing hormone 10 6 and its superactive analogue (des-Gly [D-Ala ] LHRH ethylamide) to 17 P-oestradiol ,\ primed coho salmon, Oncorhynchus kitutch (Walbaum). Journal ofFish Biology 38, 851-858. " McLean, E. and Ash, R. (1990). Modified uptake of the protein antigen horseradish IY peroxidase (HRP), following oral delivery to rainbow trout, Oncorhynchus mykiss. ~ -; Aquaculture 87, 373-379. ·-" McLean, E. and Donaldson, E.M. (1990). Absorption ofbioactive proteins by the ,, gastrointestinal tract of fish: A review. Journal ofAquatic Animal Health 2, 1-11. ', McLean, E., Donaldson, E.M., Teskeredzic, E. and Souza, L.M. (1993). Growth ,',. enhancement following dietary delivery of recombinant porcine somatotropin to diploid ~ and triploid salmon (Oncorhynchus kisutch). Fish Physiology and Biochemistry 11, j ' 363-369. McLean, E., von der Meden, A.C. and Donaldson, E.M. (1990). Direct and indirect ' 4 )\ evidence for polypeptide absorption by the teleost gastrointestinal tract. Journal ofFish I> Biology 36, 489-498. " J, Mekhjian, H.S. and Phillips, S.F. (1970). Perfusion of the canine colon with ·,:. unconjugated bile acids. Effects on water and electrolyte transport, morphology, and bile ;, acid absorption. Gastroenterology 59, 120-129. "' Michel, C., Aprahamian, M., Defontaine, L., Couvreur, P. and Damge, C. (1991). The effect of site of administration in the gastrointestinal tract on the absorption of insulin from nanoparticles in diabetic rats. Journal ofPharmacy and Pharmacology 43, 1-5. 201 Milov, D.E., Jou, D.E., Shireman, R.B. and Chun, P.W. (1992). The effect of bile salts on > carbonic anhydrase. Hepatology 15, 288-296. ,\ Mistry, G.B. (1993). Iontophoretic transdermal delivery of LHRH. Hons. Thesis, -, University of Otago, New Zealand. I , I', \ I ;I Morimoto, K., Kamiya, E., Takeeda, T., Nakamoto, Y. and Morisaka, K. (1983). /y Enhancement of rectal absorption of insulin in polyacrylic acid aqueous gel bases , . -,- containing long chain fatty acid in rats. International Journal ofPharmaceutics 14, 149-157. ->, Moriyama, S., Yamamoto, H., Sugimoto, S., Abe, T., Hirano, T. and Kawauchi, H. :~( (1993). Oral administration ofrecombinant salmon growth hormone to rainbow trout, Oncorhynchus mykiss. Aquaculture 112, 99-106. )- f Morrisey, J.H. (1981). Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity. Analytical Biochemistry 117, 307-310. .{ Murakami, K. and Noda, M. (1981). Studies on proteinases from the digestive organs of sardine. I Purification and characterization of three alkaline proteinases from the pyloric I> cacea. Biochimica et Biophysics Acta 658, 17-26. -• _, Murray, L.C. and Tucker, I.G. (1990). Gastric residence times of enteric dosage forms. ',> Australian Journal ofHospital Pharmacy 20, 235-238. ,',-, ,- Nellans, H.N. (1991). Paracellular intestinal transport: modulation of absorption. Advances in Drug Delivery Reviews 1; 339-364. 202 Nishiata, T., Liversidge, G. and Higuchi, T. (1983). Effect of aprotinin on the rectal delivery of insulin Journal ofPharmacy and Pharmacology 35, 616-617. Noaillac-Depeyre, J. and Gas, N. (1979). Structure and function of the intestinal epithelial cells in the perch (Percafluviatilis L.). Anatomical Record 195, 621-640. ".