Protective Effects of a synthetic glycosaminoglycan mimetic (OTR4132) in a rat immunotoxic lesion model of septohippocampal cholinergic degeneration

Using a partial hippocampal cholinergic denervation model, we assessed the effects of the RGTA® named OTR4132, a synthetic heparan-mimetic biopolymer with neuroprotective/neurotrophic properties. Long-Evans male rats were injected with the cholinergic immunotoxin 192 IgG-saporin into the medial septum/diagonal band of Broca (0.37 µg); vehicle injections served as controls. Immediately after surgery, OTR4132 was injected into the lateral ventricles (0.25 µg/5 µl/rat) or intramuscularly (1.5 mg/kg). To determine whether OTR4132 reached the lesion site, some rats received intracerebroventricular (ICV) or intramuscular (I.M.) injections of fluorescent OTR4132. Rats were sacrificed at 4, 10, 20, or 60 days post-lesion (DPL). Fluorescein-labeled OTR4132 injected ICV or I.M. was found in the lesion from 4 to 20 DPL. Rats with partial hippocampal cholinergic denervation showed decreases in hippocampal acetylcholinesterase reaction products and in choline acetyltransferase-positive neurons in the medial septum. These lesions were the largest at 10 DPL and then remained stable until 60 DPL. Both hippocampal acetylcholinesterase reaction products and choline acetyltransferase-positive neurons in the medial septum effects were significantly attenuated in OTR4132-treated rats. These effects were not related to competition between OTR4132 and 192 IgG-saporin for the neurotrophin receptor P75 (p75NTR), as OTR4132 treatment did not alter the internalization of Cy3-labelled 192 IgG. OTR4132 was more efficient at reducing the acetylcholinesterase reaction products and choline acetyltransferase-positive neurons than a comparable heparin dose used as a comparator. Using the slice superfusion technique, we found that the lesion-induced decrease in muscarinic autoreceptor sensitivity was abolished by intramuscular OTR4132. After partial cholinergic damage, OTR4132 was able to concentrate at the brain lesion site possibly due to the disruption of the blood-brain barrier and to exert structural and functional effects that hold promises for neuroprotection/neurotrophism.

