Mapping the effect of the antisecretory factor on GABAA receptor α1 and α6 subunits in cerebellar granule cells in vitro

The Antisecretory Factor (AF) is a protein that can reduce intestinal hypersecretion and various inflammation disorders in vivo. Discovered in many mammalian tissues and plasma, its mechanism of action remains unknown. Interestingly, its induction has been found to counteract vertigo in patients with Méniere's disease. This suggests an inherent ability to control body balance and posture, an activity that may play a role in cerebellar function. Therefore, it may be worthwhile to investigate whether this activity can inhibit neuronal cells involved in cerebellar circuitries and its potential action on enteric nervous system ganglia, which could explain its antisecretory effect in the intestine. Previously, we studied the role of AF on GABAA receptors in cerebellar granule cells, taking advantage of electrophysiology and evaluating the effects of the administration of AF-16, an AF peptide. Treatment with AF-16 increased GABAA receptor responses, especially those containing the α6 subunit. Here, we performed immunofluorescence experiments by staining α1 and α6 subunits before and after incubation with AF-16, analyzed super-resolved images comparing pre- and post-treatment maps and critically examined these experimental results with our previous electrophysiological data to shed light on the mechanisms of action of AF protein on GABAA receptor subpopulations, specifically the "fast" receptors of αn β2/3 γ2 composition that contain either the α1 or the α6 subunit. The results indicate that the α6 subunit is redistributed, with a decrease in neurites and an increase in soma. Conversely, the α1 subunit shows opposite results, with an increase in neurites and a decrease in soma.


Introduction
The antisecretory factor (AF) is a 41 kDa protein discovered, purified, and cloned in the '80 s by Lange and Lönnroth (Lange and Lönnroth, 2001).
Although AF can be found in many tissues and organs -respiratory organs, gastrointestinal tract and gallbladder, kidneys, ureters and urinary bladder, pituitary gland, and central nervous system (Lange and Lönnroth, 2001) -its mechanism of action remains unclear.
These pathological conditions arise from an imbalance of ions and water transport across the plasma membrane.However, to gain a more detailed understanding of AF's mechanism of action, it is necessary to determine its effect on cellular physiology through in vitro studies.
Studies have shown that AF functions through neuronal pathways (Lange et al., 1985;1987;Rapallino et al., 1989;2003), and its role in GABAergic transmission in the central nervous system (CNS) has been widely investigated in recent years (Kim et al., 2005;Strandberg et al., 2014;Bazzurro et al., 2018;Bazzurro et al., 2022).GABA is the principal inhibitory neurotransmitter in the mammalian CNS, acting via GABA A and GABA B receptors.
Our previous experiments using the patch-clamp technique demonstrated that the 16 amino acid AF-16 peptide (a fragment of the AF Nterminal end that contains the active site and is resistant to proteolytic degradation, Dzebo et al., 2014) affects GABA A receptors expressed on the plasma membrane in cerebellar granule cells (Bazzurro et al., 2018;2022).After AF-16 treatment, the "peak" current induced by GABA increased.The effect was dependent on the concentration of the antisecretory factor and the incubation time, with the effect on GABA A receptors rising as the concentration and the pretreatment time increased until reaching saturation (Bazzurro et al., 2018).
Previously, we investigated the action of AF-16 by electrophysiological experiments based on two-photon GABA uncaging technique (Cozzolino et al., 2020) and observed the different effects in various zones of interest (soma, axon initial segment, neurite).Using furosemide, an α 6 -containing GABA A receptor antagonist, the results showed that the action of AF-16 in the neuron cell body is almost entirely blocked, suggesting the involvement of a specific GABA A receptor population composed of α 6 subunit (Bazzurro et al., 2022).These results are interesting, especially when considering the significant impact of AF production stimulation in the body, for instance, the decrease of vertigo in Meniere's syndrome patients (Hanner et al., 2004;2010).Cerebellar granule cells are part of a circuitry that controls balance and posture by innervating cerebellar Purkinje cells.These cells, in turn, control the neurons of the lateral vestibular nucleus via a GABA-mediated input (Ito et al., 1966;Ito, 1984;Hydén et al., 2000).
This neuronal model can help understand the potential mechanisms underlying AF's role in the nervous system.It can also provide insights into the possibility that action on GABA A receptors of the ENS might mediate its antisecretory activity in the intestine (Auteri et al., 2015;Poulter et al., 1999;Seifi et al., 2014;Seifi, Swinny, 2019).
Thus, in this study, we aimed to investigate the action of AF-16 using optical microscopy and to correlate it with the results obtained with electrophysiology.Both non-permeabilized and permeabilized cells were investigated to distinguish membrane-associated GABA A receptors from those in the cell's cytoplasm.
We focused on the role of the protein AF on GABA A receptors and evaluated the variations of fluorescence intensity in control and AF-16 treated samples.
We estimated the fluorescence by measuring the fluorescence integrated density of α 6 and α 1 containing GABA A receptors by comparing, 1-hour AF-16 pretreatment, samples with control untreated cells.We used indirect immunostaining and tridimensional-stimulated emission depletion (3D-STED)-based super-resolution microscopy to assess the fluorescence.
We compared the fluorescence integrated density of α 6 and α 1 -containing GABA A receptors in 1-hour AF-16 pretreated samples with control untreated cells.Our data showed a different behavior of α 6 and α 1 containing GABA A receptors in the neurons' soma and neurites.The α 6 GABA A subtypes exhibited an increase in fluorescence, after AF incubation, in the cell body membranes and a reduction in neurites' membranes.In contrast, the α 1 GABA A subtypes presented a decreased intensity in the soma and increase of fluorescence in the neuron neurites' cytoplasm compared to the control neurons.These results obtained through fluorescence methods confirmed our previous assumption, tested with electrophysiology, that AF acts on a specific population of GABA A receptors.In turn, this suggests ideas about how the factor can influence cerebellar function and prevent vertigo in Méniere's disease, and the possibility of an action on neurons in gastrointestinal tract ganglia.