\ Norde, W. (1986). Adsorption of proteins from solution at the solid-liquid interface. ;/ Advances in Colloid and Interface Science 24, 267-340. I 7' I ' Norde, W. and Zoungrana, T. (1998). Surface-induced changes in the structure and activity of enzymes physically immobilized at solid/liquid interfaces. Biotechnology and .J., Applied Biochemistry 28, 133-143. ~ Okada, H., Yamazaki, I., Ogawa, Y., Hirari, S., Yashiki, T. and Mirna, H. (1982). \. Vaginal absorption of a potent luteinizing hormone-releasing hormone analog ( ! ,,r (leuprolide) in rats. I. Absorption by various routes and absorption enhancement. Journal ~ ofPharmaceutical Science 71, 1367-1371. I 7 Okada, H., Yamazaki, I., Yashiki, T. and Mirna, H. (1983). Vaginal absorption of a " ' potent luteinizing hormone-releasing hormone analogue (leuprolide) in rats II:· Mechanism ofabsorption enhancement with organic acids. Journal ofPharmaceutical I> Sciences 72, 75-78. ,+ l Ok:agawa, T., Fujita, T., Murakami, M., Yamamoto, A., Shimura, T., Tabata, S., Kondo, I >- S. and Muranishi, S. (1994). Susceptibility of ebiratide to proteolysis in rat intestinal fluid I,'.> and homogenates and its protection by various protease inhibitors. Life Sciences 55, \ ' 677-683. ,' 203 Owen, R.L. (1977). Sequential uptake of horseradish peroxidase by lymphoid follicle epithelium of Peyer's patches in the normal unobstructed mouse intestine: an ultrastuctural study. Gastroenterology 72, 440-451. Pappenheimer, J.R. (1987). Physiological regualtion oftransepithelial impendence in the intestinal mucosal ofrat and hamster. Journal ofMembrane Biology 100, 137-148. I 1 / Pappenheimer, J.R. and Reiss, K.Z. (1987). Contribution of solvent drag through intercellular junctions to adsorption of nutrient by the small intestine of the rat. Journal of ' Membrane Biology 100, 123-136. Patel, H.M., Stevenson, R.W., Parsons, J.A. and Ryman, B.E. (1982). Use ofliposomes ,;, to aid intestinal absorption of entrapped insulin in normal and diabetic dogs. Biochimica f I Biophysica et Acta 716, 188-193. I \ i .\ Peter, A., Devadder, S., Laus, G. and Tourwe, D. (1996). Liquid chromatography studies ) > on the enzymatic degradation of luteinizing hormone-releasing hormone analogues with I off-line identification by mass spectroscopy. Journal of Chromatography A 729, 137-142. X Peter, R.E., Lin, H. and Van der Kraak:, G. (1988). Induced ovulation and spawning of cultured fresh water fish in China: Advances in application of GnRH analogues and J I :,. dopamine anatagonists. Aquaculture 74, 1-10. ( * Peterson, G.L. (1977). A simplification of the protein assay method of Lowry et al which 't is more generally applicable. Analytical Biochemistry 87, 386-396. .:> Pickford, G.E. and Atz, J.W. (1957). The physiology of the pituitary gland of fishes. New .\ York Zoological Society, New York. 204 Powell, D.W. (1981). Barrier function of epithelia. American Journal ofPhysiology 241, ,1·,,. G275-288. ·> Pozzi, F., Furlani, P., Gazzaniga, A., Davis, S.S. and Wilding, LR. (1994). The TIME CLOCK system: A new oral dosage form for fast and complete release of drug after a predetermined lag time. Journal of Controlled Release 31, 99-108. Raae, A.J. (1990). Effects of low and high temperatures on chymotrypsin from Atlantic -,, > cod (Gadus morhua L.); comparison with bovine a-chymotrypsin. Comparative > -·- ii. Biochemistry and Physiology B 97, 145-149. ~ 4 Raae, A.J. and