- Email: [email protected]
Pharmacokinetics of trefoil peptides and their stability in gastrointestinal contents
peptides 28 (2007) 1197–1206
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/peptides
Pharmacokinetics of trefoil peptides and their stability in gastrointestinal contents§ Stine Kjellev a,*, Else Marie Vestergaard b, Ebba Nexø b, Peter Thygesen c, Maria S. Eghøj d, Palle B. Jeppesen e, Lars Thim f, Nis Borbye Pedersen g, Steen Seier Poulsen d a
Immunopharmacology, Novo Nordisk a/s, DK-2760 Maaloev, Denmark Department of Clinical Biochemistry, Aarhus University Hospital, DK-8000 Aarhus, Denmark c Exploratory ADME, Biopharmaceuticals, Novo Nordisk a/s, DK-2760 Maaloev, Denmark d Department of Medical Anatomy, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark e Department of Medicine CA-2121, Division of Gastroenterology, Rigshospitalet, University of Copenhagen, DK-2100 Copenhagen, Denmark f Protein Engineering, Novo Nordisk a/s, DK-2760 Maaloev, Denmark g Department of Medical Physiology, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark b
article info
abstract
Article history:
Trefoil factor family (TFF) peptides are considered promising for therapeutic use in gastro-
Received 18 October 2006
intestinal diseases, and there is a need to explore the fate of injected TFF and the stability of
Received in revised form
the peptides in the gastrointestinal tract. We studied the pharmacokinetics of intravenously
22 March 2007
(i.v.) administered hTFF2 in mice and rats and of hTFF3 administered i.v., intramuscularly,
Accepted 22 March 2007
intraperitoneally, and subcutaneously in mice, and estimated by ELISA the decay of the
Published on line 31 March 2007
peptides added to rat and human gastrointestinal contents. We found that i.v. injected hTFF2 and hTFF3 were cleared from the circulation within 2–3 h, exhibiting comparable
Keywords:
pharmacokinetic profiles. In contents from the rat stomach, hTFF levels remained
Trefoil factor family
unchanged for up to 6 days. In the small and large intestine of rats, the hTFF levels decreased
TFF2
markedly after 4 and 1 h, respectively. In small intestinal contents from humans, the levels
TFF3
remained stable for more than 24 h. We conclude that systemically administered hTFF2 and
Pharmacokinetics
hTFF3 are rapidly eliminated from the circulation and that the stability of hTFF2 and hTFF3
Clearance
in GI contents appeared higher in the gastric and small intestinal milieu than in the large
Stability
intestine and feces, suggesting a higher stability toward gastric acid and digestive enzymes than toward microbial degradation. # 2007 Elsevier Inc. All rights reserved.
1.
Introduction
The trefoil factor family (TFF) 1–3 comprises a group of small peptides (7–12 kDa) sharing a common structure termed the
§
trefoil domain. TFF2 contains two trefoil domains due to genomic duplication of the motif, while the two domains in TFF1 and TFF3 are linked by disulfide bonds via a seventh cysteine residue (reviewed in [18]). The peptides are produced
SK, PT, and LT are employed at Novo Nordisk a/s. * Corresponding author at: Novo Nordisk Park F6 2.30, DK-2760 Maaloev, Denmark. Tel.: +45 44438062; fax: +45 44434537. E-mail addresses: [email protected] (S. Kjellev), [email protected] (E.M. Vestergaard), [email protected] (E. Nexø), [email protected] (P. Thygesen), [email protected] (M.S. Eghøj), [email protected] (P.B. Jeppesen), [email protected] (L. Thim), [email protected] (N.B. Pedersen), [email protected] (S.S. Poulsen). Abbreviations: ELISA, enzyme-linked immunosorbant assay; GI, gastrointestinal; HPLC, high-performance liquid chromatography; ID, injected dose; i.m., intramuscular; i.p., intraperitoneal; i.v., intravenous; %ID, percentage of injected dose; PK, pharmacokinetic; s.c., subcutaneous; TFF, trefoil factor family 0196-9781/$ – see front matter # 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2007.03.016
1198
peptides 28 (2007) 1197–1206
AUC area under the curve AUC%Extrapol extrapolated percentage of the molecule left in serum after last measured time point start concentration in serum C0 maximal serum concentration Cmax CL clearance F bioavailability MRT mean residence time half-life of molecule in serum T1/2 time to maximal serum concentration Tmax Vz volume of distribution volume of distribution at steady state Vss lz terminal elimination rate constant
by mucus-secreting cells, especially in the gastrointestinal (GI) tract in a site-specific manner for each peptide, i.e., TFF1 is mainly produced by gastric surface epithelial cells, TFF2 by gastric mucous neck cells and pyloric glands, and in the duodenal Brunner’s glands, whereas TFF3 is mainly produced by the intestinal goblet cells and the antral mucous cells [6]. Upregulation and ectopic expression of the peptides have been reported in GI tissues from patients with chronic GI inflammation [24], and all three peptides have shown a therapeutic effect in experimental models of GI disease [1,5,12–14]. Earlier in vitro studies reported that TFF2 was extremely resistant to protease and acid, based on the findings that TFF2 incubated with HCl or trypsin was only slightly degraded [7]. Likewise, 125 I-hTFF2 seemed stable following incubation in small intestinal or gastric juice or in HCl- and GI protease-containing solutions [12]. Our previous studies using 125I-labeled TFFs administered intravenously (i.v.) to rats suggested that the peptides were rather slowly cleared from the circulation and that TFF2 was very resistant toward degradation in vivo [16]. These findings in concert with the compact structure of the TFF peptides have led to the assumption that all TFFs were generally highly resistant toward degradation [12,21]. Pharmacological effects of TFFs have been obtained after systemic as well as local administration, and the aim of this study was therefore to study in more detail the distribution and elimination in vivo by examining the pharmacokinetics of human TFF2 and TFF3 (hTFF2 and hTFF3) in mice and rats after systemic administration and to study the stability of TFFs added to different types of rat GI contents (to mimic the conditions following local administration) using enzymelinked immunosorbant assays (ELISA) specifically developed for hTFF2 or hTFF3. Finally, this paper presents the results of studies on the levels of hTFF2 and hTFF3 in human bowel contents from different parts of the gut, and the effect of storage at 37 8C on the concentration of these peptides.
2.
Materials and methods
2.1.
Animals
Female BALB/cAnNTac mice (20–23 g) and female Wistar Hannover GALAS Rats (HanTac:WH) (250–300 g) (Taconic
M&B, Ry, Denmark) were housed at the Panum Institute (University of Copenhagen, Denmark) or Novo Nordisk a/s. The animal studies were approved by the Danish National Committee for Animal Studies (i.e., the Animal Experiments Inspectorate).
2.2.
Subjects
From humans, samples of normal stool (n = 5) were obtained as well as wet ostomy samples from jejunostomies (n = 8), ileostomies (n = 5), and colostomies (n = 5). None of the subjects had active disease at the time of sampling. The samples were collected after lunch (the subjects were not fasted), were frozen immediately after collection, and stored at 20 8C until analysis. The local scientific ethics committee approved all procedures. All participants gave informed consent.
2.3.
Peptides
hTFF2 and dimeric hTFF3 were produced as described previously [20,21]. For the study of the molecular forms of TFF (Section 2.7), both peptides were labeled at Novo Nordisk a/s with sodium 125iodide (Na125I) using lactoperoxidase to a radiochemical purity of 98%. The specific radioactivity was 2.2 mCi/pmol (hTFF3) and 0.12 mCi/pmol (hTFF2). The tracers were diluted with 0.9% saline to obtain the final concentrations for injection (Section 2.7).
2.4. rats
Pharmacokinetics of hTFF2 and hTFF3 in mice and
hTFF2 or hTFF3 was administered to mice and rats i.v., intraperitoneally (i.p.), intramuscularly (i.m.), or subcutaneously (s.c.) according to Table 1. Prior to the i.v. injection, the mice were anesthetized with a mixture of fentanyl (0.15 mg/kg), droperidol (9.8 mg/kg), and midazolam (1 mg/ kg) administered s.c. The rats were anesthetized with methohexital (50 mg/kg) i.p. (buprenorfin 0.1 mg/kg s.c. for analgesia). The i.v. injection into the inferior vena cava was performed via a midline incision, which was closed with 3–0 silk sutures following injection. For collection of urine, the urethra was ligated when the hTFF was administered. The animals were sacrificed in groups of three to five at the time points specified in Table 1. Prior to sacrifice, blood and urine sampling was performed under general anesthesia. Blood was collected from the orbital venous plexus and urine was directly aspirated from the bladder. The blood was allowed to
Table 1 – Dosing regimens for pharmacokinetic analysis Species Peptide
Route
Dose (mg/kg)
Mouse Mouse Mouse Mouse Rat
i.m. i.p., s.c. i.v. i.v. i.v.
1 1, 5 5 5 5
a
TFF3 TFF3 TFF3 TFF2 TFF2
Simultaneous collection of urine.
Time (min) 6, 15, 30, 60, 120, 180 15, 30, 60, 180 2, 6, 15, 45a, 120a, 240 2, 6, 15, 45a, 120a, 240a 2, 6, 15, 45a, 120a, 240a
peptides 28 (2007) 1197–1206
stand for 1 h prior to centrifugation and collection of serum. Urine was frozen immediately after sampling. All samples were stored at 20 8C.
2.5. Stability of hTFF2 and hTFF3 in gastrointestinal luminal contents We measured the decay of endogenous and exogenously added hTFFs in GI contents to assess in which parts of the GI tract a therapeutic dose of TFFs can be expected to be present following oral treatment, and which dosing frequency would be appropriate. Instead of administering the hTFFs orally and performing PK analysis on the GI contents, we chose a simpler setup, where hTFFs were added directly to contents from various parts of the GI tract. In this way, we could separate the actual hTFF degradation process from influence on the hTFF levels due to different transit times in the various parts of the gut, possible enteral absorption and secretion of hTFFs. The human samples were diluted 2- or 2.5-fold in 0.9% NaCl. Jejunostomy, ileostomy, colostomy, and fecal samples were spiked with hTFF2 or hTFF3 suspended in 0.9% NaCl in sufficient amounts to increase the concentration of total hTFF2 and hTFF3 by 180 and 1900 nM, respectively. Endogenous TFF was measured in identical samples without added TFF. Furthermore, a series of samples containing only 180 nM hTFF2 or 1900 nM hTFF3 in pure 0.9% NaCl was run in parallel. Samples were incubated at 37 8C for 0, 1, 2, 4, 24, 72, and 120 h. Boiling the samples for approximately 1 min stopped incubation. (Termination of incubation by this process was confirmed by control experiments. Furthermore, the ELISAs have similar sensitivity for boiled and unboiled hTFFs (data not shown).) The samples were kept at 20 8C until analysis. Prior to analysis, the suspensions were centrifuged and the supernatants used for analysis. The stabilities of hTFF2 and hTFF3 added in vitro were also measured in contents from the rat stomach, small intestine, cecum, and colon in a similar fashion (samples were pooled from three rats, and an additional 144-h time point was included). The rats were not fasted. hTFF2 and hTFF3 were added to rat bowel discharges at a concentration of 250 nM.
