Minimum biological domain of xenin-25 required to induce anion secretion in the rat ileum

Xenin-25 has a variety of physiological functions in the gastrointestinal tract, including ion transport and motility. Xenin-25 and neurotensin show sequence homology, especially near their C-terminal regions. The sequence similarity between xenin-25 and neurotensin indicates that the effects of xenin-25 is mediated by the neurotensin receptor but some biological actions of xenin-25 are independent. We have previously reported that xenin-25 modulates intestinal ion transport and colonic smooth muscle activity. However, minimal biological domain of xenin-25 to induce ion transport was not clear. To improve the mechanistic understanding of xenin-25 and to gain additional insights into the functions of xenin-25, the present study was designed to determine the minimal biological domain of xenin-25 required for ion transport in the rat ileum using various truncated xenin fragments and analogues in an Ussing chamber system. The present results demonstrate that the minimum biological domain of xenin-25 to induce Cl-/HCO3- secretion in the ileum contains the C-terminal pentapeptide. Furthermore, Arg at position 21 is important to retain the biological activity of xenin-25 and induces Cl-/HCO3- secretion in the rat ileum.


Introduction
In 1992, the mammalian homolog of the octapeptide xenopsin from amphibian skin was identified in human gastric mucosa [12]. This peptide consists of 25 amino acids and was named xenin-25. Xenin-25 probably appeared early in evolution as its amino acid sequence is highly conserved with those in yeast and mammalian coat protein alpha [20]. Furthermore, xenin-25 shares sequence homology with the mammalian tridecapeptide neurotensin (NT) [3] and is thought to bind to and activate NT receptors [15].
Xenin immunoreactivity has been detected in the hypothalamus, lung, liver, heart, kidney, adrenal gland, pancreas, and small and large intestinal mucosa in various mammals, including rats, guinea pigs, rabbits, dogs, pigs, and humans [19]. In the small intestine, xenin-25 was previously reported to be produced by a subset of intestinal K cells containing glucose-dependent insulinotropic peptide (GIP) in humans, rhesus monkeys, and dogs [2]. However, we recently showed that xenin immunoreactivity is more broadly expressed in enteroendocrine cells (EECs); xenin-25 is produced by a subset of glucagon-like peptide 2 (GLP-2)-, cholecystokinin (CCK)-, or 5-hydroxyltryptamine (5-HT)-producing EECs, in addition to K cells in the rat duodenum [23]. Xenin-25 affects various gastrointestinal (GI) functions including motility, ion transport [24,26,27], appetite regulation [1,7,16,17,28], and pancreatic endocrine and exocrine secretion [5,14,29,38,39] although the type of responses varies according to the species and tissues type. From these observations, xenin-25 appears to be an important regulatory peptide that contributes to the maintenance of body energy homeostasis by exerting its effects on various GI functions and feeding behaviors.
The actions of NT are mediated by three known neurotensin receptors: neurotensin receptor 1 (NTSR1) [37], NTSR2 [4] and NTSR3 [32]. Since xenin-25 is structurally similar to NT, many of the effects of xenin-25 are believed to be mediated by the activation of NT receptors. [6,23,40]. We have recently shown that xenin-25 suppresses spontaneous circular muscle contractions in the rat distal colon and induces increases in basal short-circuit current (Isc) due to Cl − /HCO 3 − secretion by a noncholinergic mechanism in the rat ileum [26,27]. Our previous studies suggest that the effects of xenin-25 involve NTSR1since the response of this receptor was blocked by the NTSR1 antagonist, SR48692 [26,27]. However, some biological activities of xenin-25 are considered to be independent of NT receptors although a specific xenin-25 receptor has not been identified [8,21]. For example, a comparative study of the effects of NT and xenin-25 on motility in the guinea-pig jejunum and colon in vitro revealed a different interaction of these two structurally related peptides with the NTSR [13]. These results raise the possibility that different xenin-25 receptors may exist and be responsible for the different effects. Furthermore, intravenous injection of xenin-25 induced gallbladder contractions, but not jejunal contractions, in a dose-dependent manner in conscious dogs [24]. Xenin-25 has also been reported to act as an antagonist of NTSR2 and can activate NTSR3 with very low affinity [25]. These reports suggest that the effects of xenin-25 on the intestine may not be mediated entirely through interactions with NT receptors. Thus, the details of the biological domain inducing various physiological actions of xenin-25 are not fully understood. Similar to other peptides, xenin-25 action requires an initial binding step to specific receptors. To improve the mechanistic understanding of xenin-25 and to gain additional insight into the functions of xenin-25, we were interested in obtaining the minimal biological domain of xenin-25 required to induce ion transport in the intestine. Thus, the present study was designed to determine the minimal biological domain of xenin-25 necessary to induce ion transport in the rat ileum using various truncated xenin-25 fragments and analogues.