AF-16 peptide production
The AF-16 peptide composed of antisecretory factor (AF) amino acid sequences 36-51 (VCHSKTRSNPENNVGL) was synthesized with solidphase synthesis by Ross-Petersen AS (Copenhagen, Denmark), as reported by Rapallino et al. (2003).AF-16 has been used as an adequate molecule to study the activity of the whole-size AF, since it comprises the active site and is resistant to proteolytic degradation (Dzebo et al., 2014).

Animals
The Department of Pharmacy, Section of Pharmacology and Toxicology of the University of Genoa, housed Sprague-Dawley rats.These animals were treated in accordance with the E.U.Parliament and Council Directive of September 22nd, 2010 (2010/63/E.U.).The Italian Ministry of Health approved their use (COD.75F11.N.6DX) D.M. 116/ 1992.All necessary measures were taken to minimize animal suffering and reduce the number of animals used.

Cerebellar granule cell primary cultures
Cerebellar granule cells (CGC) were obtained from male and female 6-8 day old Sprague-Dawley rats using methods previously described (Robello et al., 1993).
To prevent glial cell proliferation, 10 μM cytosine arabinoside (Sigma-Aldrich, St. Louis, MO, USA) was added to the culture medium 18-24 hours after plating, and the medium was refreshed after 48 hours with the addition of 10 μM cytosine arabinoside.
The experiments were conducted on the 6th-10th day in vitro.