2.6.
hTFF2 and hTFF3 ELISA
The concentrations of hTFF2 and hTFF3 in serum, urine and GI contents were measured by in-house peptide-specific ELISAs as described [22,23]. Briefly, the detection limits for the hTFF2 and hTFF3 ELISAS were 6 and 3 pmol/L, respectively, and the analytical imprecision 7% for mean concentrations of 17– 77 pmol/L (hTFF2) and 13–65 pmol/L (hTFF3), respectively. No cross-reactivity was observed for human trefoil peptides, as 320 nmol/L (TFF1 and TFF3) and 40 nmol/L (TFF1 and TFF2) were below the detection limits for the TFF2 and TFF3 ELISAs, respectively. Western blot analysis of the capture antibody used in the TFF3-ELISA showed that there were no crossreactions with recombinant mouse TFF3 dimer (data not shown). Negligible levels of TFF2 and TFF3 were detected in plasma samples from mouse and rat that were not injected with the human peptides (shown for rats in [9]). The percentage of recovery in the spiked samples of GI contents from humans was determined by comparing the amount of
1199
added hTFF2 and hTFF3 with the amount measured after subtracting the endogenous hTFF2 and hTFF3 concentrations.
2.7. Molecular forms of injected hTFF2 and hTFF3 recovered in serum High-performance liquid chromatography (HPLC), gamma counting, and ELISA were performed on serum samples from animals injected s.c. with a mixture of 125I-hTFF2 (1,160,000 cpm/mouse (15 min) or 5,800,000 cpm/mouse (180 min)) and unlabeled hTFF2 (1 mg/kg) or with a mixture of 125I-hTFF3 (300,000 cpm/mouse) and unlabeled hTFF3 (1 mg/ kg). The blood was collected 15 or 180 min after injection of the tracer, and serum was prepared. The radioactivity in the serum samples was measured by a gamma counter (Wallac Wizard 1740, Perkin-Elmer, Boston, MA, USA) and the concentration of hTFF2 or hTFF3 measured by ELISA as described above (n = 4 per time point). We then calculated the percentage of the injected dose (%ID) of tracer detected in the blood by both methods. For HPLC a volume of 100–200 mL of each serum sample (two to three animals per time point) was injected onto a Vydac 214 TP54 column (250 mm 4.6 mm, Hesperia, CA, USA) equilibrated at room temperature at a flow rate of 1 mL/min with acetonitrile/water/trifluoroacetic acid (10.0:89.9:0.1, w/w/w). After application of the sample, the concentration of acetonitrile in the eluting solvent was raised to 50% (v/v) over 20 min (125I-hTFF3) or 90% (v/v) over 40 min (125I-hTFF2) using a linear gradient. Fractions of 1 mL were collected and subjected to gamma counting, and background radioactivity was subtracted to obtain the actual radioactivity of the samples. In this system, hTFF2 eluted after 20–21 min, hTFF3 dimer eluted after 15 min, hTFF3 monomer eluted after 19 min, free 125I, 125I-tyrosine, or 125I-tyrosine in small peptide fragments (1–10 amino acids) eluted after approximately 4 min.
2.8.
Size-exclusion chromatography
Fresh human stool samples not spiked with TFF were subjected to size-exclusion chromatography on a Superdex 75 HR 10/30 column using a SMART system (Amersham Pharmacia Biotech, GE Healthcare Europe, Munich, Germany). The column was pre-equilibrated and eluted with 0.05 M Tris– HCl, pH 7.2, 9.0 g/L NaCl at 0.5 mL/min with collection of 0.5mL fractions. The column was calibrated with molecular mass calibrators from Amersham Pharmacia Biotech, eluting at fractions 12 (67 kDa), 14 (43 kDa), 18 (25 kDa), and 20 (13.7 kDa). Purified hTFF3 monomer and hTFF3 dimer were used as controls. The distribution of hTFF2 and hTFF3 immunoreactivity was determined by ELISA.
2.9.
Statistical analysis
The pharmacokinetic data were analyzed by a non-compartmental analysis using WinNonlin Professional (Pharsight Inc., Mountain View, CA, USA). Calculations were performed using the mean concentration–time values from three to five animals at each time point. For data analysis of TFF contents in bowel discharge, we used the Prism1 version 4 (GraphPad, San Diego, CA, USA). Differences between the levels of
1200
peptides 28 (2007) 1197–1206
endogenous hTFFs in stomy and fecal samples were analyzed using Student’s t-test with Welsch’s correction for unequal variances.
3.
Results
3.1.
Pharmacokinetics of i.v. injected hTFF2
The pharmacokinetics of hTFF2 in mice and rats after i.v. injection are shown in Fig. 1A and B and Table 2. In both species, an initial short distribution phase was identified, followed by a first-order elimination phase (Fig. 1A). The distribution was completed faster in mice than rats (approximately 15 min vs. 30–45 min; Fig. 1B), but the impact on the serum concentration–time profile was greater in rats compared to mice indicating a more pronounced distribution and redistribution of hTFF2 in rats. The terminal elimination phase was similar in mice and rats as indicated by the two parallel serum concentration–time lines (Fig. 1B). AUC%Extrapol was less than 10% indicating almost complete clearance of from the blood by the end of the study period. The maximal mean serum concentrations observed were 3190 and 2680 nM in mice and rats, respectively (Table 2), while the exposure of hTFF2, expressed as dose-corrected area under the curve, was approximately three-fold higher in mice compared to rats (3.8 vs. 1.3 h kg/L) due to a three-fold
higher clearance in rats (13 vs. 4.4 mL/min kg). The terminal elimination half-life was similar in the two species, 48 and 50 min, respectively. The volume of distribution was threefold higher in rats compared to mice (920 vs. 300 mL/kg). When compared to average plasma volumes of 50 and 31 mL/ kg in mice and rats, respectively [4], a more extensive distribution outside the circulation or specific cellular binding of hTFF2 was observed in rats compared to mice. The mean residence time was 52 and 12 min in mice and rats, respectively, indicating that hTFF2 left the circulation much faster in rats than mice.
3.2. Pharmacokinetics of systemically administered hTFF3 in mice The pharmacokinetics of hTFF3 in mice after different routes of administration (i.v., s.c., i.p., and i.m.) are shown in Fig. 1C and D and Table 3 (AUC%Extrapol was less than 10%). The serum concentration–time profile of i.v. injected TFF3 was quite similar to that of TFF2 with an initial short distribution phase followed by a first-order elimination phase. Serum concentration–time profiles after i.p., i.m., and s.c. administration all had an absorption phase and a terminal elimination phase similar to that observed after i.v. administration. No distribution phases could be identified, most likely due to interference between the absorption and distribution processes. The pharmacokinetic parameter
Fig. 1 – The mean pharmacokinetic profiles of hTFF2 and hTFF3 injected to mice and rats. Mice (three to four per time point) were injected i.v., i.m., s.c., or i.p. with hTFF3, and rats and mice were injected i.v. with hTFF2 (three to five per time point) according to Table 1. The serum levels of the injected peptides at different time points after injection were measured by ELISA. The pharmacokinetic profiles were analyzed by a non-compartmental analysis. (A) Mean serum concentration–time profiles of hTFF2 injected i.v. in rats and mice. Note the short initial distribution phase, and the first-order elimination phase. (B) Mean semi-logarithmic serum concentration–time profiles of hTFF2 injected i.v. to rats and mice. The parallel lines of the elimination phase indicate a similar elimination phase in rats and mice. (C) Mean dose normalized serum concentration–time profiles of hTFF3 in mice after i.v., i.m., i.p., and s.c. administration (insert, same experiment but without i.v. data). Note the faster absorption following i.m. injection. (D) The mean (semi-logarithmic) dose normalized serum concentration–time profiles of hTFF3 in mice after i.v., i.m., i.p., and s.c. administration suggest comparable elimination phases (the parallel lines) regardless of the route of administration.
1201
peptides 28 (2007) 1197–1206
Table 2 – Pharmacokinetic parameter estimates of TFF2 in rats and mice after i.v. injection (mean values) Animal Mouse Rat
Dose (nmol/kg)
Tmax (min)
Cmax (nM)
C0 (nM)
AUC (h nM)
AUC/Dose (h kg/L)
lz (min 1)
T1/2 (min)
CL (mL/min kg)
Vz (mL/kg)
Vss (mL/kg)
332 307
2 2
3190 2680
4380 4010
1260 397
3.8 1.3
0.0145 0.0140
48 50
4 13
300 920
220 150
MRT (min) 52 12
Non-compartmental analysis. Tmax, time to maximal serum concentration; Cmax, maximal serum concentration; C0, starting concentration in serum (extrapolated); AUC, area under the curve; AUC%Extrapol, extrapolated percentage of the molecule left in serum after last measured time point; lz, elimination; T1/2, half-life; CL, clearance; Vz, volume of distribution; Vss, volume of distribution at steady state; MRT, mean residence time.
estimates of hTFF3 in mice are presented in Table 3. The dose-corrected maximal mean serum concentrations (Cmax) were 410, 110, 35, and 34 nM for i.v., i.m., i.p., and s.c administrations, respectively, indicating that i.m. administration results in the highest Cmax/dose when hTFF3 is given extravascularly. The time to maximal serum concentration was also shortest after i.m. administration compared to i.p. or s.c. administration (15 vs. 30 min; Fig. 1C, insert). The exposure in terms of the dose-corrected area under the curve after extravascular administration was 4400, 2300, and 3100 min/L for i.m., i.p., and s.c. administrations, respectively, resulting in absolute bioavailabilities of 30, 16, and 21% after i.m., i.p., and s.c. administrations, respectively. The terminal half-life was similar among all routes of administration and varied between 30 and 44 min. This was also the case for the body clearance and volume of distribution when corrected for bioavailability, indicating similar elimination processes and tissue distribution or cellular binding when the hTFF3 molecules reached the circulation. The mean residue times were also similar except for s.c. administration, where this value indicated a slightly slower absorption phase.
3.3. hTFF2 and hTFF3 levels in the urine of rats and mice following i.v. injection Forty-five minutes after i.v. injection of hTFF2 to mice, 9 (6, 13)% (mean (min, max)) of the injected dose (ID) was detected in the urine. After 2 h, this had increased to 13 (9, 18)% ID and after 4 h to 30 (21, 40)% ID. In rats, the corresponding values were 49 (41, 54)% after 45 min, 49 (43,56)% after 120 min, and 28% (n = 1) after 240 min. Virtually no hTFF3 was excreted in
the urine: The %ID was 0.003 (0.001, 0.005)% after 45 min and 0.01 (0.000, 0.002)% 2 h after injection.
3.4. Molecular forms of injected hTFF2 and hTFF3 recovered in mouse serum The %ID of the 125I-hTFF2/hTFF2 mixture in the serum 15 min after injection was 11.4 1.4 (S.D.) when detected by gamma counting and 15.6 7.9 (S.D.) when measured by ELISA (not significantly different). After 180 min, the corresponding values were 3.0 0.5% and 0.33 0.13%, respectively. HPLC of the serum samples showed that 15 min after injection, all detected 125I-hTFF2 eluted at a position corresponding to the tracer (intact peptide; Fig. 2A). One hundred and eighty minutes after injection 40% of the total radioactivity was detected at a position corresponding to free 125I or 125Ityrosine in small peptide fragments (degraded peptide), and 23% at a position corresponding to the tracer. About 15% of the radioactivity eluted in the following two fractions, i.e., after 22–23 min (Fig. 2B). The %ID of the 125I-hTFF3/hTFF3 mixture in the serum 15 min after injection was 5.9 0.7 (S.D.) when detected by gamma counting and 5.5 0.7 (S.D.) when measured by ELISA. After 180 min, the corresponding values were 8.4 1.3 % and 0.1 0.1 %, respectively. HPLC of the serum samples taken 15 min after s.c. injection showed that most of the injected 125I-hTFF3 eluted at a position corresponding to the dimeric form, whereas 5–10% in sample 1 and 20–25% in sample 2 eluted in a position corresponding to monomeric TFF3 (Fig. 2C). HPLC of the serum samples taken after 180 min showed that practically all injected TFF3 eluted in a position corresponding to degraded peptide (Fig. 2D).
