Changes of TLQP-21 in Response to Glucose Load

The TLQP-21 peptide peripherally potentiates glucose-stimulated insulin secretion. The aim of this study was to investigate a possible endocrine mechanism through which TLQP-21 increases the insulin secretion. Using an antibody specific for the common N-terminal portion of the TLQP peptides, we studied pancreas and plasma of mice subjected to intraperitoneal glucose load, by immunohistochemistry and immunosorbent assay (ELISA), alone or coupled to High Performance Liquid Chromatography (HLPC). Mice underwent a period of starvation hence have received a glucose load, or saline, and were sacrificed 30 or 120 minutes later. In normal endocrine pancreas, the TLQP-antiserum stained either peripheral or central cells. Interestingly, 30 min after a glucose load, TLQP immunostaining was disappeared in pancreas and, when analysed by ELISA, the TLQP-levels started to increase in plasma reaching peak concentration at 120 min. (controls vs. 30 and 120 min.: p<0.05 and p<0.001, respectively). The analysis of plasma and pancreas extracts using HLPC coupled to ELISA demonstrated the presence of the TLQP-21, with statistically significant increase of this peptide in plasma at 120 min (vs. controls p<0.05) in agreement to the changes seen by measuring the totality of the TLQP peptides. In pancreas sections we found the presence of the C3a-R1, involved in insulin secretion, and previously identified as a TLQP-21 receptor. Hence, after a glucose load, TLQP-21, may be released by the pancreas into the plasma, returning to the pancreas in order to modulate the insulin secretion through the C3a-R1.


INTRODUCTION
Insulin secretion includes a pattern of paracrine, autocrine, endocrine, and autonomic neural signaling mechanisms. The VGF precursor/proprotein (617 or 615 amino acids in rats/mice and humans, respectively) [1], codified from vgf gene (no acronym name), 5 contains a large amount prohormone convertases cleavage sites PC1/3 and PC2 [2], from which a number of peptides of different molecular weight (MW) are derived [3], including the VGF derived peptide TLQP-21 (VGF 556-564 ). The C-terminal internal fragment TLQP-21 was initially studied for its role in energy balance [4 -7]. The TLQP-21 is expressed in the pancreas [8,9] and it is able to potentiate the glucose stimulated insulin secretion, in both 10 human and rat pancreatic islets [10,11]. When peripherally injected into rats, this peptide is able to reduce the blood-glucose peak (after about 20 min. from the bolus ingestion) and to increase the plasma insulin levels. Moreover, it preserves the islet mass and slows the onset of diabetes in Zucker Diabetic Fatty rats. Overexpression of VGF in primary rat islets resulted in a 46% increase in glucose-stimulated insulin secretion, without affecting basal 15 insulin secretion [11]. The TLQP-21 action mechanisms are not clarified yet, but two different G-coupled protein receptors have been discovered as TLQP-21 receptors. The complement component 3a receptor (C3a-R1) [12 -14] is involved with the TLQP-21 in modulating the lipolysis [14], while the receptor for the globular heads of C1q (gC1q-R) acts together with the TLQP-21 as a modulator of neuropathic pain [15]. The modality of 20 action with which TLQP-21 acts in potentiating insulin secretion has not been established yet. Hence, the aim of the present study was to investigate the possible endocrine activity of TLQP-21 in response to glucose stimuli. So, we analyzed possible alterations of the TLQP-21 in both plasma and endocrine pancreas through immunohistochemistry (IHC) and immunosorbent assay (ELISA) alone or coupled to High Performance Liquid 25 Chromatography (HLPC).

Animals and tissue samples
Male CD1 mice (Charles River, Lecco, Italy; aged 12 weeks, weight 35-45 g) have been housed (6 animals/cage) at the enclosure of the University of Cagliari under controlled 5 temperature (23  2°C), light (12-12h; light phase: from 8am to 8pm and dark phase: from 8pm to 8am) and relative humidity (60  10 %) conditions. Mice underwent a period of starvation 18h before the glucose tolerance test and the sacrifice. The animals have been kept with free access to water. Mice (n=12 per group) have been injected with glucose (3g/kg) or saline and sacrificed 30 or 120 min. after. Immediately before the sacrifice, a 10 peripheral venous blood sample (from tail vein) was obtained for the measurement of blood glucose level (mg/ml, MultiCare-in, Biochemical System, S.r.l, Arezzo, Italy). For the sacrifice, mice have been anesthetized (under gas anesthesia induced by isofluran, IsoFlo, Italy) and blood (100-200µl) has been drawn from the left heart with syringes pretreated with ethylenediaminetetraacetic acid disodium salt (EDTA; 1.78 mg/ml). The blood 15 samples collected have been drawn into tubes, centrifuged (11,000 rpm; 10 min.) and frozen until use. After blood collection, the perfusion has been performed with 40 g/L paraformaldehyde (40 g/L in 0.1 mol/L PO 4 , 15 min.) injected into the left ventricle. Fixed pancreas samples have been extracted from the body and rinsed in phosphate buffer saline (PBS: 0.01 mol/L PO 4 , pH 7.2, 0.15 mol/L NaCl) containing 70g/L sucrose and 20 0.1g/L NaN 3 , treated with cryoembedding media (PVA 56-98 59 g/L, Tween-20 10 g/L, and Peg-400 40 g/L in PBS-NaN 3 1 g/L) [16] and subsequently frozen with liquid nitrogen.