Immunolabeling and 3D-STED microscopy
For the immunolabeling experiments, control cells and cells that were incubated with 1.0 µM AF-16 for 1 hour at 37 • C were used.
The neurons were fixed in 4 % paraformaldehyde in PBS (Sigma-Aldrich, St. Louis, MO, USA) for 15 minutes at room temperature and washed three times with PBS.Some control and AF-16 pretreated samples were permeabilized with 0.1 % Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 5 minutes.All the samples were then blocked with 2 % BSA (bovine serum albumin) in PBS for 30 minutes to prevent non-specific binding and were incubated overnight at 4 • C with primary antibodies.
The primary antibodies used were as follows: -Polyclonal rabbit antibody against rat GABA A receptor α 6 subunit (Alomone Labs Ltd., Israel) at a concentration of 1:500 in 2 % BSA; V. Bazzurro et al.
-Polyclonal rabbit antibody against rat GABA A receptor α 1 subunit (Alomone Labs Ltd., Israel) at a concentration of 1:500 in 2 % BSA.The cells were washed three times with PBS and incubated with secondary antibodies for 45 minutes at room temperature.
The samples were washed three times with PBS and mounted in Mowiol® (Sigma-Aldrich, St. Louis, MO, USA).
The images were captured using a Stellaris 8 Falcon TauSTED (Leica Microsystems, Mannheim, Germany) inverted microscope.A white light laser provided the excitation wavelengths with a notch filter set at 561 nm and a STED beam laser with a wavelength of 775 nm.An avalanche photodiode was used as a detector in the 571-700 nm range.The samples were imaged through a plan-apochromatic oil immersion objective 100×/1.40NA. 3D Z-stacks (512 × 512 pixels, 16-bit) were acquired with a Z-step size of 0.07 µm, setting the pinhole at 1.0 Airy size, speed of 200 Hz, and field of view of 47.68 μm × 47.68 μm.
Image processing and statistical analysis Z-stacks were analyzed using the software Fiji.We evaluated the sum of the fluorescence intensity of each single frame of the 3D stack using the Z-projection function.After subtracting the image background, we determined the fluorescence intensity by drawing a region of interest (ROI) around every cell's profile (soma or neurite) (see Fig. 1).
The obtained values of the ROI integrated density were normalized, taking into consideration the cell volume of each neuron.Then, we calculated the weighted average of the integrated density/volume ratio for every 3D acquisition.
Finally, we assessed the effect of AF-16 by calculating the percentage effect with the following formula: The values used in the final calculations (I AF-16 and I CTRL ) were the means across different evaluations in the various sets of cells according to the treatment (AF-16 or controls), permeabilization or notpermeabilization, and α 6 or α 1 subunit, cell soma or neurites: overall, sixteen different sets.To assess the statistical significance of E% comparisons in the various instances, the E% (see formula above) SEM's were evaluated as described in the following paragraph.
Data are reported as mean ± SEM.Error propagation was used to calculate SEM for E% in all comparisons with the controls.In the comparison of the various pairs of E% values, we employed the Student's ttest.Statistical significance was set at p < 0.05.

Results
We used a concentration of 1.0 μM AF-16 to test the effect of AF on cerebellar granule cells.This concentration allowed us to work with the lowest concentration, while still having the maximum effect as demonstrated in our previous work's dose-response curve (Bazzurro et al., 2018).
We then used immunofluorescence microscopy approaches to investigate the impact of AF-16 on two different subpopulations of GABA A receptors, α 1 and α 6 -containing GABA A Rs, in both nonpermeabilized and permeabilized cerebellar granule cultures.
The results of the immunofluorescence tests showed a different behavior for α 1 and α 6 -containing GABA A receptors.
We also carried out experiments in samples permeabilized with 0.1 % Triton X-100 to compare the results with the non-permeabilized ones.
Tables 1 and 2 summarize the results presented as histograms in Figs. 2 and 3. Table 3 shows that in all comparisons (either for nonpermeabilized or permeabilized cells; either α 6 or α 1 subunit) between soma and neurites, the effect (E%) of AF-16 was different.In particular, the effect on the subunit density for α 6 was an increase in the soma statistically different from the (decreased) value in the neurites, both for permeabilized (referring to cytoplasm plus plasma membranes) and non-permeabilized (referring only to the plasma membranes) CGC neurons.The contrary applies to the α 1 subunit.Overall, this demonstrates an opposite trafficking of the two subunits, induced by the action of AF-16.From neurites to soma for α 6 subunits, vice versa for the α 1 ones.
Table 4 shows the comparison of cells cytoplasm plus plasma membranes with only plasma membrane subunit density in soma or neurites separately.Only for α 6 and in neurites, the subunit density change was different in non-permeabilized (only plasma membranes) vs. permeabilized (cytoplasm plus plasma membranes) neurons.The results show that for α 6, its decrease in plasma membranes is significantly greater than in cell cytoplasm.
V. Bazzurro et al. permeabilized CGC cells, which are shown in Figs. 4 and 5, respectively.

