The inverse relationship between fibronectin and neuroligins in the embryonic rat colon

Our previous study identified that the abnormal expression of fibronectin (FN), neuroligin-1 (NL1) and neuroligin-2 (NL2) in the colons of children with Hirschsprung disease (HSCR), but the correlated relationship between the three in the development of the enteric nervous system (ENS) remains unclear. Colons of Wistar rats and PC12 neurons were used to investigate the relationship between FN and the neuroligins (NLs). Colon tissues from thirty healthy embryonic rats, including fifteen at embryonic day 16 (E16), eight at E18, seven at E20, and fifteen newborn rats within 24 hours (Ep0) were analyzed to determine the correlated expression of FN and NLs using Western blot (WB) analysis and real-time fluorescence quantitative PCR (qRT-PCR) methods. Small interfering ribonucleic acid (siRNA) targeting and gene plasmids were used to explore the functional interaction between FN and NLs by using PC12 neuron cells. Furthermore, we used recombinant FN and NL proteins to confirm their interactions. Our studies showed that there were downregulatory effects between FN and the NLs in the embryonic rat colon and different cell lines, indicating that FN and NLs could directly regulate each other, and there is a negative linear correlation between them. The imbalanced interaction between extracellular matrix and synapse-related genes may provide a new perspective for the pathogenesis and treatment of HSCR and neuronal intestinal malformations (NIMs).


Introduction
Neuroligins (NLs), which are expressed predominantly at the postsynaptic terminal, interact with neurexins (NXs) across the synaptic cleft and play critical roles in proper synapse maturation, function and neuronal function. [1] Limited research has shown that NLs may be expressed in the enteric nervous system (ENS). The extracellular matrix (ECM) is secreted by cells and composed of proteins and glycosaminoglycans (GAGs). [2] ECM constitutes the cell microenvironment and plays critical roles in guiding neighboring cell behaviors, such as migration, shape, survival, differentiation, and proliferation. [3][4][5][6][7][8][9] Fibronectin (FN) is one of the most vital ECM glycoproteins produced by mesenchymal cells in the colon and plays a crucial role in modulating the neural crest cell response. [10,11] ECM homeostasis imbalance might affect the microenvironment surrounding enteric neural crest cells (eNCCs), which may be responsible for innervation deficiency due to incomplete colonization of eNCCs in the gut in patients with Hirschsprung disease. [12] Hirschsprung disease (HSCR) is attributable to failed migration of neural crest cells in the distal bowel and extending proximally for varying distances from the anus. HSCR causes a life-threatening bowel obstruction in approximately 1 in 5000 live births. 1 The earlier the arrest of the migration is, the longer the aganglionic segment. We have found a temporal trend in the abnormal expression of FN, neuroligin-1 (NL1) and neuroligin-2 (NL2) in the diseased intestine of children with HSCR. [13,14] These pieces of evidence led us to hypothesize that the abnormal development of FN, NL1 and NL2 may lead to intestinal lesions but does not explicitly indicate whether there is a real interaction between FN and NLs. Therefore, we further investigated the expression of FN, NL1 and NL2 on different embryonic days and the relationship between them and confirmed the interaction between FN and NLs in this study. This work may contribute to the understanding of mechanistic research and confirms that synapse-related genes interact with the ECM microenvironment.

Animals
Thirty healthy Wistar rats at embryonic days 16 (E16), 18 (E18), 20 (E20) and fifteen newborns within 24 hours (Ep0) were used in this study. All Wistar rats were purchased from the animal facility of Shandong University. All procedures were approved by the local ethics committee and were performed at Qilu Hospital, Shandong University. Rats were anesthetized with chloral hydrate and then sacrificed by cervical dislocation.
Segments of the full thickness colon, approximately 1.0 cm distal to the ileocecal junction, were harvested. The entire surgery was performed on ice, and the colon specimens were stored at -80°C.

Cell culture, transfection, and treatments
293T cells (a human embryonic renal epithelial cell line that is widely used in cell biology for their reliable growth and propensity for transfection) [15]

Antibodies and reagents
Detailed information on the antibodies and primers is listed in Tables  analyzed by the 2 −ΔΔCt method. [17] The 0 μg/mL group served as a negative control, and 2 −Δ Δ Ct was calculated as the relative expression for further analysis. The expression of GAPDH in each sample was used as an internal control.

Statistical analyses
All data shown in this study were analyzed with Graph Pad Prism® 5 software (La Jolla, CA, USA) and excel. One-way analysis of variance (ANOVA) with Tukey's multiple comparison test was used for comparisons among three or four groups, and unpaired t-tests were used for comparisons between two groups. The original index data were standardized by Zero-mena normalization. All results are expressed as the mean values (±SD), and values < 0.05 were considered statistically significant.