Walther, B.T. (1989). Purification and characterization of chymotrypsin, ,. trypsin and elastase like proteinases from cod (Gadus morhua L.). Comparative I> Biochemistry and Physiology B 93, 317-324. '; ., ,. Racicot, W.F. and Hultin, H.O. (1987). A comparison of dogfish and bovine chymotrypsins. Archives ofBiochemistry Biophysics 256, 131-143. Ramakrishna, M., Hultin, H.0. and Racicot, W.F. (1987). Some kinetic properties of -,, " dogfish chymotrypsin. Comparative Biochemistry and Physiology B 87, 25-30. I , , i {)- Rao, K.N. and Lombardi, B. (1975). Substrate solubilization for the Hummel a-chymotrypsin assay. Analytical Biochemistry 65, 548-551. / ;r l, ii Roda, A., Minutelo, A., Angello, M.A. and Fini, A. (1990). Bile acid structure-activity ,~ .. relationship: Evaluation of bile acid lipophilicity using 1-octanol/water partition coefficient and reverse phase HPLC. Journal ofLipid Research 31, 1433-1443. ' 205 Rombout, J.H.W., Lamers, C.M.J., Helfrich, M.H., Dekker, A. and Taveme-Thiele, J.J. . , (1985). Uptake and transport of intact macromolecules in the intestinal epithelium of carp (Cyprinus carpio L.) and the possible immunological implications. Cell Tissue Research 239, 519-530. Rombout, J.H.W.M., Bot, H.E. and Taveme-Thiele, J.J. (1989). Immunological importance of the second gut segement of carp. 2. Characterization of mucosa! leucocytes. Journal ofFish Biology 35, 167-178 . .. ' ~ Rombout, J.H.W.M. and van der Berg, A.A. (1989). Immunological importance of the ·,' second gut segment of carp. 1. Uptake and processing of anitgens by epithelial cells and . ~ macrophages. Journal Fish Biology 35, 13-22. 0- ,. Roy, A.D., Campell, R. and Goldberg, D.M. (1967). Effect of diet on the trypsin and chymotrypsin output in the stools of patient with an ileostomy. Gastroenterology 53, .i " 584-589 . '> _). ti,. Rubinstein, A., Tirosh, B., Baluom, M., Nassar, T., David, A., Radai, R., Gliko-Kabir, I. I and Friedman, M. (1997). The rationale of peptide drug delivery to the colon and the ~ potential of polymeric carriers as effective tools. Journal of Controlled Release 46, \ " 59-73. /> ... Russell-Jones, G.J. (1998). Use of vitamin B 12 conjugates to deliver protein drugs by the :,. oral route. Critical Reviews in Therapeutic Drug Carrier Systems 15, 557-586 1, ~ ·' ' Russin, DJ., Floyd, B.F., Toomey, T.P., Brady, A.H. and Awad, W.M. (1974). The . proteolytic enzymes of the K-1 strain of Streptomyces griseus obtained from a • commercial preparation (pronase). Stabilization of the trypsin component by calcium and guanidine. Journal ofBiological Chemistry 249, 6144-6148. 206 Saffran, M., Kumar, G.S., Neckers, D.C., Pena, J., Jones, R.H. and Field, J.B. (1990). Biodegradable axopolymer coating for oral delivery of peptide drugs. Biochemical Society Transactions 18, 7S2-754. i>- ':, I' Saffran, M., Kumar, G.S., Savariar, C., Burnham, J.C., Williams, F. and Neckers, D.C. (1986). A new approach to the oral administration of insulin and other peptide drugs. Science 233, 1081-1084. Sakai, M. Imai, T., Ohtake, H., Azuma, H. Otagiri, M. (1997). Effects of absorption ,> '. enhancers on the transport of model compounds in Caco-2 cell monolayers: assessment by confocal laser scanning microscopy. Journal ofPharmaceutical Sciences 86, 779-785. '""1 ;. Sangadala, S., Walters, F.S., English, L.H. and Adang, M.J. (1994). A mixture of ii- Manduca sexta aminopeptidase and phosphatase enhances Bacillus thuringiensis 86 ';. _insecticidal CryIA(c) toxin binding and Rb+-K+ efflux in vitro Journal ofBiological ,A, t, Chemistry 269, 10088-10092. i> Satow, T., Machida, A., Funakushi, K. and Palmieri, R.L. (1991). Effects of the sample matrix on the separation of peptides by high performance capillary electrophoresis. 