Animals and tissue preparation
Male Sprague-Dawley rats weighing 230~300 g (Shimizu Laboratory Supplies, Kyoto, Japan) were fed a pellet diet and water ad libitum. This study was approved by the Committee for Animal Research of Ritsumeikan University (approval no. BKC2019-006). The animals received proper care, and the experiments were conducted according to the guidelines of the committee. Animals were anesthetized with isoflurane and euthanized by terminal exsanguination. The ileum (10~20 cm from the ileocecal valve) was removed and immediately placed in ice-cold Krebs-Ringer solution. The ileum was cut open along the mesenteric border, the mesenteric and fat tissues were removed from the ileum, and the luminal contents were gently removed. Then, the tissues were pinned flat on silicone rubber-lined petri dishes containing Krebs-Ringer solution, and mucosa-submucosal preparations were prepared by removal of the external muscle layers containing the myenteric plexus with fine forceps under a stereomicroscope. The tissue sheets were longitudinally divided into four pieces, and each piece was mounted between the halves of an Ussing flux chamber (cross-sectional area of 0.64 cm 2 ).

Short-circuit current measurements
Muscle-stripped, mucosa-submucosal preparations were used. The mucosal and submucosal surfaces of the tissues were bathed in 10 mL of Krebs-Ringer solution by circulation from a reservoir maintained at 37 • C. The Krebs-Ringer solution contained the following (in mM): 117 NaCl, 4.7 KCl, 1.2 MgCl 2 , 1.2 NaH 2 PO 4 , 25 NaHCO 3 , 2.5 CaCl 2 , and 5 glucose (except the mucosal bathing solution) and was bubbled with 95 % O 2 and 5% CO 2 to maintain a pH of 7.4. The Isc was continuously measured in voltage-clamp mode at zero potential (CEZ-9100; Nihon-Koden, Tokyo, Japan) and recorded on a Power-Laboratory System 4/ 26 (ADInstruments, Cattle Hill, Australia). Positive Isc values indicate a negative electrical charge flux from the serosal to mucosal side as a result of anion secretion or cation absorption. Experiments were performed within 3~4 h after the tissues were mounted on the Ussing chambers, because the mucosa of the small intestine sustains after being mounted on an Ussing chamber for >4 h [22]. Furthermore, the viability of the preparation was confirmed by the addition of carbachol (10 − 5 M) to the serosal bath after the experiment (data not shown). Tissues were stabilized for 40-50 min before the effects of xenin-25 fragments and analogues were examined. Concentration-response curves for xenin-25 and several fragments and analogues were obtained noncumulatively up to 10 -6 M.

Peptide synthesis
Xenin-25, other truncated xenin fragments, the xenin 21-25 analogue, NT and truncated NT fragments were synthesized by a solidphase Fmoc methodology, using an automated peptide synthesizer (Model Pioneer; Thermo Fisher Scientific, Waltham, MA, USA). The crude peptides were purified using reversed-phase HPLC (Delta 600 HPLC system; Waters, Milford, MA, USA) with a Develosil ODS-HG-5 column (2 × 25 cm; Nomura Chemical, Seto, Japan). The purity of each peptide was confirmed using analytical HPLC and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Table 1 illustrates the sequences of xenin-25, NT and their fragments synthesized and used in the present study.