Table 3 – Pharmacokinetic parameter estimates of TFF2 and TFF3 in mice after i.v., i.m., i.p., and s.c. administration (mean values) Peptide
RoA
TFF2 TFF3 TFF3 TFF3 TFF3
i.v. i.v. i.m. i.p. s.c.
Dose Tmax Cmax (nmol) (min) (nM) 8.3 9.3 1.6 1.6 6.5
2 2 15 30 30
3190 3830 169 55 225
C0 (nM) 4380 4690
Cmax/ T1/2 CL Vz Vss MRT F AUC AUC/dose lz dose (min nM) (min/L) (min 1) (min) (mL/min) (mL) (mL) (min) (%) 380 410 110 35 34
75,900 136,000 6,830 3,570 20,300
9,100 14,700 4,400 2,300 3,100
0.0145 0.0158 0.0193 0.0236 0.0159
48 44 36 30 44
0.1 0.1 0.2 0.4 0.3
8 4 12 19 20
5.7 4.1
52 60 53 58 76
30 16 21
Non-compartmental analysis. RoA, route of administration; Tmax, time to maximal serum concentration; Cmax, maximal serum concentration; C0, start concentration in serum (extrapolated); AUC, area under the curve; AUC%Extrapol, extrapolated percentage of the molecule left in serum after last measured time point; lz, elimination; T1/2, half-life; CL, clearance; Vz, volume of distribution; Vss, volume of distribution at steady state; MRT, mean residence time; F, bioavailability.
1202
peptides 28 (2007) 1197–1206
Fig. 2 – HPLC profiles of systemically administered 125I-hTFF2 or 125I-hTFF3. HPLC was performed on serum samples from mice injected s.c. with a mixture of 125I-hTFF2/unlabeled hTFF2 (A and B) or 125I-hTFF3/unlabelled (C and D) at 15 min (A and C) and 180 min (B and D) prior to blood sampling. Intact hTFF2 elutes after 20–21 min, hTFF3 dimer after 15 min, hTFF3 monomer after 19 min and free 125I, 125I-tyrosine, and other small 125I-labeled peptide fragments after 1–4 min.
3.5. Concentrations and stability of endogenous hTFF2 and hTFF3 in intestinal discharges and in stool The concentrations of endogenous hTFF2 and hTFF3 in the contents collected from human jejunostomies, ileostomies, colostomies, and stool are shown in Table 4. The highest concentrations were found for hTFF2, and these were from 2000 to 10,000 times higher than the concentrations found in human serum. For hTFF3, the concentrations were from 140 to 500 times the serum concentrations. Using size exclusion chromatography, we found, for hTFF2 in jejunostomy samples, a sharp peak corresponding to a molecular mass of approximately 25 kDa, whereas in the samples from the ileostomy and the colostomy, the peak was broader (Fig. 3A). Since un-glycosylated hTFF2 has a molecular weight of 12 kDa and glycosylated hTFF2 weighs around 20 kDa, this suggests
Table 4 – Mean (S.D.) concentrations of endogenous hTFF2 and hTFF3 in human small bowel discharges and in stool hTFF2 (nM) Jejunostomy Ileostomy colostomy Normal stool Normal serum a
736 (541) 462 (443) 147 (86)* 241 (182) 0.076a
*
TFF3 (nM) 64 (61) 42 (33) 70 (52) 19 (10) 0.140b
Ref. [22]. Ref. [23]. n = 300 and 296 (serum, median values) and four to eight (bowel discharges). * P < 0.05 (jejunostomy vs. colostomy). b
that the hTFF2 in the samples is mainly present in the glycosylated form, as described previously for the stomach [11]. For hTFF3, a single major peak at 6.6 kDa, corresponding to the reported molecular mass of monomeric hTFF3 was detected (Fig. 3B). We measured the levels of endogenous hTFF2 and hTFF3 in these samples after incubation at 37 8C, as shown in Fig. 3C and D. In the jejunal contents, the hTFF2 and hTFF3 levels were essentially unchanged after 24 h of incubation, and 50% of the initial amount could be retrieved after 120 h. In ileum, 66% of the initial hTFF2 and 56% of the initial hTFF3 were detectable after 120 h. In the colonic contents, the levels of both peptides declined gradually (although higher concentrations of hTFF3 were measured at 4 and 8 h). After 24 h, 50% of the initial amount of hTFF3 and 35% of hTFF2 was detectable. In the fecal samples, the levels of hTFF2 and hTFF3 were low but stable for at least 24 h after, when the levels declined markedly (for unknown reasons, the TFF levels in these samples were lower than in the remaining fecal samples).
3.6. Stability of added hTFF2 and hTFF3 in rat and human GI contents 3.6.1.
Human bowel discharges
We then assessed the stability of hTFF2 and hTFF3 dimer added in vitro to the bowel discharges in order to asses how long time locally administered TFFs would be potentially active in the different parts of the intestinal tract. The added hTFFs were generally difficult to detect in the samples: In the samples without incubation at 37 8C, 49% (range 21–67%) of added hTFF2 and 25% (9–48%) of added hTFF3 could be recovered. In contrast, the recoveries of hTFF2 and hTFF3 in
peptides 28 (2007) 1197–1206
1203
Fig. 3 – Analysis of molecular sizes and decay of endogenous TFF2 and TFF3 in human bowel ostomy discharges and in normal stool. (A and B) Size-exclusion chromatography. Three fresh samples (no incubation) were fractionated by use of a SMART system on a precalibrated Superdex 75 HR 10/30 column. Fractions were analyzed for hTFF2 or hTFF3 by ELISA. The columns were calibrated with molecular mass calibrators, eluting at fractions 12 (67 kDa), 14 (43 kDa), 18 (25 kDa), and 20 (13.7 kDa) (TFF3 monomer = 7 kDa, TFF3 dimer = 14 kDa, TFF2 = 12 kDa). (C and D) Decays of endogenous hTFF2 and hTFF3 in human intestinal contents following incubation at 37 8C (determined by ELISA).
saline within the first 4 h were 102% (86–116%) and 92% (67– 121%), respectively. Despite the low recovery rate, the detectable levels of hTFF2 and hTFF3 in the small intestinal contents remained stable for at least 24 h (after an initial
increase in the hTFF3 levels during the first hours of incubation) (Fig. 4A and B). In the colonic contents, a similar stability was also found for hTFF3, while the hTFF2 levels decreased steadily and were reduced by 70% after 24 h of
Fig. 4 – Recoveries of added hTFF2 and hTFF3 in human stool and ostomy discharges, and in samples of rat gastrointestinal contents after incubation at 37 8C for increasing time intervals. hTFF2 or hTFF3 were added to samples of small and large intestinal contents and feces from humans and from gastric, small and large intestinal contents from rats and incubated at 37 8C for up to 144 h. Samples were collected, and the hTFF levels at the indicated time points were detected by ELISA. Values are expressed as percentage of the amount of TFF detected to time 0 (endogenous TFF levels are subtracted). (A and B) Contents from human jejunum, ileum, colon and feces. (C and D) Contents from rat stomach, small intestine (SI), cecum, and colon.
1204
peptides 28 (2007) 1197–1206
incubation at 37 8C. In the fecal samples, the levels of hTFF2 were reduced eight-fold from the first to the second hour of incubation and decreased to less than 1% of the initially detected concentration after 8 h. The levels of hTFF3 in fecal samples declined gradually from the first hours to reach 6% of the initially measured concentration after 24 h.
3.6.2.
Rat gastrointestinal contents
The mean initial recovery of hTFF2 and hTFF3 from the spiked samples was 50% (range 13–94%) and 59% (27–85%), respectively. The recoveries of TFF2 and TFF3 in saline within the first 4 h were 94% (71–128%) and 102% (80–136%). In the gastric contents, the levels of TFF2 increased continually from 1 h after addition until 144 h after, whereas in the intestinal contents, the hTFF2 concentrations were markedly decreased already after 2 h (Fig. 4C). After 24 h, the levels of hTFF2 in the small and large intestines were decreased to 1–2% of the initial measured concentrations, while the hTFF2 levels in the cecum remained stable from 2 to 144 h after addition. For hTFF3, the general pattern of decay was comparable to that of hTFF2, except in the cecum where the hTFF3 levels decreased with the same rate as in the large intestine (Fig. 4D).
4.