7. } Journal ofHigh Resolution Chromatography 14, 276-279. I ~ Sayani, A.P. and Chien, Y.W. (1996). Systemic delivery of peptides and proteins across absorptive mucosae. Critical Reviews in Therapeutic Drug Carrier Systems 13, 85-184. " , > Schechter, I. and Berger, A. (1967). On the size of the active site in proteases. I. Papain. _\., J.' Biochemical and Biophysical Research Communications 27, 157-162. ,. Schellenberger, V., Schelknberger, U., Mitin, Y.V. and Jakubke, H.-D. (1989). Characterization of the S'-subsite specificity of porcine pancreatic elastase. European Journal ofBiochemistry 179, 161-163. 207 Schellenberger, V., Turck, C.W., Hedstrom, L. and Rutter, W.J. (1993). Mapping the S' subsites of serine proteases using acyl transfer to mixtures of peptide nucleophiles. Biochemistry 32, 4349-4353. { \> Schep, L.J. (1999). Delivery ofbioactive peptides across the salmonid posterior intestine. ' ,, PhD thesis, University ofOtago, New Zealand. ., Schep, L.J., Tucker, LG., Young, G. and Butt, A.G. (1997). Regional permeability differences between the proximal and distal portions of the isolated salmonid posterior ' ' ,, intestine. Journal of Comparative Physiology B 167, 370-377. ~-" I> Schep, L.J., Tucker, LG., Young, G., Ledger, R. and Butt, A.G. (1998). Permeability of " the salmon (Oncorhynchus tshawytscha) posterior intestine in vivo to hydrophilic \ markers. Journal of Comparative Physiology B 168, 562-568. Schipper, N.G.M., Vaorum, K.M. and Artursson, P. (1996). Chitosans as absorption ·::. :> enhancers for poorly absorbable drugs. 1: Influence of molecular weight and degree of 7.,. acetylation on drug transport across human intestinal epithelium (Caco-2) cells. i Pharmaceutical Research 13, 1686-1692. ·' > /> Schwert, G.W. and Takenaka, Y. (1955). A spectrophotometric determination of trypsin ..... and chymotrypsin. Biochimica et Biophysica Acta 16, 570-574. > .... (> Seidl, D.S. and Liener, L (1972). Isolation and properties of complexes of Bowman-Birk -'.. soybean inhibitor with trypsin and chymotrypsin. Journal ofBiological Chemistry 247, ,,_ 3533-3538 Shiau, Y. (1981). Mechanisms of intestinal fat absorption. American Journal of Physiology 240, Gl-9. 208 Sire, M.F. and Vernier, J.M. (1992). Intestinal absorption of protein in teleost fish. Comparative Biochemistry Physiology A 103, 771-781. Sleisenger, M.H., Burston, D., Dalrymple, J.A., Wilinson, S. and Mathews, D.M. (1979). Evidence for a single common carrier for uptake of dipeptide and a tripeptide by hamster jejunum in vitro. Gastroenterology 71, 76-81 ~·. Smart, J.D., Kellaway, I.W. and Worthington, H.E.C. (1984). An in-vitro investigations ... of mucosa-adhesive materials for use in controlled drug delivery. Journal ofPharmacy Pharmacology 36, 295-299. ~J- l.;. b,. Smith, RN., Hansch, C. and Ames, M.M. (1975). Selection of a reference .:>, partitioning system for drug design work. Journal ofPharmaceutical Sciences 64, .) 