Discussion
The Antisecretory Factor was discovered by Lönnroth and Lange in 1985.Since then, it has been characterized, cloned, and sequenced (Johansson et al., 1995) and has been found to be present in several organs, in addition to the pituitary and intestine (Lange et al., 1999).Originally, it was indicated and studied as an endogenous protein that could counteract intestinal hypersecretion (Lönnroth and Lange, 1985;Lange and Lönnroth, 2001) and has been successfully used in clinical settings to counteract pathologies resulting in diarrhea (Björck et al., 2000;Eriksson et al., 2003a;Eriksson et al., 2003b;Laurenius et al., 2003;Zaman et al., 2018).The idea that it could also play a role in regulating water in body fluids led to trials in pathologies such as Méniere's disease, where hypersecretion of endolymph in the inner ear is thought to be the cause of the disease (Gibson, 2010;Brown et al., 2013;Mohseni-Dargah et al., 2023).In these clinical studies, the major therapeutic result was a decrease in vertigo symptoms (Hanner et al., 2004;2010).However, a possible mechanism of action of AF via the nervous system has been proposed (Lange and Lönnroth, 2001;Grøndahl et al., 2002).For such reasons, studies have been conducted to investigate the interaction of AF with the physiology of brain nervous system structures (Kim et al., 2005;Strandberg et al., 2014).In addition to searching for specific functions in the CNS of the factor AF, it is important to understand whether its activity in counteracting hypersecretion in the intestine is via enteric nervous system ganglia.When considering the beneficial effect against vertigo, it is appropriate to consider actions in terms of cerebellar circuitries, which contribute to body posture and balance control.The circuit mossy fibers-cerebellar granule cells (under inhibitory control by cerebellar Golgi cells)-Purkinje cells-deep cerebellar nuclei-lateral vestibular nucleus Deiters' neurons-spinal cord anti-gravity motor neurons (Ito et al., 1966;Ito, 1984;Hydén et al., 2000) certainly has a role in this connection.The model of primary rat cerebellar granule cells culture has been long used by our group for studying their inhibitory GABA A system (Robello et al., 1993;Gatta et al., 2009;Cupello et al., 2013).Thus, we conducted a study to investigate the effects of AF-16 on the inhibitory GABA-mediated chloride currents in CGCs in culture.The results clearly showed that 1.0 µM AF-16 caused an increase (by 40 %) of GABA A receptor-mediated "peak" chloride currents (Bazzurro et al., 2018).This current component (Robello et al., 1999) is due to GABA A receptors of specific subunit compositions: α 6 β 2/3 γ 2 , α 1 β 2/3 γ 2 and α 1 α 6 β 2/3 γ 2 (Gatta et al., 2009).It corresponds to the "fast" synaptic inhibition of cerebellar granule cells in situ (Farrant and Nusser, 2005), whereas the "steady state" component (Robello et al., 1999) corresponds to the "tonic" inhibition in situ (Farrant and Nusser, 2005).Later, we demonstrated that GABA A receptors containing the α 6 subunit played a major role in that increase (Bazzurro et al., 2022).Our present results indicate a redistribution of GABA A receptors containing the α 6 subunit in CGC with an increased presence in the cells' body vs. that in neurites.We found also the opposite change of those containing the α 1 subunit.Another interesting result (Table 4) is that in neurites the plasma membrane pool of α 6 -containing receptors undergoes a greater decrease than the cytoplasmic one.This happens in correspondence with its significant increase in somatic membranes (Fig. 2).
Considering the different functional profile of α 6 β 2/3 γ 2 and α 1 β 2/3 γ 2 GABA A receptors both in terms of affinity for GABA (Sigel and Baur, 2000) and deactivation/desensitization (Tia et al., 1996), this redistribution on plasma membranes can bring about a different efficiency of the inhibitory control on such neurons.On the other hand, our present results cannot exclude that the α 6 subunit relative overall increase in the cell body is related to "tonic" δ subunit containing receptors (Farrant and Nusser, 2005) which are extra-synaptic and mainly present in cerebellar granules cell body (Merlo et al., 2004).This seems to be denied by the absence of changes in the "steady state" current component in CGC's in culture (Bazzurro et al., 2018).However, subtler differences in α 6 β 2/3 δ and α 1 β 2/3 δ.GABA A receptors (e.g., in desensitization kinetics or affinity for GABA, Sieghart et al., 2022) may have a role, in absence of changes in the current intensity as found in our model.Within cerebellar circuitries, changes in the CGC inhibition processes may be factors in the anti-vertigo action of AF.Recent results show in a mouse model of Essential Tremor (ET), a neurological disorder, which probably involves synchronous rhythmic firing of cerebellar Purkinje cells, a protective effect by drugs which positively modulate cerebellar granule cell GABA A receptors containing the α 6 subunit.Among these, ethanol at low levels Table 1 Percentage effect (E%) of 1.0 μM AF-16 in non-permeabilized samples of cerebellar granule cultures.E% are given as mean ± SEM.P-value was evaluated with Student's t-test; n.s. is not significant.n CTRL is the number of control cells, and n AF-16 is the number of treated neurons.This was suggested to happen via increased inhibition of cerebellar granule cells and a resulting less active excitatory input by these cells on Purkinje cells (Huang et al., 2023).Something similar could happen by the enhancing action of AF on inhibition of cerebellar granule cells in the circuit involving mossy fibers-cerebellar granule cells-Purkinje cells-deep cerebellar nuclei-vestibular neurons-mossy fibers originating in the nucleus reticularis tegmenti pontis (Fanardjian and Sarkisian, 1988).It has to be established whether changes in GABA A receptors composition and distribution, involving the α 6 subunit, may play a role in the intestinal antisecretory activity of AF, possibly via an action on ENS ganglia.The presence of α 6 subunit in ENS ganglia is only marginal (Poulter et al., 1999).However, this subunit has been found in neurons in the human and pig small intestine Meissner's plexus and human Auerbach plexus (Jennische and Lange, 2021, unpublished results).Thus, it cannot be excluded a mechanism of counteraction of intestinal hypersecretion via ENS, although it remains to be demonstrated.Summing up, our present experiments and results support the idea that AF can interfere with cerebellar mechanisms, possibly with those related to body posture and balance, via modification of the inhibitory input to cerebellar granule cells.Whether changes in the function of GABA A receptors of ENS ganglia play a role in its antisecretory action in the intestine has yet to be demonstrated.