Discussion
Since the trefoil peptides were first identified, there has been a major effort to elucidate the occurrence and function of these molecules. Yet, we still have limited knowledge on the fate of the peptides, when these are administered in vivo. The prevailing view of the peptides as being extremely stable is based on a few reports mainly of in vitro experiments and our own early in vivo work with porcine 125I-TFF2 in rats. We found it relevant to extend this work to facilitate a more reliable prediction of the behavior of the peptides in a potential therapeutic setting. We therefore included studies of the serum concentration–time profile following various systemic routes of administration and established a setup that could represent the conditions in different parts of the GI tract following either oral or rectal application of the peptides. hTFF2 as well as hTFF3 was included in this study, since our previous studies suggested different stabilities of the peptides [15]. The PK analysis showed that i.v.-injected hTFF2 was more extensively distributed to the peripheral compartment and cleared from the circulation in rats than in mice. The previously observed two-fold higher binding of TFF3 in the rat kidney compared to the mouse kidney [9,15] may contribute to this observation. However, in both species, the TFFs exhibited a relatively short distribution phase and a volume of distribution several fold higher than the plasma volume, which is in accord with the previously demonstrated rapid binding of injected TFFs to the proximal tubules of the kidney and to mucus-producing cells, particularly in the GI tract. These data, together with our earlier studies indicating a high stability of i.v.-injected 125I-TFF2 in the stomach and kidney tissues and our recent finding that systemically administered TFF2 is secreted to the gastric lumen in a biologically active form, which is furthermore detectable by ELISA (as was i.v.-injected TFF3) [9], suggest that the peptides
are quickly distributed to specific TFF-binding cells in the GI tract, from where at least a fraction of the TFF is secreted intact and functional. Despite the similar serum concentration–time profiles for i.v.-injected hTFF2 and hTFF3, we found that 4 h after injection in mice, approximately 50% of the injected hTFF2 was detected in the urine, whereas essentially no hTFF3 was measured in the urine of mice injected with hTFF3. Since porcine 125I-TFF2 and 125I-hTFF3 administered to rats both bind to the renal proximal tubules nearly to the same extent, the different levels in urine found in this study may reflect a more extensive renal degradation of TFF3 than of TFF2 [15,16]. This is in accord with our unpublished findings in humans which showed that while the serum levels of hTFF2 and hTFF3 were comparable, the urinary levels of hTFF2 were more than eight times higher than those of hTFF3. However, a significant degree of reabsorption of intact TFF3 from the kidneys to the circulation resulting in negligible urine levels of TFF3 cannot be ruled out. Studies involving, e.g., isolated perfused kidneys may aid in clarifying this issue. We have previously examined the distribution of the TFFs by means of 125I-labeled peptides and gamma counting. In this study, we have primarily used unlabeled TFFs and detected these by ELISAs specific for either hTFF2 or hTFF3, and we generally found that the peptides were markedly faster cleared from the circulation compared to the earlier studies. Our analysis of the molecular forms of hTFF2 and hTFF3 by HPLC (Section 3.4) suggested that 15 min after s.c. injection, practically all hTFF2 and hTFF3 recovered from the serum was intact—for hTFF3 either in dimeric or monomeric forms—and that both gamma counting and the ELISAs detected all the labeled and unlabeled peptides, respectively, present in the sample. The corresponding results obtained after 180 min showed that at this time point, the ELISA measurements correlated much better with the HPLC data than with gamma counting data, particularly for hTFF3, indicating that serum measurements using 125I-labeled TFF2 or TFF3 and gamma counting can be reliably applied to studies on the initial distribution of the peptides, and that the measurements of serum concentrations after longer time intervals should be replaced by ELISA-based measurements. Our finding that the levels of endogenous hTFF2 were significantly lower in the luminal contents from the distal part of the gut compared to the proximal were not unexpected, as hTFF2 is mainly produced in the stomach and proximal duodenum. However, the differences between the proximal and distal gut were small, suggesting that the hTFF2 produced proximally is subject to only limited degradation during the passage through the intestines. The hTFF2 levels measured in the intestinal contents corresponded to the levels found by [17] in the stomach, and are thereby in accord with previous findings [7,12] and our present findings in rats, that hTFF2 is extremely stable in gastric contents and probably reaches the small intestine from the stomach essentially intact. For hTFF3, we found comparable levels in the small and large intestines, which was unexpected. Since hTFF3 is produced in goblet cells throughout the gut as well as in the stomach and salivary glands [6,10], we had expected somewhat higher levels in areas with the highest frequency of goblet cells in the mucosa, i.e., the ileum and the colon. It is possible that the presence of
peptides 28 (2007) 1197–1206
bacterial enzymes in the colon causes a faster breakdown of the endogenous hTFF3 than in the small intestine, which could result in similar levels throughout the gut. However, the material used for measuring hTFF2 and hTFF3 is small and the results may be substantiated by further studies. Size-exclusion chromatography showed that the detected endogenous hTFF3 was solely in a monomeric form, which corresponds to the previous findings in human serum [23], and in GI tissue homogenates from rats [19]. In contrast, both the monomeric and the dimeric forms of TFF3 were detected in intestinal tissue from a patient with Crohn’s disease [3]. To the best of our knowledge, a naturally occurring TFF3 dimer in samples from healthy mammals has not been finally demonstrated, yet the presence of such is mentioned in numerous publications. The reasons for our and others’ failure to detect endogenous TFF3 dimer remain to be clarified, and it is possible that endogenous TFF3 in non-pathological conditions mainly occurs as a monomer. The functional importance of TFF3 monomers versus dimers is unsettled. In vitro studies have shown that the antiapoptotic effect of TFF3 requires a dimer, whereas the pro-migratory effect neither requires the dimer nor an intact trefoil motif [8]. In vivo, both the dimeric and the monomeric forms of TFF3 ameliorated experimental necrotizing enterocolitis [2]. In this respect, it is interesting that part of the systemically injected hTFF3 seems to be reduced to monomer already after 15 min, as determined by HPLC, and it cannot be excluded that the majority of the injected TFF3 actually reach the target tissue as monomer. Thus, it is highly relevant for the understanding of the physiological and pathophysiological roles of TFF3 to investigate in more detail the occurrence in vivo of the different molecular forms of the peptide. The recoveries of exogenous hTFF2 and hTFF3 immediately after addition to GI contents were generally poor. In addition, the levels of hTFF3 in human intestinal contents and of both peptides in rat stomach contents measured after 1 h and onward were in some instances higher than the levels initially measured. We have no clear explanation for this, but it could be due to initial masking of the epitopes on hTFF3 by crosslinking to mucus in the samples and subsequent release from the mucus, when this is degraded. Incubation of the samples both from human and rat GI contents showed that the hTFF levels in samples from the small intestine were generally more stable than those from colon and feces. This could suggest that the trefoil peptides are more susceptible to bacterial proteolytic activity than to the digestive enzymes of the upper GI tract, and might explain the previous reports of the extreme stability of trefoil peptides, since these studies were performed with typical upper GI tract enzymes. Furthermore, the decay of the peptides was markedly faster in the rat GI contents than in the human intestinal contents, which may be caused by a higher proteolytic activity of the gut in the rats. In contrast to the findings in urine after i.v. injection, there were no major differences between the rate of decrease of hTFF2 and hTFF3 in the GI contents incubated at 37 8C, suggesting that the stability of TFF3 is not generally lower than that of TFF2, but depends on the current conditions including the types of enzymes present. Both peptides would be expected to be active in the upper GI tract following oral treatment, since
1205
most of the initially recovered trefoil peptide is still present in the human small intestine even 24 h after incubation. For treatment of the terminal parts of the colon, oral administration may be insufficient to obtain therapeutic levels of TFF. In conclusion, we have described the pharmacokinetic profiles of hTFF2 in mice and rats and that of hTFF3 in mice following several routes of administration and evaluated the stability of the peptides in GI contents from humans and rats. We found that the peptides are rapidly distributed to the peripheral tissues, and that hTFF3 and hTFF2 are handled differently by the kidneys. Our findings confirm that the trefoil factors are very stable peptides, but also that their stabilities depend on the types of enzymes they are exposed to. Oral administration of the peptides may be sufficient for potential therapy of the upper GI tract, while the rapid degradation of TFFs in the terminal parts of the large intestine questions an effect of the luminally administered trefoil peptides in this location, unless they are protected from degradation by, for example, chemical modifications or a formulation strategy. Our data give new information applicable in future in vivo experiments when determining dosing regimens and routes of administration.
Acknowledgements Stine Kjellev is the recipient of a co-financed Ph.D. fellowship from the Danish Ministry of Science, Technology and Innovation and the Corporate Research Affairs at Novo Nordisk a/s. This work was supported by grants from the Danish Medical Research Council. The technical assistance rendered by Inger Marie Jensen and Birgitte Klitgaard is warmly acknowledged.
references
[1] Babyatsky MW, deBeaumont M, Thim L, Podolsky DK. Oral trefoil peptides protect against ethanol- and indomethacininduced gastric injury in rats. Gastroenterology 1996;110:489–97. [2] Carrasco R, Pera M, May FEB, Westley BR, Martinez A, Morales L. Trefoil factor family peptide 3 prevents the development and promotes healing of ischemiareperfusion injury in weanling rats. J Pediatr Surg 2004;39:1693–700. [3] Chinery R, Bates PA, De A, Freemont PS. Characterisation of the single copy trefoil peptides intestinal trefoil factor and pS2 and their ability to form covalent dimers. FEBS Lett 1995;357:50–4. [4] Davies B, Morris T. Physiological parameters in laboratory animals and humans. Pharm Res 1993;10:1093–5. [5] FitzGerald AJ, Pu M, Marchbank T, Westley BR, May FE, Boyle J, et al. Synergistic effects of systemic trefoil factor family 1 (TFF1) peptide and epidermal growth factor in a rat model of colitis. Peptides 2004;25:793–801. [6] Hoffmann W, Jagla W, Wiede A. Molecular medicine of TFFpeptides: from gut to brain. Histol Histopathol 2001;16:319– 34. [7] Jorgensen KH, Thim L, Jacobsen HE. Pancreatic spasmolytic polypeptide (PSP): I. Preparation and initial chemical
1206
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
peptides 28 (2007) 1197–1206
characterization of a new polypeptide from porcine pancreas. Regul Pept 1982;3:207–19. Kinoshita K, Taupin DR, Itoh H, Podolsky DK. Distinct pathways of cell migration and antiapoptotic response to epithelial injury: structure–function analysis of human intestinal trefoil factor. Mol Cell Biol 2000;20:4680–90. Kjellev S, Nexo E, Thim L, Poulsen SS. Systemically administered trefoil factors are secreted into the gastric lumen and increase the viscosity of gastric contents. Br J Pharmacol 2006;149:92–9. Kouznetsova I, Peitz U, Vieth M, Meyer F, Vestergaard EM, Malfertheiner P, et al. A gradient of TFF3 (trefoil factor family 3) peptide synthesis within the normal human gastric mucosa. Cell Tissue Res 2004;316:155–65. May FEB, Semple JI, Newton JL, Westley BR. The human two domain trefoil protein, TFF2, is glycosylated in vivo in the stomach. Gut 2000;46:454–9. Playford RJ, Marchbank T, Chinery R, Evison R, Pignatelli M, Boulton RA, et al. Human spasmolytic polypeptide is a cytoprotective agent that stimulates cell migration. Gastroenterology 1995;108:108–16. Poulsen SS, Kissow H, Hare K, Hartmann B, Thim L. Luminal and parenteral TFF2 and TFF3 dimer and monomer in two models of experimental colitis in the rat. Regul Pept 2005;126:163–71. Poulsen SS, Thulesen J, Christensen L, Nexo E, Thim L. Metabolism of oral trefoil factor 2 (TFF2) and the effect of oral and parenteral TFF2 on gastric and duodenal ulcer healing in the rat. Gut 1999;45:516–22. Poulsen SS, Thulesen J, Hartmann B, Kissow HL, Nexo E, Thim L. Injected TFF1 and TFF3 bind to TFF2immunoreactive cells in the gastrointestinal tract in rats. Regul Pept 2003;115:91–9.