599-606 . ,,. > Smithson, K.W. and Gray, G.M. (1977). Intestinal assimilation oftetrapeptide in the rat. Obligate :function of brush border aminopeptidase. Journal' of Clinical Investigations 60, 665-674. "'f"r Soderquist, M.E. and Walton, A.G. (1980). Structural changes in proteins adsorbed on t> polymer surfaces. Journal of Colloid and Interface Science 75, 386-397. ,. ,. />. ' ... Solar, I.I., McLean, E., Baker, I.J., Sherwood, N.M. and Donaldson, E.M. (1990). Short communication: Induced ovulation of sablefish (Anoplopoma fimbria) following oral i • administration of des Gly10-(D-Ala6)LH-RH ethylamide. Fish Physiology and J> Biochemistry 8, 497-499 . • 209 Sterchi, E.E., Green, J.R. and Lentze, M.J. (1982). Non-pancreatic hydrolysis ofN . r > benzoyl-p-amino benzoic acid (P ABA-peptide) in the human small intestine. Clinical I " Science 62, 557-560. I ~' I> Stroband, H.W.J. and van der Veen, F.H. (1981). Localisation of protein absorption during transport of food in the intestine of the grass carp, Ctenopharyngodon idella (Val). ., Journal ofExperimental Zoology 218, 149-156 . '· -, . Stryer, L. (1988). Biochemistry. W.H. Freeman and Company: New York. p. 226. ,, ', I ' Su, S.-F. and Amidon, G.L. (1995). Investigation into the intestinal metabolism of 1 [D-Ala ] Peptide T amide implication for oral delivery. Biochimica et Biophysica Acta ,;. 1245, 62-68 . .... Suda, H., Aoyagi, T., Takeuchi, T. and Umezawa, H. (1976). Inhibition of ) .1, i aminopeptidase B and leucine aminopeptidase by bestatin and its stereoisomer. I i. i Archives ofBiochemistry and Biophysics 177, 196-200. • > ,,,,,., Sue, T.K., Jaques, L.B. and Yuen, E. (1976). Effects of acidity, cations and alcoholic fractionation on absorption of heparin from gastrointestinal tract. Canadian Journal ( 7 Physiology Pharmacology 54, 613-617. I,,. .. ' Sukumasavin, N., Leelapatra, W., McLean, E. and Donaldson, E.M. (1992). Orally ;. ,",, induced spawning of Thai carp (Puntius gonionotus, Bleeker) following co 10 6 administration of des Gly (D-Arg ) sGnRH ethylamide and domperidone. Journal of ! ' . Fish Biology 40, 477-479. I ~ I 210 Sutcliffe, N. and Corran, P.H. (1993). Comparison of selectivities ofreversed-phase high • performance liquid chromatography, capillary zone electrophoresis and micellar capillary electrophoresis in the separation ofneurophypophyseal peptides and analogues. Journal \-> of Chromatography 636, 95-103. ' Suzuki, Y., Kobayashi, M., Aida, K. and Hanyu, I. (1988a). Transport of physiologically J active salmon gonadotropin into the circulation in goldfish, following oral administration of salmon pituitary extract. Journal of Comparative Physiology B 157, 753-758. -r,., •, ', Suzuki, Y., Kobayashi, M., Nakamura, 0., Aida, K. and Hanyu, I. (1988b ). Induced ovulation of the goldfish by oral administration of salmon pituitary extract. Aquaculture ... ,,, 74, 379-384. ~ Takashima, F. and Hibiya, T. (1995). An atlas of fish histology - normal and pathalogical ~ features, Kodansha Ltd, Tokyo. l ~ '> Takeuchi, H., Yamamoto, H., Niwa, T., Hino, T. and Kawashima, Y. (1996). Enteral ,, absorption of insulin in rats from mucoadhesive chitosan-coated liposomes. ,, 't Pharmaceutical Research 13, 896-901. 