Data analysis and statistics
The results are expressed as the mean ± SE of the mean (SEM; n = number of experiments from different animals) and the median effective concentration (EC 50 values) with 95 % confidence intervals. Concentration-response corves were analyzed by nonlinear regression and were compared using two-way ANOVA followed by Bonferroni's posttest correction. All plotting, calculations, and statistical analyses were performed using GraphPad software (GraphPad 9, San Diego, California, USA). Diff ;erences between groups were considered significant if p < 0.05.

Effects of xenin-25 and its C-terminal fragments on basal Isc in the rat ileum
Basal electrical parameters were measured after stabilization of the basal Isc and conductance (Gt); after 40 min of incubation in the Ussing chamber, the basal Isc and Gt stabilized to 30.4 ± 2.3 μA/cm 2 and 25.7 ± 1.7 mS/cm 2 , respectively, in mucosal-submucosal preparations of the ileum (n =99). Serosal application of the parent peptide, xenin-25, immediately increased the Isc in a concentration-dependent manner with an EC 50 = 1.8 × 10 − 9 M, as reported previously [26] (Fig. 1, Table 2). The maximal response occurred at 10 -7 M and the maximal net increase in Isc after application of 10 -7 M xenin-25 was 102.7 ± 16.5 μA/cm 2 (n = 10). However, mucosal application of xenin-25 had no effect on the basal Isc similar to results from a previous study [26]. The increase in Isc by xenin-25 showed a similar time course as that previously reported; the Isc peaked approximately 30 s after application and then gradually decreased to baseline at approximately 3 min even if the peptide was kept in contact with the tissue (Fig. 1A, xenin 1-25). Fig. 1A shows representative tracings of xenin-25 and its fragments.
As with many gut peptides, xenin-25 secreted from EECs is rapidly metabolized by enzymes in the circulation but the physiological importance of these degradation fragments is still largely unknown [29][30][31]36,38]. Therefore, examining the enzymatic degradation fragment peptides of xenin-25 in plasma is important to define the minimum biological domain of xenin-25 together with the synthesized truncated fragment. The plasma enzymatic degradation fragment peptides of native xenin-25, xenin 9-25, xenin 11-25, xenin 14-25 and xenin 18-25 induced increases in Isc in a concentration-dependent manner similar to that of the parent peptide xenin-25 as shown in Fig. 1A & B. Furthermore, the synthetic fragments xenin 5-25, xenin 20-25 and xenin 21-25 also showed the same action on ion transport in rat ileum as the natural peptide (Table 2). These fragments showed ED 50 and 95 % CI values similar to those of xenin-25 as displayed in Table 2. Because xenin 21-25 was biologically active, shorter fragment, xenin 22-25 was synthesized to define the more precise active domain of xenin-25. The synthesized peptide, xenin 22-25 did not show any effect on basal Isc even at a concentration of 10 − 6 M, as shown in Fig. 1A & B (p < 0.0001, n = 8)

Effects of xenin 21-25 analogues on basal Isc in the rat ileum
Xenin 21-25 showed potency similar to that of the parent peptide xenin-25 but the shorter fragment xenin 22-25 failed to induce an increase in Isc in the rat ileum. Alternations to various amino acid residues, amino acid deletions or substitution of D-amino acids into xenin-25 have been reported to dramatically affect the biological activity of this peptide [9][10][11]17,29,34]. From the present results, we thought that the minimum amino acid sequence of xenin-25 required to induce ion transport might exist on its C-terminal pentapeptide and we hypothesized that the arginine (Arg) residue at position 21 was important to induce an increase in Isc of the rat ileum because xenin 22-25 did not contain Arg. To assess the importance of Arg at position 21, we synthesized xenin 21-25 analogues with single amino acid substitutions at position 21. Arg 21 was substituted with lysine (Lys 21 ), glutamate (Glu 21 ), and D-Arg 21 to test whether Arg 21 was essential to induce ion transport by xenin-25.  Fig. 2A & B). Moreover, these inactive analogues did not show any of the antagonistic effects of xenin-25