[16] Poulsen SS, Thulesen J, Nexo E, Thim L. Distribution and metabolism of intravenously administered trefoil factor 2/ porcine spasmolytic polypeptide in the rat. Gut 1998;43:240–7. [17] Semple JI, Newton JL, Westley BR, May FEB. Dramatic diurnal variation in the concentration of the human trefoil peptide TFF2 in gastric juice. Gut 2001;48:648–55. [18] Taupin D, Podolsky DK. Trefoil factors: initiators of mucosal healing. Nat Rev Mol Cell Biol 2003;4:721–32. [19] Taupin DR, Pang KC, Green SP, Giraud AS. The trefoil peptides spasmolytic polypeptide and intestinal trefoil factor are major secretory products of the rat gut. Peptides 1995;16:1001–5. [20] Thim L, Norris K, Norris F, Nielsen PF, Bjorn SE, Christensen M, et al. Purification and characterization of the trefoil peptide human spasmolytic polypeptide (hSP) produced in yeast. FEBS Lett 1993;318:345–52. [21] Thim L, Woldike HF, Nielsen PF, Christensen M, LynchDevaney K, Podolsky DK. Characterization of human and rat intestinal trefoil factor produced in yeast. Biochemistry 1995;34:4757–64. [22] Vestergaard EM, Brynskov J, Ejskjaer K, Clausen JT, Thim L, Nexo E, et al. Immunoassays of human trefoil factors 1 and 2: measured on serum from patients with inflammatory bowel disease. Scand J Clin Lab Invest 2004;64:146–56. [23] Vestergaard EM, Poulsen SS, Gronbaek H, Larsen R, Nielsen AM, Ejskjaer K, et al. Development and evaluation of an ELISA for human trefoil factor 3. Clin Chem 2002;48:1689– 95. [24] Wright NA, Poulsom R, Stamp G, Van Norden S, Sarraf C, Elia G, et al. Trefoil peptide gene expression in gastrointestinal epithelial cells in inflammatory bowel disease. Scand J Gastroenterol Suppl 1992;193:76–82.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/peptides
Pharmacokinetics of trefoil peptides and their stability in gastrointestinal contents§ Stine Kjellev a,*, Else Marie Vestergaard b, Ebba Nexø b, Peter Thygesen c, Maria S. Eghøj d, Palle B. Jeppesen e, Lars Thim f, Nis Borbye Pedersen g, Steen Seier Poulsen d a
Immunopharmacology, Novo Nordisk a/s, DK-2760 Maaloev, Denmark Department of Clinical Biochemistry, Aarhus University Hospital, DK-8000 Aarhus, Denmark c Exploratory ADME, Biopharmaceuticals, Novo Nordisk a/s, DK-2760 Maaloev, Denmark d Department of Medical Anatomy, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark e Department of Medicine CA-2121, Division of Gastroenterology, Rigshospitalet, University of Copenhagen, DK-2100 Copenhagen, Denmark f Protein Engineering, Novo Nordisk a/s, DK-2760 Maaloev, Denmark g Department of Medical Physiology, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark b
article info
abstract
Article history:
Trefoil factor family (TFF) peptides are considered promising for therapeutic use in gastro-
Received 18 October 2006
intestinal diseases, and there is a need to explore the fate of injected TFF and the stability of
Received in revised form
the peptides in the gastrointestinal tract. We studied the pharmacokinetics of intravenously
22 March 2007
(i.v.) administered hTFF2 in mice and rats and of hTFF3 administered i.v., intramuscularly,
Accepted 22 March 2007
intraperitoneally, and subcutaneously in mice, and estimated by ELISA the decay of the
Published on line 31 March 2007
peptides added to rat and human gastrointestinal contents. We found that i.v. injected hTFF2 and hTFF3 were cleared from the circulation within 2–3 h, exhibiting comparable
Keywords:
pharmacokinetic profiles. In contents from the rat stomach, hTFF levels remained
Trefoil factor family
unchanged for up to 6 days. In the small and large intestine of rats, the hTFF levels decreased
TFF2
markedly after 4 and 1 h, respectively. In small intestinal contents from humans, the levels
TFF3
remained stable for more than 24 h. We conclude that systemically administered hTFF2 and
Pharmacokinetics
hTFF3 are rapidly eliminated from the circulation and that the stability of hTFF2 and hTFF3
Clearance
in GI contents appeared higher in the gastric and small intestinal milieu than in the large
Stability
intestine and feces, suggesting a higher stability toward gastric acid and digestive enzymes than toward microbial degradation. # 2007 Elsevier Inc. All rights reserved.
1.
Introduction
The trefoil factor family (TFF) 1–3 comprises a group of small peptides (7–12 kDa) sharing a common structure termed the
§
trefoil domain. TFF2 contains two trefoil domains due to genomic duplication of the motif, while the two domains in TFF1 and TFF3 are linked by disulfide bonds via a seventh cysteine residue (reviewed in [18]). The peptides are produced
SK, PT, and LT are employed at Novo Nordisk a/s. * Corresponding author at: Novo Nordisk Park F6 2.30, DK-2760 Maaloev, Denmark. Tel.: +45 44438062; fax: +45 44434537. E-mail addresses: [email protected] (S. Kjellev), [email protected] (E.M. Vestergaard), [email protected] (E. Nexø), [email protected] (P. Thygesen), [email protected] (M.S. Eghøj), [email protected] (P.B. Jeppesen), [email protected] (L. Thim), [email protected] (N.B. Pedersen), [email protected] (S.S. Poulsen). Abbreviations: ELISA, enzyme-linked immunosorbant assay; GI, gastrointestinal; HPLC, high-performance liquid chromatography; ID, injected dose; i.m., intramuscular; i.p., intraperitoneal; i.v., intravenous; %ID, percentage of injected dose; PK, pharmacokinetic; s.c., subcutaneous; TFF, trefoil factor family 0196-9781/$ – see front matter # 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2007.03.016
1198
peptides 28 (2007) 1197–1206
AUC area under the curve AUC%Extrapol extrapolated percentage of the molecule left in serum after last measured time point start concentration in serum C0 maximal serum concentration Cmax CL clearance F bioavailability MRT mean residence time half-life of molecule in serum T1/2 time to maximal serum concentration Tmax Vz volume of distribution volume of distribution at steady state Vss lz terminal elimination rate constant
by mucus-secreting cells, especially in the gastrointestinal (GI) tract in a site-specific manner for each peptide, i.e., TFF1 is mainly produced by gastric surface epithelial cells, TFF2 by gastric mucous neck cells and pyloric glands, and in the duodenal Brunner’s glands, whereas TFF3 is mainly produced by the intestinal goblet cells and the antral mucous cells [6]. Upregulation and ectopic expression of the peptides have been reported in GI tissues from patients with chronic GI inflammation [24], and all three peptides have shown a therapeutic effect in experimental models of GI disease [1,5,12–14]. Earlier in vitro studies reported that TFF2 was extremely resistant to protease and acid, based on the findings that TFF2 incubated with HCl or trypsin was only slightly degraded [7]. Likewise, 125 I-hTFF2 seemed stable following incubation in small intestinal or gastric juice or in HCl- and GI protease-containing solutions [12]. Our previous studies using 125I-labeled TFFs administered intravenously (i.v.) to rats suggested that the peptides were rather slowly cleared from the circulation and that TFF2 was very resistant toward degradation in vivo [16]. These findings in concert with the compact structure of the TFF peptides have led to the assumption that all TFFs were generally highly resistant toward degradation [12,21]. Pharmacological effects of TFFs have been obtained after systemic as well as local administration, and the aim of this study was therefore to study in more detail the distribution and elimination in vivo by examining the pharmacokinetics of human TFF2 and TFF3 (hTFF2 and hTFF3) in mice and rats after systemic administration and to study the stability of TFFs added to different types of rat GI contents (to mimic the conditions following local administration) using enzymelinked immunosorbant assays (ELISA) specifically developed for hTFF2 or hTFF3. Finally, this paper presents the results of studies on the levels of hTFF2 and hTFF3 in human bowel contents from different parts of the gut, and the effect of storage at 37 8C on the concentration of these peptides.
2.
Materials and methods
2.1.
Animals
Female BALB/cAnNTac mice (20–23 g) and female Wistar Hannover GALAS Rats (HanTac:WH) (250–300 g) (Taconic
M&B, Ry, Denmark) were housed at the Panum Institute (University of Copenhagen, Denmark) or Novo Nordisk a/s. The animal studies were approved by the Danish National Committee for Animal Studies (i.e., the Animal Experiments Inspectorate).
2.2.
Subjects
From humans, samples of normal stool (n = 5) were obtained as well as wet ostomy samples from jejunostomies (n = 8), ileostomies (n = 5), and colostomies (n = 5). None of the subjects had active disease at the time of sampling. The samples were collected after lunch (the subjects were not fasted), were frozen immediately after collection, and stored at 20 8C until analysis. The local scientific ethics committee approved all procedures. All participants gave informed consent.
2.3.
Peptides
hTFF2 and dimeric hTFF3 were produced as described previously [20,21]. For the study of the molecular forms of TFF (Section 2.7), both peptides were labeled at Novo Nordisk a/s with sodium 125iodide (Na125I) using lactoperoxidase to a radiochemical purity of 98%. The specific radioactivity was 2.2 mCi/pmol (hTFF3) and 0.12 mCi/pmol (hTFF2). The tracers were diluted with 0.9% saline to obtain the final concentrations for injection (Section 2.7).
2.4. rats
Pharmacokinetics of hTFF2 and hTFF3 in mice and
hTFF2 or hTFF3 was administered to mice and rats i.v., intraperitoneally (i.p.), intramuscularly (i.m.), or subcutaneously (s.c.) according to Table 1. Prior to the i.v. injection, the mice were anesthetized with a mixture of fentanyl (0.15 mg/kg), droperidol (9.8 mg/kg), and midazolam (1 mg/ kg) administered s.c. The rats were anesthetized with methohexital (50 mg/kg) i.p. (buprenorfin 0.1 mg/kg s.c. for analgesia). The i.v. injection into the inferior vena cava was performed via a midline incision, which was closed with 3–0 silk sutures following injection. For collection of urine, the urethra was ligated when the hTFF was administered. The animals were sacrificed in groups of three to five at the time points specified in Table 1. Prior to sacrifice, blood and urine sampling was performed under general anesthesia. Blood was collected from the orbital venous plexus and urine was directly aspirated from the bladder. The blood was allowed to
Table 1 – Dosing regimens for pharmacokinetic analysis Species Peptide
Route
Dose (mg/kg)
Mouse Mouse Mouse Mouse Rat
i.m. i.p., s.c. i.v. i.v. i.v.
1 1, 5 5 5 5
a
TFF3 TFF3 TFF3 TFF2 TFF2
Simultaneous collection of urine.
Time (min) 6, 15, 30, 60, 120, 180 15, 30, 60, 180 2, 6, 15, 45a, 120a, 240 2, 6, 15, 45a, 120a, 240a 2, 6, 15, 45a, 120a, 240a
peptides 28 (2007) 1197–1206
stand for 1 h prior to centrifugation and collection of serum. Urine was frozen immediately after sampling. All samples were stored at 20 8C.
2.5. Stability of hTFF2 and hTFF3 in gastrointestinal luminal contents We measured the decay of endogenous and exogenously added hTFFs in GI contents to assess in which parts of the GI tract a therapeutic dose of TFFs can be expected to be present following oral treatment, and which dosing frequency would be appropriate. Instead of administering the hTFFs orally and performing PK analysis on the GI contents, we chose a simpler setup, where hTFFs were added directly to contents from various parts of the GI tract. In this way, we could separate the actual hTFF degradation process from influence on the hTFF levels due to different transit times in the various parts of the gut, possible enteral absorption and secretion of hTFFs. The human samples were diluted 2- or 2.5-fold in 0.9% NaCl. Jejunostomy, ileostomy, colostomy, and fecal samples were spiked with hTFF2 or hTFF3 suspended in 0.9% NaCl in sufficient amounts to increase the concentration of total hTFF2 and hTFF3 by 180 and 1900 nM, respectively. Endogenous TFF was measured in identical samples without added TFF. Furthermore, a series of samples containing only 180 nM hTFF2 or 1900 nM hTFF3 in pure 0.9% NaCl was run in parallel. Samples were incubated at 37 8C for 0, 1, 2, 4, 24, 72, and 120 h. Boiling the samples for approximately 1 min stopped incubation. (Termination of incubation by this process was confirmed by control experiments. Furthermore, the ELISAs have similar sensitivity for boiled and unboiled hTFFs (data not shown).) The samples were kept at 20 8C until analysis. Prior to analysis, the suspensions were centrifuged and the supernatants used for analysis. The stabilities of hTFF2 and hTFF3 added in vitro were also measured in contents from the rat stomach, small intestine, cecum, and colon in a similar fashion (samples were pooled from three rats, and an additional 144-h time point was included). The rats were not fasted. hTFF2 and hTFF3 were added to rat bowel discharges at a concentration of 250 nM.