TLQP antiserum
The guinea-pig primary antiserum against TLQP peptides specific for their common Nterminal portion, previously described in detail [18] was extensively used in different rat organs and tissues [8, 19 -21]. Briefly, a synthetic peptide corresponding to rat VGF 556-564, with the addition of a C-terminal cysteine residue, was conjugated via its C-terminus 15 to keyhole limpet hemocyanin (KLH), and used for immunizations.

IHC and image analysis
Pancreas sections have been used to perform IHC. After a first treatment with Triton X-100 (0.1% in PBS 1x for 45 min.), the sections have been washed with PBS and incubated overnight with the guinea-pig TLQP antiserum diluted with PBS containing 30ml/L of 20 normal donkey serum, 30ml/L of normal mouse serum and 0.02g/L NaN 3 . Guineapig/mouse insulin, rabbit glucagon (concession from Professors J.M. Polak and S.R. Bloom), rabbit somatostatin (AB_1143012), rabbit gC1q-R (AB_10675815) and C3a-R1 (AB_2687440) antibodies have been also used either single or in double staining (Table 1) while the relevant species-specific donkey secondary antibody/ies, conjugated with either  10 the optical density has been measured at 450nm using a multilabel plate reader (Chameleon: Hidex, Turku, Finland). Recovery of synthetic peptide (same used for immunization, plate coating and measurement standard) added to plasma, or to tissue samples at extraction was >85%. TLQP assay has been previously characterized [22] using multiple synthetic peptides, with IC 50 (pmol/ml) = 1.1; CV1=4% and CV2=6%.
Instead, the C3a-R1 staining is well represented in the entire islet, even if more represented in the central cells (Fig. 3d), hence only partially colocalized with TLQPp ( Fig.   3e,f). The staining of the two receptors seems comparable between normal and glucose-15 treated animals (not shown). Raw data are included in S1_Fig.tif.

Reverse-phase HPLC
Upon HPLC using pooled pancreas extracts from the control group (Fig. 5), TLQP

DISCUSSION
At 30 min. after the glucose load, when the glycaemic peak is reached, the immunoreactivity of the entire population of TLQPp is scarcely detectable in the pancreas, concurrent with a TLQP-21 increase in plasma reaching a high peak at 120 min. Our results are in agreement with the role of TLQP-21 as hypoglycemic agent [11]. Indeed, 10 TLQP-21 peripherally injected into rats, is able to reduce the blood-glucose peak (after about 20 min. from the bolus ingestion) and to increase the plasma insulin levels [11].
Obviously, TLQP-21 peptide in plasma can also originate from different organs/tissues. In fact, in addition to pancreatic islets [8,9], TLQP-21 can also be released by the adrenal gland [19], stomach [20], pituitary-ovary axis [24], as well as (albeit less expected) median 15 eminence [24]. Nonetheless, after a glucose load, the massive pancreatic degranulation of TLQPp at 30 min. in parallel with the starting of the TLQP-21 increase in plasma, should be in favor of considering the endocrine pancreas as the origin of at least part of the TLQP-21 amount in plasma. At 120 min., although the TLQP-21 return to be immunoreactive in the pancreas, it still increased in the plasma, as previously reported [9].
20 When plasma stability was measured, it was reported that TLQP-21 has a terminal half-life of about 110 min, hence roughly comprising our temporal range of measurements [25].
The major plasmatic increase of the TLQP-21 at 120 min rather than at 30min at the glycaemic peak is not clear, however we may hypothesize a modulatory activity of TLQP-21 at a late-phase of glucose-induced insulin secretion. The restore of the production in 25 the pancreas could be compatible with a hypothetical re-accumulation phase, correlated to the de novo biosynthesis of TLQP-21. We have also found the presence of the two TLQP-21 receptors (gC1q-R and C3a-R1) in pancreatic cells. While limited information is so far available regarding the pancreas immunolocalization of gC1q-R [26], C3aR1 is known to be expressed in human and mouse islets, within β-and α-cells [27] and to potentiate 5 glucose-induced insulin secretion from human and mouse islets [27]. Furthermore, C3a-R1, compared to gC1q-R, has been more studied as TLQP-21 receptor [13,28,29] and has been characterized in its precise reactivity depending on the C-terminal sequence of the TLQP-21 peptide [13]. We can hypothesize that TLQP-21, by returning to the pancreas from the plasma, could interact with the C3a-R1, in order to modulate the insulin secretion.
10 Other hormones are known to act in this way, for example the gastrin peptide released from the stomach into the bloodstream, returns to the stomach in order to stimulate secretion and motility. In agreement with our data, targeted deletion of VGF in mice [30,31] decreases islet cell mass, and reduces circulating insulin levels. Moreover, patients with obesity and Type 2 Diabetes (T2D) showed an altered TLQP response compared to 15 normal weight subjects, while the chronic injection of TLQP-21 in pre-diabetic rats preserves islet mass and slows T2D onset [11]. In conclusion, we can speculate that, after glucose stimuli, TLQP-21 could modulate the insulin secretion through an endocrine activity.