Results
We have repeated our previous studies [13,14] and confirmed that gene dysplasia of NLs and FN are associated with HSCR. HSCR is divided into three segments: aganglionic segments, transitional segments and normal tissue. NL1 and NL2 expression gradually increased in the three sections, while FN showed the opposite effect (data not shown). Previous studies suggest that there must be a relationship among the three, and the imbalance in their expression must be associated with HSCR.

Downregulation of FN and the correlated expression of NLs in the enteric nervous system of embryonic rats
A WB assay was performed to investigate the expression of FN, NL1 and NL2 in the later embryonic stage with the aim of confirming the temporal increasing trend of NL1 and NL2 expression and the temporal decreasing trend of FN expression. Compared with the expression levels in E16 rats, the protein expression of NL1 and NL2 at E18, E20, and Ep0 were significantly increased, while the expression levels of FN were significantly decreased (P<0.05) (Fig 1A and B). And the results were consistent with those of the q-PCR analysis (P<0.05) (Fig 1C). Fig 1 reveals a temporal   were statistically significant (P< 0.05). d P <0.05 versus E16, e P < 0.05 versus E18, and f P < 0.05 versus E20.

Suppression of FN can promote NL expression
To confirm the direct relationship between FN and NLs, we transfected PC12 cells with NL1, NL2 and FN siRNA, referred to as knockdown (KD) lines of NL1-KD, NL2-KD and FN-KD, respectively. Then, the effect of genetic deletion of FN on NL1 and NL2 was detected by transfecting siRNA into PC12 cells. As shown in Fig 2C and 3C Moreover, the levels of NL1 and NL2 (Fig 2A and B) were significantly higher in the FN-KD group than that in the NC and control (Ctr) groups , while the FN was the opposite, indicating that FN inhibits the expression of NL1 and NL2 (P<0.05).

Suppression of NL1 and NL2 can promote FN expression
As shown in

Negatively correlated relationship between FN and NLs in 293T cells
Our previous studies on HSCR have proved that FN expression in aganglionic segments, transitional segments and ganglionic segments showed a diminishing trend, while the expression of NL1 and NL2 was just the opposite. [18] Then, FN1 recombinant proteins (FN1rp), NL1 recombinant proteins (NL1rp) and NL2 recombinant proteins (NL2rp) were used to examine whether there was a linear relationship among FN, NL1 and NL2.
High-level expression models of FN and NLs were produced by recombinant proteins, and their effects on the expression of NLs and FN were observed.

WB and RT-PCR were performed in 293T cells to measure the expression of FN and
NLs for exploring the putative relationship between them. To investigate the optimal stimulation time, 20 μg/mL FN and 0.2 μM NLsrp were selected to stimulate for 0, 0.5, 1, or 4 hours. As depicted in Fig S1 and S2, the results demonstrate that the inhibitory effect of FN and NLs was the best when stimulated for 0.5 hours, which was selected as the basis for later experiments.
The abundance of transcripts encoding FN was significantly lower under each increasing concentration group, which means the relative expression of FN decreased as the recombinant protein concentrations increased (0 μM > 0.2 μM > 1 μM > 5 μM) in both NL1-FN and NL2-FN groups ( Fig 4F).

Discussion
To date, 15 genes represented by RET have been implicated in HSCR development, only approximately 30% of which have mutations, suggesting the involvement of other genes or the abnormal ECM. [7,[19][20][21][22] Currently, it is widely accepted that the surrounding microenvironment is closely related to the corresponding nervous system, such as that in the intestinal tract. NL2 is similar in structure and sequence with NL1, [23] and it has been confirmed that NL1 and NL2 have similar activities in vitro in mice, [24] while NL1 is only expressed at excitatory synapses and NL2 at inhibitory synapses in vivo. [25] 293T cells, a human embryonic renal epithelial cell line, is widely used in cell biology for their reliable growth and propensity for transfection and the rat PC12 cell line was widely used in studies of neuronal disease , such as Alzheimer's disease, Huntington's disease and HSCR. [26,27]  NLs plays an important role in ENS morphogenesis and functional maturity.
Moreover, there is indeed a certain regulatory relationship between the surrounding microenvironment and the ENS. In addition, our current studies have confirmed that there is a negative dose-dependent correlation between the expression of FN and NLs.
Therefore, our further studies will focus on whether the in vitro induction of FN and NL expression can improve or cure the effects of HSCR in mouse models. We may be able to use the optimal physiological dose to achieve a cure or reduction in the symptoms of NIMs.