2.6.
hTFF2 and hTFF3 ELISA
The concentrations of hTFF2 and hTFF3 in serum, urine and GI contents were measured by in-house peptide-specific ELISAs as described [22,23]. Briefly, the detection limits for the hTFF2 and hTFF3 ELISAS were 6 and 3 pmol/L, respectively, and the analytical imprecision 7% for mean concentrations of 17– 77 pmol/L (hTFF2) and 13–65 pmol/L (hTFF3), respectively. No cross-reactivity was observed for human trefoil peptides, as 320 nmol/L (TFF1 and TFF3) and 40 nmol/L (TFF1 and TFF2) were below the detection limits for the TFF2 and TFF3 ELISAs, respectively. Western blot analysis of the capture antibody used in the TFF3-ELISA showed that there were no crossreactions with recombinant mouse TFF3 dimer (data not shown). Negligible levels of TFF2 and TFF3 were detected in plasma samples from mouse and rat that were not injected with the human peptides (shown for rats in [9]). The percentage of recovery in the spiked samples of GI contents from humans was determined by comparing the amount of
1199
added hTFF2 and hTFF3 with the amount measured after subtracting the endogenous hTFF2 and hTFF3 concentrations.
2.7. Molecular forms of injected hTFF2 and hTFF3 recovered in serum High-performance liquid chromatography (HPLC), gamma counting, and ELISA were performed on serum samples from animals injected s.c. with a mixture of 125I-hTFF2 (1,160,000 cpm/mouse (15 min) or 5,800,000 cpm/mouse (180 min)) and unlabeled hTFF2 (1 mg/kg) or with a mixture of 125I-hTFF3 (300,000 cpm/mouse) and unlabeled hTFF3 (1 mg/ kg). The blood was collected 15 or 180 min after injection of the tracer, and serum was prepared. The radioactivity in the serum samples was measured by a gamma counter (Wallac Wizard 1740, Perkin-Elmer, Boston, MA, USA) and the concentration of hTFF2 or hTFF3 measured by ELISA as described above (n = 4 per time point). We then calculated the percentage of the injected dose (%ID) of tracer detected in the blood by both methods. For HPLC a volume of 100–200 mL of each serum sample (two to three animals per time point) was injected onto a Vydac 214 TP54 column (250 mm 4.6 mm, Hesperia, CA, USA) equilibrated at room temperature at a flow rate of 1 mL/min with acetonitrile/water/trifluoroacetic acid (10.0:89.9:0.1, w/w/w). After application of the sample, the concentration of acetonitrile in the eluting solvent was raised to 50% (v/v) over 20 min (125I-hTFF3) or 90% (v/v) over 40 min (125I-hTFF2) using a linear gradient. Fractions of 1 mL were collected and subjected to gamma counting, and background radioactivity was subtracted to obtain the actual radioactivity of the samples. In this system, hTFF2 eluted after 20–21 min, hTFF3 dimer eluted after 15 min, hTFF3 monomer eluted after 19 min, free 125I, 125I-tyrosine, or 125I-tyrosine in small peptide fragments (1–10 amino acids) eluted after approximately 4 min.
2.8.
Size-exclusion chromatography
Fresh human stool samples not spiked with TFF were subjected to size-exclusion chromatography on a Superdex 75 HR 10/30 column using a SMART system (Amersham Pharmacia Biotech, GE Healthcare Europe, Munich, Germany). The column was pre-equilibrated and eluted with 0.05 M Tris– HCl, pH 7.2, 9.0 g/L NaCl at 0.5 mL/min with collection of 0.5mL fractions. The column was calibrated with molecular mass calibrators from Amersham Pharmacia Biotech, eluting at fractions 12 (67 kDa), 14 (43 kDa), 18 (25 kDa), and 20 (13.7 kDa). Purified hTFF3 monomer and hTFF3 dimer were used as controls. The distribution of hTFF2 and hTFF3 immunoreactivity was determined by ELISA.
2.9.
Statistical analysis
The pharmacokinetic data were analyzed by a non-compartmental analysis using WinNonlin Professional (Pharsight Inc., Mountain View, CA, USA). Calculations were performed using the mean concentration–time values from three to five animals at each time point. For data analysis of TFF contents in bowel discharge, we used the Prism1 version 4 (GraphPad, San Diego, CA, USA). Differences between the levels of
1200
peptides 28 (2007) 1197–1206
endogenous hTFFs in stomy and fecal samples were analyzed using Student’s t-test with Welsch’s correction for unequal variances.
3.
Results
3.1.
Pharmacokinetics of i.v. injected hTFF2
The pharmacokinetics of hTFF2 in mice and rats after i.v. injection are shown in Fig. 1A and B and Table 2. In both species, an initial short distribution phase was identified, followed by a first-order elimination phase (Fig. 1A). The distribution was completed faster in mice than rats (approximately 15 min vs. 30–45 min; Fig. 1B), but the impact on the serum concentration–time profile was greater in rats compared to mice indicating a more pronounced distribution and redistribution of hTFF2 in rats. The terminal elimination phase was similar in mice and rats as indicated by the two parallel serum concentration–time lines (Fig. 1B). AUC%Extrapol was less than 10% indicating almost complete clearance of from the blood by the end of the study period. The maximal mean serum concentrations observed were 3190 and 2680 nM in mice and rats, respectively (Table 2), while the exposure of hTFF2, expressed as dose-corrected area under the curve, was approximately three-fold higher in mice compared to rats (3.8 vs. 1.3 h kg/L) due to a three-fold
higher clearance in rats (13 vs. 4.4 mL/min kg). The terminal elimination half-life was similar in the two species, 48 and 50 min, respectively. The volume of distribution was threefold higher in rats compared to mice (920 vs. 300 mL/kg). When compared to average plasma volumes of 50 and 31 mL/ kg in mice and rats, respectively [4], a more extensive distribution outside the circulation or specific cellular binding of hTFF2 was observed in rats compared to mice. The mean residence time was 52 and 12 min in mice and rats, respectively, indicating that hTFF2 left the circulation much faster in rats than mice.
3.2. Pharmacokinetics of systemically administered hTFF3 in mice The pharmacokinetics of hTFF3 in mice after different routes of administration (i.v., s.c., i.p., and i.m.) are shown in Fig. 1C and D and Table 3 (AUC%Extrapol was less than 10%). The serum concentration–time profile of i.v. injected TFF3 was quite similar to that of TFF2 with an initial short distribution phase followed by a first-order elimination phase. Serum concentration–time profiles after i.p., i.m., and s.c. administration all had an absorption phase and a terminal elimination phase similar to that observed after i.v. administration. No distribution phases could be identified, most likely due to interference between the absorption and distribution processes. The pharmacokinetic parameter
Fig. 1 – The mean pharmacokinetic profiles of hTFF2 and hTFF3 injected to mice and rats. Mice (three to four per time point) were injected i.v., i.m., s.c., or i.p. with hTFF3, and rats and mice were injected i.v. with hTFF2 (three to five per time point) according to Table 1. The serum levels of the injected peptides at different time points after injection were measured by ELISA. The pharmacokinetic profiles were analyzed by a non-compartmental analysis. (A) Mean serum concentration–time profiles of hTFF2 injected i.v. in rats and mice. Note the short initial distribution phase, and the first-order elimination phase. (B) Mean semi-logarithmic serum concentration–time profiles of hTFF2 injected i.v. to rats and mice. The parallel lines of the elimination phase indicate a similar elimination phase in rats and mice. (C) Mean dose normalized serum concentration–time profiles of hTFF3 in mice after i.v., i.m., i.p., and s.c. administration (insert, same experiment but without i.v. data). Note the faster absorption following i.m. injection. (D) The mean (semi-logarithmic) dose normalized serum concentration–time profiles of hTFF3 in mice after i.v., i.m., i.p., and s.c. administration suggest comparable elimination phases (the parallel lines) regardless of the route of administration.
1201
peptides 28 (2007) 1197–1206
Table 2 – Pharmacokinetic parameter estimates of TFF2 in rats and mice after i.v. injection (mean values) Animal Mouse Rat
Dose (nmol/kg)
Tmax (min)
Cmax (nM)
C0 (nM)
AUC (h nM)
AUC/Dose (h kg/L)
lz (min 1)
T1/2 (min)
CL (mL/min kg)
Vz (mL/kg)
Vss (mL/kg)
332 307
2 2
3190 2680
4380 4010
1260 397
3.8 1.3
0.0145 0.0140
48 50
4 13
300 920
220 150
MRT (min) 52 12
Non-compartmental analysis. Tmax, time to maximal serum concentration; Cmax, maximal serum concentration; C0, starting concentration in serum (extrapolated); AUC, area under the curve; AUC%Extrapol, extrapolated percentage of the molecule left in serum after last measured time point; lz, elimination; T1/2, half-life; CL, clearance; Vz, volume of distribution; Vss, volume of distribution at steady state; MRT, mean residence time.
estimates of hTFF3 in mice are presented in Table 3. The dose-corrected maximal mean serum concentrations (Cmax) were 410, 110, 35, and 34 nM for i.v., i.m., i.p., and s.c administrations, respectively, indicating that i.m. administration results in the highest Cmax/dose when hTFF3 is given extravascularly. The time to maximal serum concentration was also shortest after i.m. administration compared to i.p. or s.c. administration (15 vs. 30 min; Fig. 1C, insert). The exposure in terms of the dose-corrected area under the curve after extravascular administration was 4400, 2300, and 3100 min/L for i.m., i.p., and s.c. administrations, respectively, resulting in absolute bioavailabilities of 30, 16, and 21% after i.m., i.p., and s.c. administrations, respectively. The terminal half-life was similar among all routes of administration and varied between 30 and 44 min. This was also the case for the body clearance and volume of distribution when corrected for bioavailability, indicating similar elimination processes and tissue distribution or cellular binding when the hTFF3 molecules reached the circulation. The mean residue times were also similar except for s.c. administration, where this value indicated a slightly slower absorption phase.
3.3. hTFF2 and hTFF3 levels in the urine of rats and mice following i.v. injection Forty-five minutes after i.v. injection of hTFF2 to mice, 9 (6, 13)% (mean (min, max)) of the injected dose (ID) was detected in the urine. After 2 h, this had increased to 13 (9, 18)% ID and after 4 h to 30 (21, 40)% ID. In rats, the corresponding values were 49 (41, 54)% after 45 min, 49 (43,56)% after 120 min, and 28% (n = 1) after 240 min. Virtually no hTFF3 was excreted in
the urine: The %ID was 0.003 (0.001, 0.005)% after 45 min and 0.01 (0.000, 0.002)% 2 h after injection.
3.4. Molecular forms of injected hTFF2 and hTFF3 recovered in mouse serum The %ID of the 125I-hTFF2/hTFF2 mixture in the serum 15 min after injection was 11.4 1.4 (S.D.) when detected by gamma counting and 15.6 7.9 (S.D.) when measured by ELISA (not significantly different). After 180 min, the corresponding values were 3.0 0.5% and 0.33 0.13%, respectively. HPLC of the serum samples showed that 15 min after injection, all detected 125I-hTFF2 eluted at a position corresponding to the tracer (intact peptide; Fig. 2A). One hundred and eighty minutes after injection 40% of the total radioactivity was detected at a position corresponding to free 125I or 125Ityrosine in small peptide fragments (degraded peptide), and 23% at a position corresponding to the tracer. About 15% of the radioactivity eluted in the following two fractions, i.e., after 22–23 min (Fig. 2B). The %ID of the 125I-hTFF3/hTFF3 mixture in the serum 15 min after injection was 5.9 0.7 (S.D.) when detected by gamma counting and 5.5 0.7 (S.D.) when measured by ELISA. After 180 min, the corresponding values were 8.4 1.3 % and 0.1 0.1 %, respectively. HPLC of the serum samples taken 15 min after s.c. injection showed that most of the injected 125I-hTFF3 eluted at a position corresponding to the dimeric form, whereas 5–10% in sample 1 and 20–25% in sample 2 eluted in a position corresponding to monomeric TFF3 (Fig. 2C). HPLC of the serum samples taken after 180 min showed that practically all injected TFF3 eluted in a position corresponding to degraded peptide (Fig. 2D).