Effect of neurotensin and its analogues on the basal Isc in the rat ileum
Xenin-25 and NT share close structural similarity 5 out of the 8 amino acids on their C termini being the same, and only xenin 18-25 (xenin-8) has been reported to possess biological activity. Xenin 18-25 augments GIP-induced insulin secretion in the perfused rat pancreas [36] but xenin 9-25, xenin 11-25 and xenin14-25 had no effect in an in vitro study using clonal pancreatic BRIN-BD11 cells [30]. However, in the present experiments, we showed that xenin 5-25, xenin 9-25, xenin 11-25, xenin 14-24, xenin18-25, xenin20-25, and xenin 21-25 retained their biological activity on ion transport in the rat ileum. The present results suggest that the C-terminal pentapeptide is essential to induce ion transport in the rat ileum. Furthermore, the C-terminal pentapeptide of NT has a similar sequence except for one amino acid reside, Tyr. Thus, we examined whether C-terminal NT fragments affect ion transport similar to that of the C-terminal fragment xenin 21-25 in the rat ileum. Fig. 3A shows representative traces of NT and its shorter C-terminal fragments. Serosal application of the parent peptide NT immediately Table 1 Amino acid sequences of xenin 1-25, the xenin fragments and other neurotensin and their fragments.  Fig. 3B shows the concentration-response curves of NT and its C-terminal fragments. The concentration-response curves of NT and NT 9-13 were nearly identical as shown in Fig. 3B. However, a shorter fragment, NT 10-13, with a sequence similar to that of xenin 22-25 had no effect on the basal Isc in the rat ileum. NT 10-13 (10 -6 M) slightly induced an increase in Isc to 3.7 ± 2.5 μA/cm 2 (n = 3; p = 0.044 vs. NT). In addition, NT 10-13 had no antagonistic activity even at the 10 -6 M. This result suggests that Arg at position 9 of NT is also important to maintain the biological activity of NT, similar to that of xenin-25.