'r Taki, Y., Sakane, T., Nadai, T., Sezaki, H., Amidon, G.L., Langguth, P. and :, Yamashita, S. (1995). Gastrointestinal absorption of peptide drug: quantitative evaluation ~ h of the degradation and permeation ofmetkephamid in rat small intestine. Journal of ;,., Pharmacology and Experimental Therapeutics 274, 373-377. ~ \ ~/ Taran, L.D. and Smovdyr, I.N. (1992). Comparative kinetic investigations of primary specificity of bovine and salmon trypsins. Biokhimiya 57, 55-60. _., 211 Teem, M.V. and Philips, S._F. (1972). Perfusion of the hamster jejunum with conjugated " and unconjugated bile acids: inhibition of water absorption and effects on morphology. Gastroenterology 62, 261-267. ~ I Temkin, R.J. and Millan, M. (1986). Gut associated lymphoid tissue (GALT) of the goldfish, Carassius auratus. Journal ofMorphology 190, 9-26. ·' Thamotharan, M., Gomme, J., Zonno, V., Maffia, M., Storelli, C. and Ahearn, G.A. ·t \ (1996). Electronic, proton-coupled, intestinal dipeptide transport in herbivores and •, carnivorous teleosts. American Journal ofPhysiology 270, R939-947. ~,c Thiebaut, F., Tsuruo, T., Hamada, H., Gottesman, M.M., Pastan, I. and Willingham, M.C. I>- (1987). Cellular localization of the multi-drug resistance gene product P-glycoprotein in -~ normal human tissues. Proceedings of the National Academy ofScience ofthe U.S.A 84, 7735-7738 . .l "· ~ Thomas, P. and Boyd, N.M. (1988). Induced spawning of spotted seatrout, red drum and orangemouth corvina (family: Sciaenidae) with luteinizing hormone-releasing hormone ·• 't analog injection. Contributions to Marine Science 30 S, 43-47. ~ Thomas, P. and Boyd, N.W. (1989). Dietary administration of an LHRH analogue '.> induces spawning of spotted seatrout (Cynoscion nebulosus). Aquaculture 80, 363-370. I>- Jc Tidball, C.S. (1964). Magnesium and calcium as regulators of intestinal permeability. American Journal ofPhysiology 206, 243-246. ' ._ Tomita, M., Shiga, M., Hayashi, M. and Awazu, S. (1988). Enhancement of colonic drug > absorption by the paracellular permeation route. Pharmaceutical Research 5, 341-346. 212 Toppmeyer, D.L., Slapak, C.A., Croop, J. and Kufe, D.W. (1994). Role of P-glycoprotein l . in dolastatin 10 resistance. Biochemical Pharmacology 48, 609-612. Torrissen, K.R. (1984). Characterization of proteases in the digestive tract of Atlantic 1 ·1 I ., salmon (Salmo salar) in comparison with rainbow trout (Salmo gairdneri). Comparative ' ' Biochemistry and Physiology B 77, 669-674 . ./. Touitou, E. and Rubinstein, A. (1986). Targeted entemal delivery of insulin to rats . ., " .. International Journal ofPharmaceutics 30, 95-99 . Towns, J.K. and Regnier, F.E. (1992). Capillary electrophoretic separations of -., proteins using nonionic surfactant coatings. Analytical Chemistry 64, 2473-2476. ,,.. 1 I ~ Tozaki, H., Komoike, J., Tada, C., Maruyama, T., Terabe, A., Suzuki, T., Yamamoto, A. J and Muranishi, S. (1997). Chitosan capsules for colon-specific drug delivery: ) * Improvement of insulin absorption from the rat colon. Journal ofPharmaceutical I ~ I Sciences 86, 1016-1021. ' " Tsai, T., Mehta, R.C. and DeLuca, P.P. (1996). Adsorption of peptides to poly(D,L t ~ lactide-co-glycolide): 2. Effect of solution properties on the adsorption. International T Journal ofPharmaceutics 127, 43-52. .:> 7 ' Tzannis, S.T., Hrushesky, W.J.M., Wood, P.A. and Przybycien, T.M. (1997). Adsorption > of a formulated protein on a drug delivery device surface. Journal of Colloid and ~ >- Interface Science 189, 216-228. ~ .. ,. Uchida, N., Tsukayama, K. and Nishide, E. (1984a). Properties of two main anionic ;/ trypsins from the plyoric caeca of chum salmon. Bulletin ofthe Japanese Society of Scientific Fisheries 50, 313-322. 