Table 3 – Pharmacokinetic parameter estimates of TFF2 and TFF3 in mice after i.v., i.m., i.p., and s.c. administration (mean values) Peptide
RoA
TFF2 TFF3 TFF3 TFF3 TFF3
i.v. i.v. i.m. i.p. s.c.
Dose Tmax Cmax (nmol) (min) (nM) 8.3 9.3 1.6 1.6 6.5
2 2 15 30 30
3190 3830 169 55 225
C0 (nM) 4380 4690
Cmax/ T1/2 CL Vz Vss MRT F AUC AUC/dose lz dose (min nM) (min/L) (min 1) (min) (mL/min) (mL) (mL) (min) (%) 380 410 110 35 34
75,900 136,000 6,830 3,570 20,300
9,100 14,700 4,400 2,300 3,100
0.0145 0.0158 0.0193 0.0236 0.0159
48 44 36 30 44
0.1 0.1 0.2 0.4 0.3
8 4 12 19 20
5.7 4.1
52 60 53 58 76
30 16 21
Non-compartmental analysis. RoA, route of administration; Tmax, time to maximal serum concentration; Cmax, maximal serum concentration; C0, start concentration in serum (extrapolated); AUC, area under the curve; AUC%Extrapol, extrapolated percentage of the molecule left in serum after last measured time point; lz, elimination; T1/2, half-life; CL, clearance; Vz, volume of distribution; Vss, volume of distribution at steady state; MRT, mean residence time; F, bioavailability.
1202
peptides 28 (2007) 1197–1206
Fig. 2 – HPLC profiles of systemically administered 125I-hTFF2 or 125I-hTFF3. HPLC was performed on serum samples from mice injected s.c. with a mixture of 125I-hTFF2/unlabeled hTFF2 (A and B) or 125I-hTFF3/unlabelled (C and D) at 15 min (A and C) and 180 min (B and D) prior to blood sampling. Intact hTFF2 elutes after 20–21 min, hTFF3 dimer after 15 min, hTFF3 monomer after 19 min and free 125I, 125I-tyrosine, and other small 125I-labeled peptide fragments after 1–4 min.
3.5. Concentrations and stability of endogenous hTFF2 and hTFF3 in intestinal discharges and in stool The concentrations of endogenous hTFF2 and hTFF3 in the contents collected from human jejunostomies, ileostomies, colostomies, and stool are shown in Table 4. The highest concentrations were found for hTFF2, and these were from 2000 to 10,000 times higher than the concentrations found in human serum. For hTFF3, the concentrations were from 140 to 500 times the serum concentrations. Using size exclusion chromatography, we found, for hTFF2 in jejunostomy samples, a sharp peak corresponding to a molecular mass of approximately 25 kDa, whereas in the samples from the ileostomy and the colostomy, the peak was broader (Fig. 3A). Since un-glycosylated hTFF2 has a molecular weight of 12 kDa and glycosylated hTFF2 weighs around 20 kDa, this suggests
Table 4 – Mean (S.D.) concentrations of endogenous hTFF2 and hTFF3 in human small bowel discharges and in stool hTFF2 (nM) Jejunostomy Ileostomy colostomy Normal stool Normal serum a
736 (541) 462 (443) 147 (86)* 241 (182) 0.076a
*
TFF3 (nM) 64 (61) 42 (33) 70 (52) 19 (10) 0.140b
Ref. [22]. Ref. [23]. n = 300 and 296 (serum, median values) and four to eight (bowel discharges). * P < 0.05 (jejunostomy vs. colostomy). b
that the hTFF2 in the samples is mainly present in the glycosylated form, as described previously for the stomach [11]. For hTFF3, a single major peak at 6.6 kDa, corresponding to the reported molecular mass of monomeric hTFF3 was detected (Fig. 3B). We measured the levels of endogenous hTFF2 and hTFF3 in these samples after incubation at 37 8C, as shown in Fig. 3C and D. In the jejunal contents, the hTFF2 and hTFF3 levels were essentially unchanged after 24 h of incubation, and 50% of the initial amount could be retrieved after 120 h. In ileum, 66% of the initial hTFF2 and 56% of the initial hTFF3 were detectable after 120 h. In the colonic contents, the levels of both peptides declined gradually (although higher concentrations of hTFF3 were measured at 4 and 8 h). After 24 h, 50% of the initial amount of hTFF3 and 35% of hTFF2 was detectable. In the fecal samples, the levels of hTFF2 and hTFF3 were low but stable for at least 24 h after, when the levels declined markedly (for unknown reasons, the TFF levels in these samples were lower than in the remaining fecal samples).
3.6. Stability of added hTFF2 and hTFF3 in rat and human GI contents 3.6.1.
Human bowel discharges
We then assessed the stability of hTFF2 and hTFF3 dimer added in vitro to the bowel discharges in order to asses how long time locally administered TFFs would be potentially active in the different parts of the intestinal tract. The added hTFFs were generally difficult to detect in the samples: In the samples without incubation at 37 8C, 49% (range 21–67%) of added hTFF2 and 25% (9–48%) of added hTFF3 could be recovered. In contrast, the recoveries of hTFF2 and hTFF3 in
peptides 28 (2007) 1197–1206
1203
Fig. 3 – Analysis of molecular sizes and decay of endogenous TFF2 and TFF3 in human bowel ostomy discharges and in normal stool. (A and B) Size-exclusion chromatography. Three fresh samples (no incubation) were fractionated by use of a SMART system on a precalibrated Superdex 75 HR 10/30 column. Fractions were analyzed for hTFF2 or hTFF3 by ELISA. The columns were calibrated with molecular mass calibrators, eluting at fractions 12 (67 kDa), 14 (43 kDa), 18 (25 kDa), and 20 (13.7 kDa) (TFF3 monomer = 7 kDa, TFF3 dimer = 14 kDa, TFF2 = 12 kDa). (C and D) Decays of endogenous hTFF2 and hTFF3 in human intestinal contents following incubation at 37 8C (determined by ELISA).
saline within the first 4 h were 102% (86–116%) and 92% (67– 121%), respectively. Despite the low recovery rate, the detectable levels of hTFF2 and hTFF3 in the small intestinal contents remained stable for at least 24 h (after an initial
increase in the hTFF3 levels during the first hours of incubation) (Fig. 4A and B). In the colonic contents, a similar stability was also found for hTFF3, while the hTFF2 levels decreased steadily and were reduced by 70% after 24 h of
Fig. 4 – Recoveries of added hTFF2 and hTFF3 in human stool and ostomy discharges, and in samples of rat gastrointestinal contents after incubation at 37 8C for increasing time intervals. hTFF2 or hTFF3 were added to samples of small and large intestinal contents and feces from humans and from gastric, small and large intestinal contents from rats and incubated at 37 8C for up to 144 h. Samples were collected, and the hTFF levels at the indicated time points were detected by ELISA. Values are expressed as percentage of the amount of TFF detected to time 0 (endogenous TFF levels are subtracted). (A and B) Contents from human jejunum, ileum, colon and feces. (C and D) Contents from rat stomach, small intestine (SI), cecum, and colon.
1204
peptides 28 (2007) 1197–1206
incubation at 37 8C. In the fecal samples, the levels of hTFF2 were reduced eight-fold from the first to the second hour of incubation and decreased to less than 1% of the initially detected concentration after 8 h. The levels of hTFF3 in fecal samples declined gradually from the first hours to reach 6% of the initially measured concentration after 24 h.
3.6.2.
Rat gastrointestinal contents
The mean initial recovery of hTFF2 and hTFF3 from the spiked samples was 50% (range 13–94%) and 59% (27–85%), respectively. The recoveries of TFF2 and TFF3 in saline within the first 4 h were 94% (71–128%) and 102% (80–136%). In the gastric contents, the levels of TFF2 increased continually from 1 h after addition until 144 h after, whereas in the intestinal contents, the hTFF2 concentrations were markedly decreased already after 2 h (Fig. 4C). After 24 h, the levels of hTFF2 in the small and large intestines were decreased to 1–2% of the initial measured concentrations, while the hTFF2 levels in the cecum remained stable from 2 to 144 h after addition. For hTFF3, the general pattern of decay was comparable to that of hTFF2, except in the cecum where the hTFF3 levels decreased with the same rate as in the large intestine (Fig. 4D).
4.