Compliance with ethical standards
Please note that all authors have read and have abided by the statement of ethical standards for manuscripts submitted to IBRO Neuroscience Reports.All authors have approved the final article.
-16 is the treated cell fluorescence integrated density, and I CTRL is the control neuron fluorescence integrated density.

Fig. 1 .
Fig. 1.Example of a Z-stack analyzed using the sorftware Fiji.After evaluating the volume of the neuron, we projected the sum of the fluorescence intensity of the 3D frames on the Z-axis.We subtracted the fluorescence noise background and drew a region of interest (ROI) around the soma (A) or neurite (B).The obtained values of the ROI integrated density were normalized with the cell volume.
et al.and pyrazoloquinolinones(Huang et al., 2023;Sieghart et al., 2022).Previous work had indicated in the cerebellar granule cells α 6 βδ receptors those whose activation protects from Essential Tremor (Handforth et al., 2018).The explanation of the protective effect on the ET model by drugs positively activating α 6 subunit containing receptors was a reequilibration of a deranged activity of the Inferior Olive Nucleus (ION)-Purkinje cells (PC)-Deep Cerebellar Nuclei (DCN)-Inferior Olive Nucleus (ION) circuit.

Table 2
Percentage effect (E%) of 1.0 μM AF-16 in permeabilized samples of cerebellar granule cultures.E% are given as mean ± SEM.The p-value was evaluated with Student's t-test; n.s. is not significant.n CTRL is the number of control cells, and n AF-16 is the number of treated neurons.

Table 4
Statistical significance, calculated with Student's t-test, for non-permeabilized and permeabilized samples either for α 1 and α 6 subunits in soma or neurite.n.
s. is not significant.