can protect neurons from degeneration. However, the protective effect of RGTA® has not been fully assessed in neural cells, such as the cholinergic cells involved in neurodegeneration. Here, we studied the neurotrophic and neuroprotective effects of OTR4132 in a model that met the following two criteria: (i) partial cholinergic damage reproducible in terms of both severity and location, and (ii) limited postinjury sprouting of the cholinergic neurons or other spontaneous restorative mechanisms. A partial immunotoxic lesion was induced by local administration of the immunotoxin 192 IgG-saporin in the forebrain cholinergic neurons located in the medial septum (MS) and vertical limb of the diagonal band of Broca (vDBB). 192 IgG-saporin carries the ribosome-inactivating protein (saporin) combined with 192 IgG (a monoclonal antibody against rat p75 NTR receptors) [25,26]. Since such receptors are specific to cholinergic neurons in the basal forebrain, 192 IgG-saporin injected intraseptally selectively damages this population of neurons [3,27,28]. Interestingly, recovery of choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) activities after the induction of partial lesions is extremely limited in the hippocampus, even after 1 year [29][30][31]. Assessment of the RGTA® named OTR4132 neuroprotective effects in the 192 IgG-saporin model was conducted with several complementary experiments: evaluation of the morphological and histochemical effects of OTR4132 administered directly into the cerebral ventricles or by intramuscular (I.M). injection, followed by the assessment of the neuroprotection effects of OTR4132 after induction of larger lesions. Next, the effects of OTR4132, which is a heparan sulfates/heparin mimetic, were compared to those of heparin. Then, competition between OTR4132 and 192 IgG for the p75NTR receptor was investigated. Finally, additional information about the effects of OTR4132 on the function of cholinergic terminals arising from a partially damaged septum, were obtained using a hippocampal-slice superfusion model. The effect of OTR4132 on the muscarinic modulation of electrically evoked acetylcholine (ACh) release [32,33] was determined using different pharmacological approaches: inhibition of the acetylcholinesterase (with Physostigmine) associated or not to a muscarinic antagonist (Atropine) or a muscarinic agonist (Oxotremorine), which Introduction After injury to the adult mammalian central nervous system (CNS), spontaneous repair is generally limited [1][2][3][4]. Nevertheless, both endogenous and exogenous neuroprotective/ neurotrophic substances may prevent neurons from degeneration and/or enable regeneration [5][6][7]. Experimentally, in vivo CNS repair via direct or indirect effects on neuron survival and/or axon regrowth can be induced by acute, subchronic, or chronic injections of growth factors including nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), or fibroblast growth factors (FGF) [8][9][10][11]. Heparin binding growth factors (HBGF), such as FGF and transforming growth factors (TGF), protected against ischemic injuries, reducing infarct size in an animal model of stroke [10,[12][13][14][15][16][17]. However, administration of exogenous growth factors may induce adverse effects due to penetration of the factors into unlesioned brain regions leading to pleiotropic effects and to immune responses [6,18]. On the other hand, endogenously synthetized neurotrophic factors can also be released in response to injury, but generally only within or near the injured or denervated regions [14]. However, injury-related phenomena, such as enzymatic degradation, which occurs rapidly [19,20], may limit the effects of the endogenous factors. Therefore, compounds that slow down the degradation of endogenously released neurotrophic factors, or that protect neurons from degeneration, may hold therapeutic potential.
Here, we used the glycosaminoglycan mimetic OTR4132, a member of the RGTA® family of polymers, which, by mimicking glycosaminoglycans in the extracellular matrix, protect endogenous growth factors from proteolytic degradation. Some of these polymers are currently used in the clinic as regenerative agents [9,15,21,22]. By protecting HBGF from proteolytic degradation, RGTA® fostered substantial tissue repair in various in vivo injury models [9,21]. Recently, we showed that OTR4132, also known as Hsm4131, administered I.V. after stroke induction in a rat model, conferred long-lasting neuroprotection and significantly reduced functional deficits [23]. It also restored the extracellular matrix and increased brain plasticity processes (i.e. angiogenesis and neurogenesis) in the affected brain hemisphere. In the same set of experiments, we additionally showed that OTR4132 was safe, and that it could specifically locate in the injured hemisphere. Moreover, the therapeutic interest of RGTA®-based therapies was shown in mice models of Alzheimer's disease-related tauopathy in which a RGTA® prototype of small size (called F6) was administrated and blocked neuronal uptake, seeded aggregation and transcellular propagation of stereotactically injected tau fibrils [24]. These data suggest that RGTA® might offer benefits in neurodegenerative diseases where it a fine drill, and a stainless steel cannula was lowered until it reached the meninx to establish "0". The cannula was subsequently placed at vertical (i.e. dorso-ventral coordinate) -7.2 mm and -6.5 mm (V= -7.2, and -6.5) [36]. The injections were given through a stainless steel cannula, which was left in situ for 5 min after each injection to enable diffusion of the toxin. The cannula was then slowly removed. The scalp was sutured with sterile Vicryl*Plus 2-0 (Ethicon, Johnson & Johnson). After surgery, each animal was allowed to recover from anesthesia under a warm lamp before returning to its cage and the animal facility.
Fluorescein-labeled OTR4132 was prepared by incubating a solution of OTR4132 (250 mg, 6.25 µmol, in 1 ml of aqueous buffer, pH= 7.5) with fluorescein-5-thiosemicarbazide (FTSC, 53 mg, 125 µmol, from Molecular Probes/ Invitrogen) at 30 °C for 24 h. Under these conditions, only the single reducing glucose unit present in OTR4132 can be labeled without affecting the global structure of the drug [37][38][39][40]. After purification by ultrafiltration, the FTSCderived-OTR4132 was obtained as a powder and conserved under dry conditions as previously described [37].

OTR4132 treatment
OTR4132 was injected immediately after the induction of hippocampal lesion by a single intracerebroventricular (ICV) injection or by a single I.M. injection as follows.