213 Uchida, N., Tsukayama, K. and Nishide, E. (1984b). Purification and some properties of ' ... trypsins from pyloric caeca of chum salmon. Bulletin ofthe Japanese Society ofScientific \ '~ ,I Fisheries 50, 129-138. ' > I I I I Umezawa, H., Aoyagi, T., Okura, A., Morishima, H., Takeuchi, T. and Okami, Y. (1973). ~ Elastinal, a new elastase inhibitor produced by Actinomycetes. Journal ofAntibiotics 26, ,.. 787-789. ~ ; s, ~-., Van Der Kraak, G., Lin, R.R., Donaldson, E.M., Dye, H.M. and Hunter, G.A. (1983). I ~ 10 6 Effects ofLH-RH and des-Gly [D-Ala ] LH-RH-ethylamide on plasma gonadotropin t. levels and oocyte maturation in adult female coho salmon (Oncorhynchus kisutch). > General and Comparative Endocrinology 49, 470-476. ., .l ~- van Hoogdalem, E.J., Heijligers-Feijen, C.D., de Boer, A.G., Verhoef, J.C. and Breimer, i D.D. (1989). Rectal absorption enhancement of des-enkephalin-y-endorphin (DEyE) by medium-chain glycerides and EDTA in conscious rats. Pharmaceutical Research 6, '· \ 91-95. 't :, Varallyay, E., Pal, G., Patthy, A., Sziagyi, L. and Graf, L. (1998). Two mutations in rat 7 i.,., .. trypsin confer resistance against autolysis. Biochemical and Biophysical Research 243, ·( Communications 56-69. :~ Varesio, L., Landolf, S., Giovarelli, M. and Forni, G. (1980). The macrophages as the social interconnection within the immunesystem. Development of Comparative ;. Immunology 4, 11-19. 214 Verri, T.M., Maffia, M. and Storelli, C. (1992). H+/glycyl-glycine cotransport in eel . intestinal brush-border membrane vesicles: studies with pH-sensitive dye acridine orange . Biochimica et Biophysica Acta 110, 123-126. Vithayathil, A.J., Buck, F., Bier, F. and Nord, F.F. (1961). On the mechanism of enzyme -- ' action. LXXII. Comparative studies of trypsins of various origins. Archives of ,,_ Biochemistry and Biophysics 92, 532-540. Walde, P., Sunamoto, J. and O'Connor, C.J. (1987). The mechanism ofliposomal damage '"" >- by taurocholate. Biochimica et Biophysics Acta 905, 30-38. • >" Walker, W.A. (1981). Intestinal transport of macromolecules, In: Physiology ofthe " Gastrointestinal tract, L.R. Johnson, Raven Press: New York. p. 1271. Walsh, K.A. (1970). Trypsinogens and trypsins of various species. Methods in 'j, Enzymology 19, 41-63. l "' Watanabe, Y. (1982). Intracellular digestion of horseradish peroxidase by intestinal cells of teleost larve and juveniles. Bulletin ofthe Japanese Society ofScientific Fisheries 48, ·- ~ ~ 37-42. 't Watzig, H. (1995). Appropriate calibration functions for capillary electrophoresis I. Precision and sensitivity using peak areas and heights. Journal of Chromatography A t> ;. 700, 1-7. > Weinberger, R., Sapp, E. and Moring, S. (1990). Capillary electrophoresis of urinary porphyrins with absorbance and fluorescence detection. Journal of Chromatography 516, 271-285. 