Discussion
Since the trefoil peptides were first identified, there has been a major effort to elucidate the occurrence and function of these molecules. Yet, we still have limited knowledge on the fate of the peptides, when these are administered in vivo. The prevailing view of the peptides as being extremely stable is based on a few reports mainly of in vitro experiments and our own early in vivo work with porcine 125I-TFF2 in rats. We found it relevant to extend this work to facilitate a more reliable prediction of the behavior of the peptides in a potential therapeutic setting. We therefore included studies of the serum concentration–time profile following various systemic routes of administration and established a setup that could represent the conditions in different parts of the GI tract following either oral or rectal application of the peptides. hTFF2 as well as hTFF3 was included in this study, since our previous studies suggested different stabilities of the peptides [15]. The PK analysis showed that i.v.-injected hTFF2 was more extensively distributed to the peripheral compartment and cleared from the circulation in rats than in mice. The previously observed two-fold higher binding of TFF3 in the rat kidney compared to the mouse kidney [9,15] may contribute to this observation. However, in both species, the TFFs exhibited a relatively short distribution phase and a volume of distribution several fold higher than the plasma volume, which is in accord with the previously demonstrated rapid binding of injected TFFs to the proximal tubules of the kidney and to mucus-producing cells, particularly in the GI tract. These data, together with our earlier studies indicating a high stability of i.v.-injected 125I-TFF2 in the stomach and kidney tissues and our recent finding that systemically administered TFF2 is secreted to the gastric lumen in a biologically active form, which is furthermore detectable by ELISA (as was i.v.-injected TFF3) [9], suggest that the peptides
are quickly distributed to specific TFF-binding cells in the GI tract, from where at least a fraction of the TFF is secreted intact and functional. Despite the similar serum concentration–time profiles for i.v.-injected hTFF2 and hTFF3, we found that 4 h after injection in mice, approximately 50% of the injected hTFF2 was detected in the urine, whereas essentially no hTFF3 was measured in the urine of mice injected with hTFF3. Since porcine 125I-TFF2 and 125I-hTFF3 administered to rats both bind to the renal proximal tubules nearly to the same extent, the different levels in urine found in this study may reflect a more extensive renal degradation of TFF3 than of TFF2 [15,16]. This is in accord with our unpublished findings in humans which showed that while the serum levels of hTFF2 and hTFF3 were comparable, the urinary levels of hTFF2 were more than eight times higher than those of hTFF3. However, a significant degree of reabsorption of intact TFF3 from the kidneys to the circulation resulting in negligible urine levels of TFF3 cannot be ruled out. Studies involving, e.g., isolated perfused kidneys may aid in clarifying this issue. We have previously examined the distribution of the TFFs by means of 125I-labeled peptides and gamma counting. In this study, we have primarily used unlabeled TFFs and detected these by ELISAs specific for either hTFF2 or hTFF3, and we generally found that the peptides were markedly faster cleared from the circulation compared to the earlier studies. Our analysis of the molecular forms of hTFF2 and hTFF3 by HPLC (Section 3.4) suggested that 15 min after s.c. injection, practically all hTFF2 and hTFF3 recovered from the serum was intact—for hTFF3 either in dimeric or monomeric forms—and that both gamma counting and the ELISAs detected all the labeled and unlabeled peptides, respectively, present in the sample. The corresponding results obtained after 180 min showed that at this time point, the ELISA measurements correlated much better with the HPLC data than with gamma counting data, particularly for hTFF3, indicating that serum measurements using 125I-labeled TFF2 or TFF3 and gamma counting can be reliably applied to studies on the initial distribution of the peptides, and that the measurements of serum concentrations after longer time intervals should be replaced by ELISA-based measurements. Our finding that the levels of endogenous hTFF2 were significantly lower in the luminal contents from the distal part of the gut compared to the proximal were not unexpected, as hTFF2 is mainly produced in the stomach and proximal duodenum. However, the differences between the proximal and distal gut were small, suggesting that the hTFF2 produced proximally is subject to only limited degradation during the passage through the intestines. The hTFF2 levels measured in the intestinal contents corresponded to the levels found by [17] in the stomach, and are thereby in accord with previous findings [7,12] and our present findings in rats, that hTFF2 is extremely stable in gastric contents and probably reaches the small intestine from the stomach essentially intact. For hTFF3, we found comparable levels in the small and large intestines, which was unexpected. Since hTFF3 is produced in goblet cells throughout the gut as well as in the stomach and salivary glands [6,10], we had expected somewhat higher levels in areas with the highest frequency of goblet cells in the mucosa, i.e., the ileum and the colon. It is possible that the presence of
peptides 28 (2007) 1197–1206
bacterial enzymes in the colon causes a faster breakdown of the endogenous hTFF3 than in the small intestine, which could result in similar levels throughout the gut. However, the material used for measuring hTFF2 and hTFF3 is small and the results may be substantiated by further studies. Size-exclusion chromatography showed that the detected endogenous hTFF3 was solely in a monomeric form, which corresponds to the previous findings in human serum [23], and in GI tissue homogenates from rats [19]. In contrast, both the monomeric and the dimeric forms of TFF3 were detected in intestinal tissue from a patient with Crohn’s disease [3]. To the best of our knowledge, a naturally occurring TFF3 dimer in samples from healthy mammals has not been finally demonstrated, yet the presence of such is mentioned in numerous publications. The reasons for our and others’ failure to detect endogenous TFF3 dimer remain to be clarified, and it is possible that endogenous TFF3 in non-pathological conditions mainly occurs as a monomer. The functional importance of TFF3 monomers versus dimers is unsettled. In vitro studies have shown that the antiapoptotic effect of TFF3 requires a dimer, whereas the pro-migratory effect neither requires the dimer nor an intact trefoil motif [8]. In vivo, both the dimeric and the monomeric forms of TFF3 ameliorated experimental necrotizing enterocolitis [2]. In this respect, it is interesting that part of the systemically injected hTFF3 seems to be reduced to monomer already after 15 min, as determined by HPLC, and it cannot be excluded that the majority of the injected TFF3 actually reach the target tissue as monomer. Thus, it is highly relevant for the understanding of the physiological and pathophysiological roles of TFF3 to investigate in more detail the occurrence in vivo of the different molecular forms of the peptide. The recoveries of exogenous hTFF2 and hTFF3 immediately after addition to GI contents were generally poor. In addition, the levels of hTFF3 in human intestinal contents and of both peptides in rat stomach contents measured after 1 h and onward were in some instances higher than the levels initially measured. We have no clear explanation for this, but it could be due to initial masking of the epitopes on hTFF3 by crosslinking to mucus in the samples and subsequent release from the mucus, when this is degraded. Incubation of the samples both from human and rat GI contents showed that the hTFF levels in samples from the small intestine were generally more stable than those from colon and feces. This could suggest that the trefoil peptides are more susceptible to bacterial proteolytic activity than to the digestive enzymes of the upper GI tract, and might explain the previous reports of the extreme stability of trefoil peptides, since these studies were performed with typical upper GI tract enzymes. Furthermore, the decay of the peptides was markedly faster in the rat GI contents than in the human intestinal contents, which may be caused by a higher proteolytic activity of the gut in the rats. In contrast to the findings in urine after i.v. injection, there were no major differences between the rate of decrease of hTFF2 and hTFF3 in the GI contents incubated at 37 8C, suggesting that the stability of TFF3 is not generally lower than that of TFF2, but depends on the current conditions including the types of enzymes present. Both peptides would be expected to be active in the upper GI tract following oral treatment, since
1205
most of the initially recovered trefoil peptide is still present in the human small intestine even 24 h after incubation. For treatment of the terminal parts of the colon, oral administration may be insufficient to obtain therapeutic levels of TFF. In conclusion, we have described the pharmacokinetic profiles of hTFF2 in mice and rats and that of hTFF3 in mice following several routes of administration and evaluated the stability of the peptides in GI contents from humans and rats. We found that the peptides are rapidly distributed to the peripheral tissues, and that hTFF3 and hTFF2 are handled differently by the kidneys. Our findings confirm that the trefoil factors are very stable peptides, but also that their stabilities depend on the types of enzymes they are exposed to. Oral administration of the peptides may be sufficient for potential therapy of the upper GI tract, while the rapid degradation of TFFs in the terminal parts of the large intestine questions an effect of the luminally administered trefoil peptides in this location, unless they are protected from degradation by, for example, chemical modifications or a formulation strategy. Our data give new information applicable in future in vivo experiments when determining dosing regimens and routes of administration.
Acknowledgements Stine Kjellev is the recipient of a co-financed Ph.D. fellowship from the Danish Ministry of Science, Technology and Innovation and the Corporate Research Affairs at Novo Nordisk a/s. This work was supported by grants from the Danish Medical Research Council. The technical assistance rendered by Inger Marie Jensen and Birgitte Klitgaard is warmly acknowledged.
references
[1] Babyatsky MW, deBeaumont M, Thim L, Podolsky DK. Oral trefoil peptides protect against ethanol- and indomethacininduced gastric injury in rats. Gastroenterology 1996;110:489–97. [2] Carrasco R, Pera M, May FEB, Westley BR, Martinez A, Morales L. Trefoil factor family peptide 3 prevents the development and promotes healing of ischemiareperfusion injury in weanling rats. J Pediatr Surg 2004;39:1693–700. [3] Chinery R, Bates PA, De A, Freemont PS. Characterisation of the single copy trefoil peptides intestinal trefoil factor and pS2 and their ability to form covalent dimers. FEBS Lett 1995;357:50–4. [4] Davies B, Morris T. Physiological parameters in laboratory animals and humans. Pharm Res 1993;10:1093–5. [5] FitzGerald AJ, Pu M, Marchbank T, Westley BR, May FE, Boyle J, et al. Synergistic effects of systemic trefoil factor family 1 (TFF1) peptide and epidermal growth factor in a rat model of colitis. Peptides 2004;25:793–801. [6] Hoffmann W, Jagla W, Wiede A. Molecular medicine of TFFpeptides: from gut to brain. Histol Histopathol 2001;16:319– 34. [7] Jorgensen KH, Thim L, Jacobsen HE. Pancreatic spasmolytic polypeptide (PSP): I. Preparation and initial chemical
1206
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
peptides 28 (2007) 1197–1206
characterization of a new polypeptide from porcine pancreas. Regul Pept 1982;3:207–19. Kinoshita K, Taupin DR, Itoh H, Podolsky DK. Distinct pathways of cell migration and antiapoptotic response to epithelial injury: structure–function analysis of human intestinal trefoil factor. Mol Cell Biol 2000;20:4680–90. Kjellev S, Nexo E, Thim L, Poulsen SS. Systemically administered trefoil factors are secreted into the gastric lumen and increase the viscosity of gastric contents. Br J Pharmacol 2006;149:92–9. Kouznetsova I, Peitz U, Vieth M, Meyer F, Vestergaard EM, Malfertheiner P, et al. A gradient of TFF3 (trefoil factor family 3) peptide synthesis within the normal human gastric mucosa. Cell Tissue Res 2004;316:155–65. May FEB, Semple JI, Newton JL, Westley BR. The human two domain trefoil protein, TFF2, is glycosylated in vivo in the stomach. Gut 2000;46:454–9. Playford RJ, Marchbank T, Chinery R, Evison R, Pignatelli M, Boulton RA, et al. Human spasmolytic polypeptide is a cytoprotective agent that stimulates cell migration. Gastroenterology 1995;108:108–16. Poulsen SS, Kissow H, Hare K, Hartmann B, Thim L. Luminal and parenteral TFF2 and TFF3 dimer and monomer in two models of experimental colitis in the rat. Regul Pept 2005;126:163–71. Poulsen SS, Thulesen J, Christensen L, Nexo E, Thim L. Metabolism of oral trefoil factor 2 (TFF2) and the effect of oral and parenteral TFF2 on gastric and duodenal ulcer healing in the rat. Gut 1999;45:516–22. Poulsen SS, Thulesen J, Hartmann B, Kissow HL, Nexo E, Thim L. Injected TFF1 and TFF3 bind to TFF2immunoreactive cells in the gastrointestinal tract in rats. Regul Pept 2003;115:91–9.
[16] Poulsen SS, Thulesen J, Nexo E, Thim L. Distribution and metabolism of intravenously administered trefoil factor 2/ porcine spasmolytic polypeptide in the rat. Gut 1998;43:240–7. [17] Semple JI, Newton JL, Westley BR, May FEB. Dramatic diurnal variation in the concentration of the human trefoil peptide TFF2 in gastric juice. Gut 2001;48:648–55. [18] Taupin D, Podolsky DK. Trefoil factors: initiators of mucosal healing. Nat Rev Mol Cell Biol 2003;4:721–32. [19] Taupin DR, Pang KC, Green SP, Giraud AS. The trefoil peptides spasmolytic polypeptide and intestinal trefoil factor are major secretory products of the rat gut. Peptides 1995;16:1001–5. [20] Thim L, Norris K, Norris F, Nielsen PF, Bjorn SE, Christensen M, et al. Purification and characterization of the trefoil peptide human spasmolytic polypeptide (hSP) produced in yeast. FEBS Lett 1993;318:345–52. [21] Thim L, Woldike HF, Nielsen PF, Christensen M, LynchDevaney K, Podolsky DK. Characterization of human and rat intestinal trefoil factor produced in yeast. Biochemistry 1995;34:4757–64. [22] Vestergaard EM, Brynskov J, Ejskjaer K, Clausen JT, Thim L, Nexo E, et al. Immunoassays of human trefoil factors 1 and 2: measured on serum from patients with inflammatory bowel disease. Scand J Clin Lab Invest 2004;64:146–56. [23] Vestergaard EM, Poulsen SS, Gronbaek H, Larsen R, Nielsen AM, Ejskjaer K, et al. Development and evaluation of an ELISA for human trefoil factor 3. Clin Chem 2002;48:1689– 95. [24] Wright NA, Poulsom R, Stamp G, Van Norden S, Sarraf C, Elia G, et al. Trefoil peptide gene expression in gastrointestinal epithelial cells in inflammatory bowel disease. Scand J Gastroenterol Suppl 1992;193:76–82.