Discussion
To better understand the physiological role of xenin-25, we made various truncated fragments of xenin-25 and it's analogue and found that the smallest peptide that retained significant biological activity to Fig. 1. Effects of xenin 1-25 and its fragments on shortcircuit current (Isc) in mucosa-submucosal preparations of the rat ileum mounted in Ussing chambers. (A) Representative traces of the increases in basal Isc induced by 10 − 7 M xenin 1-25 and its fragments. Serosal application of xenin 1-25 and its fragments, except xenin 22-25 (10 -11 ~ 10 -6 M), induced increases in the Isc, whereas mucosal application of the peptides had no effect. (B) Concentration-response curves of xenin 1-25 and its fragments. The concentrationresponse curves were analyzed by nonlinear regression and compared using two-way ANOVA followed by Bonferroni's posttest correction using Graphpad software (Graphpad 9, San Diego, California, USA). Values are presented as the mean ± SEM (n = 10). The diff ;erence between groups was considered significant if p < 0.05. induce ion transport in the rat ileum was the C-terminal pentapeptide including Arg at position 21.
In the present experiments, xenin-25 and its related fragments induced rapid increases in the basal Isc that were not maintained at a stable plateau and faded even if the peptides were kept in contact with the tissues. This decline was hard to consider to be for tissue fatigue or desensitization of the effect because the integrity of the villi and crypts of ileum is reported to be maintained for 4 h after mounted on the Ussing chambers and basal potential difference (PD) and forskolin-induced PD are maintained up to 4 h [22]. Furthermore, the present experiments were performed within 3− 4 h after the tissues were mounted on the Ussing chamber and viability of the preparations was confirmed by the addition of carbachol after the experiment. Thus, these results suggest that xenin-25 and its fragments may be inactivated by the enzyme Indeed, as with many gut peptides, xenin-25 secreted from EECs has been reported to be rapidly metabolized by serum enzymes in circulation [29][30][31].
The presence of four plasma degradation products of xenin-25 including xenin 9-25, xenin 11-25, xenin 14-25 and xenin 18-25 was confirmed by mass spectrometry [29]. Among these fragments, xenin 9-25, xenin 11-25 and xenin 14-25 have no biological activity, as they do not show either insulinotropic or GIP-potentiating actions in clonal pancreatic BRIN-BD11 cells. Only xenin 18-25 was demonstrated to have insulin-releasing actions [30]. Xenin 18-25-induced insulin release is considered to be a direct action on β cells [36]. In addition, a previous study reported that xenin-25-induced insulin release in BRIN-BD11 cells was not involved in significant changes in membrane potential or increases in intracellular Ca 2+ concentration and cAMP production [38]. These results suggest that xenin-25 does not induce insulin release by binding its receptors on the plasma membrane and does not utilize cAMP as a second messenger system in BRIN-BD11 cells. In in vivo studies, xenin 18-25 stimulates basal insulin release and potentiates the insulin response to glucose in a dose-dependent manner in the perfused rat pancreas [33,36]. An early study showed that a peptide sequence of 16 C-terminal amino acids (xenin 10-25) is necessary to elicit full longitudinal muscle contraction in the rat jejunum, whereas only 6 C-terminal amino acids (xenin 20-25) are enough to induce relaxation of longitudinal smooth muscles in the rat colon but xenin 21-25 had significantly reduced potency on relaxation in vitro [13]. Another study showed that xenin 22-25 failed to induce relaxation of KCl-precontracted ileal longitudinal smooth muscle [6]. Taken together, these results show that the type of responses varies according to species and tissues but xenin 18-25 has significant action on both the pancreas and intestinal smooth muscle under in vitro and in vivo conditions. In the present experiments, xenin-25 and the xenin fragments not including xenin 22-25 induced concentration-dependent increases in the basal Isc in rat ileum. Variation of the length of the molecule in the direction toward the C-terminus showed similar effects on the biological activity of xenin-25 and xenin 21-25 and all larger fragments had similar biological activity. No marked difference could be demonstrated in comparison to the application of the parent peptide xenin-25 (Table 2). However, xenin 22-25 lost activity. Thus, the present results indicate that the minimal biological domain consists of the C-terminal pentapeptides and that the length of this pentapetide is sufficient to induce ion transport in the rat ileum. Previous studies have shown that truncated xenin fragments have antagonistic activity; xenin 9-25, xenin 11-25 and xenin 14-25 significantly inhibited xenin-25 (10 − 6 M)-induced insulin release in  BRIN-BD11 cells [30], but the same truncated fragments and xenin 22-25 did not show any antagonistic activity on ion transport in the rat ileum in the present experiment. Further studies are needed to show this discrepancy because BRIN-BD11 cells might use different mechanisms to induce insulin secretion as mentioned above.