215 Western, J.R.H. (1971). Feeding and digestion in two cottid fishes, freshwater Cottus - ~; gobio L. and the marine Enophrys bubalis (Euphrasen). Journal ofFish Biology 3, 225-246. -i.'3,_ .j ,, 'i Wilding, LR., Davis, S.S., Pozzi, F., Furlani, P. and Gazzaniga, A. (1994). Enteric coated '.. time released systems for colonic targeting. International Journal ofPharmaceutics 111, ,,, 99-102. Wirt, R. (1974). Methods ofEnzymatic analysis. Academic Press Inc., New York. y ,,. p. 1009-1012. .\ f tr' "' Woodley, J.F. (1994). Enzymatic barriers for GI peptide and protein delivery. Critical Reviews in Drug Therapy 11, 61-95. ~ .:!, Yagi, T., Ishizaki, K. and Takebe, H. (1980). Cytotoxic effects of protease inhibitors on l human cells. 2. Effect of elastinal. Cancer Letters 10, 301-307. L. -', Yamada, K., Murakami, M., Yamaoto, A., Takada, K. and Muranishi, S. (1992). ~ Improvement of intestinal absorption of thyrotropin-releasing hormone by chemical ~ modified with lauric acid. Journal ofPharmacy and Pharmacology 44, 717-721. ) Yamamoto, A., Hayakawa, E. and Lee, V.H.L. (1990). Insulin and proinsulin proteolysis I> .\. in mucosal homogenates of the albino rabbit: implications in peptide delivery from \\ nonoral routes. Life Science 47, 2465-2474. Yamamoto, A., Taniguchi, T., Rikyuu, K., Tsuji, T., Fujita, T., Murakami, M. and Murakami, S. (1994). Effect of various protease inhibitors on the intestinal absorption and degradation of insulin in rat. Pharmaceutical Research 11, 1496-1500. . 216 Yasutake, W.T. and Wales, J.H. (1983). Microscopic anatomy of Salmonids: An atlas. US Department of the Interior, Washington. Young, B.R., Pitt, W.G. and Cooper, S.L. (1988). Protein adsorption on polymeric . '! biomaterials. II Adsorption kinetics. Journal of Colloid and Interface Science 125, 246- \ 260. '} ,., Yoshinaka, R., Sato, M. and Ikeda, S. (1981). Distribution of trypsin and chymotrypsin -,- and their zymogens in the digestive system of catfish. Bulletin of the Japanese Society of . ,,_ Scientific Fisheries 47, 1615-1618 . ... · Yoshinaka, R., Sato, M., Sato, T. and Ikeda, S. (1984a). Enzymatic characterization of ,,. ... anionic trypsin of the catfish (Parasilurus asotus). Comparative Biochemistry and PhysiologyB 77, 1-6. > :! Yoshinaka, R., Sato, M., Tanaka, H. and Ilceda, S. (1984b). Purification and .\. characterization of a new metalloproteinase with elastolytic activity from the catfish 1, pancreas. Biochimica et Biophysica Acta 798, 240-246 . . ~ ., Yoshinaka, R., Sato, M., Tanaka, H. and Ikeda, S. (1985). Distribution of pancreatic ' elastase and metalloproteinase in several species of fish. Comparative Biochemistry and -~ Physiology B 80, 227-234. ' . C> ..\.. Ziv, E., Eldor, A., Kleinman, Y., Bar-On, H. and Kibron, M. (1983). Bile salts facilitate the absorption of heparin from the intestine. Biochemical Pharmacology 32, 773-776. -•, ~ Ziv, E., Kidron, M., Berry, E.M. and Bar-On, H. (1981). Bile salts promote the absorption of insulin from the rat colon. Life Sciences 29, 803-809. ( 217 Ziv, E., Lior, 0. and Kidron, M. (1987). Absorption of protein via the intestinal wall. A quantitative model. Biochemical Pharmacology 36, 1035-1039. , j Zohar, Y., Tosky, M., Pagelson, G. and Finkelman, Y. (1989). Induction of spawning in j the gilthead seabream, Sparus aurata, using [D-Ala6-Pro9NETJ-LHRH: comparison with I ) \ the use ofhCG. Israel Journal ofAquaculture 4, 105-113. !', .'; ·.·'K. ':'.J,.. !j " ,. /.-- J \'l. l, ~ t I, i I_} ,1 -1:::-r, !': 1 > I ii_' .. I i • ~