Recent studies have shown the modification of C-terminal fragment (ψ-xenin-6) or hybrid type peptide ((D-Ala2) GIP/xenin-8-Gln)) are still retain the biological activity of xenin-25; these peptides are enzymatically stable and have prolonged insulinotropic action [9][10][11]35]. These studies indicate that appropriate modification of the C-terminal fragments to improve enzymatic degradation did not affect their biological activity. Therefore, these modification peptides may be useful tool for diabetic management. Furthermore, these modification fragments may be valuable tool to explore new biological action of Xenin-25 in the body because Xenin has more broad biological activity than thought previously [8].
In the present study, we used a mucosa-submucosal preparation to identify the minimal biological domain to induce ion transport in the rat ileum with an Ussing chamber system. Using this same system, we recently showed that xenin-25-induced Cl − /HCO 3 − secretion in the rat ileum is involved in NTSR1 activation on intrinsic afferent neurons followed by the release of substance P and subsequent activation of neurokinin receptor 1 (NK1) expressed on noncholinergic VIP secretomotor neurons [26]. This result is further supported by the following observation that intravenous infusion of xenin-25 evoked an increase in VIP and pancreatic peptide YY in dogs [14]. In addition, we showed that NTSR1 immunoreactivity colocalizes with NK1 in intrinsic primary afferent neurons in the submucosal plexus. Combining the present results with our functional studies, it can be speculated that xenin 21-25 retains the minimum biological activity of xenin-25 and can recognize NTSR1 on enteric neurons to induce Cl-/HCO 3 -secretion in the rat ileum. Thus, these results indicate that the minimum biological domain of xenin-25 required to exert ion transport in the rat ileum is the C-terminal pentapeptide (Arg-Pro-Trp-Ile-Leu) of xenin-25 and that the integrity of the C-terminal sequence of xenin-25 is sufficient to induce Cl-/HCO 3 -secretion in the rat ileum. This result is further supported by the observation that the xenin 22-25 did not show any biological activity to stimulate ion transport. Since xenin 22-25 lost its biological activity to induce Cl − /HCO 3 − secretion in the rat ileum, we considered that Arg at position 21 was important to retain the biological activity of xenin-25. Therefore, we substituted the amino acids at this position with other amino acids although the study was not designed to examine the detailed function of this residue. Replacement of Arg 21 by Lys 21 retained the biological activity of xenin 21-25 and both of these amino acids are basic. In addition, the peptide that substituted D-Arg for Arg at position 21 showed similar activity to that of xenin 21-25. On the other hand, substitution of Arg by Glu at position 21 lost the biological activity. Glutamic acid is an acidic amino acid. Thus, these results indicate that basic amino acids are important for inducing ion transport in the rat ileum. However, further studies are needed to evaluate the importance of the amino acid residue at position 21.
We previously showed that the specific NTSR1 antagonist SR48692 antagonizes xenin-25-induced Cl − /HCO 3 − secretion in the rat ileum [26]. Thus, we speculated that xenin 21-25 could recognize NTSR1 in a manner similar to the parent peptide xenin-25 since xenin-25 is structurally related to NT [3,16]. SR48692 inhibited xenin 21-25-induced Cl − /HCO 3 − secretion. Furthermore, SR48692 inhibited the increases in basal Isc induced by NT 9-13 to approximately the same degree as xenin 21-25. These results suggest that the truncated fragment xenin 21-25 may still contain the receptor recognition site of xenin-25, which is similar to that of NT. In addition, NT and NT 9-13 retained their biological activity on the ion transport in the ileum but NT 10-13 lost the biological activity. These results suggest that Arg at position 9 of NT is important to maintain the biological activity of NT, similar to that of xenin-25.
In conclusion, 12 truncated neurotensin/xenin family peptides (9 truncated xenin 1-25 peptides and 3 truncated neurotensin peptides) were systematically investigated to identify the minimum biological domain required for Cl − /HCO 3 − secretion in the rat ileum. The present results show that the minimum biological domain is the C-terminal pentapeptide and Arg at position 21 seems to be important to recognize NTSR1 and induce Cl − /HCO 3 − secretion in the rat ileum. These data provide the basis for understanding the physiological function of xenin-25.

Author contributions
YK and AK conceived the study, designed the experiments, performed the experiments and analyzed data. YK and AK drafted the manuscript. YK and AK revised the manuscript. IK made peptide to use experiments. All authors approved the final manuscript.

Declaration of Competing Interest
The authors report no declarations of interest.