Animal care
Animal care and procedures were conducted in conformity with our institution's guidelines. 3-month-old Long-Evans male rats were obtained from R. Janvier (C.E.R.J., Le Genest-St-Isle, France). They were housed individually in transparent Makrolon cages (42 × 26 × 15 cm) placed in rooms on a 12 h-12 h light-dark cycle (lights on at 7:00 am) under controlled temperature (21 °C). Food and water were available ad libitum. All procedures have been performed in compliance with 3R's ethical rules on animal experimentation and in accordance with the European Directive concerning animal experimentation (2010/63/EU). The protocol was submitted to the Ministry of Higher Education and Research to obtain the ethic approval for using animals for scientific purposes (Animal protocol numbers: 67-215 and 67-101 for J-CC and SS, respectively under their authorities). Every effort was made to minimize suffering and the number of animals was as small as possible, ranging from 3 to 5 given the experimental and statistical constraints.

Induction of the hippocampal lesion
Procedures, dosages, and time intervals required to detect morphological effects of the immunotoxin were determined in preliminary experiments using the same batch of 192 IgG-saporin (batch 21#107; ATS, San Diego, CA) [2]. This batch was used in all experiments. The size of the immunolesion was monitored using AChE histochemistry in the hippocampus and anti-ChAT immunohistochemistry in the septal area (see below). All surgical procedures were performed under aseptic conditions. A mixture of ketamine (98 mg/kg intraperitoneally) and xylazine (13 mg/kg intraperitoneally) was used for anesthesia. A stereotactic method was used to inject 192 IgG-saporin into the MS. The 192 IgG-saporin dose was 0.37 µg for the induction of a partial lesion (0.46 µg/µl, 0.2 µl/site, 0.1 µl/min, in sterile artificial cerebrospinal fluid) and 0.60 µg for the induction of an extensive lesion (0.75 µg/µl, 0.2 µl/site, 0.1 µl/min). Control animals were injected with artificial cerebrospinal fluid (aCSF, Harvard Apparatus). After the fur was shaved, the scalp was prepared using povidone-iodine and alcohol 70%, and the rats were placed into a stereotaxic device (Kopf Instruments) with lambda and bregma at the same vertical coordinate. A small midline incision was made over the scalp to provide access to the cranium, and the cranial surface was cleaned with oxygenated water. Coordinates for cannula placement to target the MS were anteroposterior to bregma 0.6 mm (AP= + 0.6) and 0.20 mm lateral to the midline (L= ± 0.2). A skull window was outlined with Fig. 1 Chemical structure of OTR4132 molecule. OTR4132, also known as Hsm4131 [23], is a sodium salt of α-1,6 poly(carboxymethylglucose acetyl sulfate) corresponding to the following molecular structure [((C 6  ) and acetyl (-OCCH 3 ) groups with a degree of substitution of 0.60 ± 0.20, 1.50 ± 0.20 and 0.20 ± 0.05, respectively, and represented into the brackets (CAS RN: 2342614-00-8). R is the proportion of each substituted group at both carbons C-3 and the C-4 positions. Each substituted group is randomly distributed on the glucopyranose units [37]. post-fixed in the same fixative for about 2 h, washed with 0.1 M PBS, and transferred into a phosphate-buffered 20% sucrose solution for 36-40 h at 4 °C. The brains were then quickly frozen in isopentane and subsequently cut in a cryostat.

Tissue preparation
Coronal Sect. 30 μm in thickness were collected onto gelatin-coated slides. For rats treated with FTSC-labeled OTR4132, sections were mounted with anti-fading medium and stored at 4 °C in the dark until examination. For other experiments, sections were dried at room temperature for 36 h and stained for AChE according to Koelle's revised protocol [45,46]. Ethopropazine (0.3 mM; Sigma, St-Louis, MO) was used to block non-specific cholinesterases, and acetylthiocholine iodide (4 mM; Sigma, St. Louis, MO) was used as the substrate. The details of the following procedures have been described in our previous works [47,48].

Evaluation of acetylcholinesterase staining
The extent of cholinergic innervation, proportional to the amount of AChE-positive reaction products in the hippocampus [49,50], was determined using a computer-assisted image analysis system (SAMBA Technologies, Paris, France) coupled to a monochrome CCD digital video camera (Model XC 77CE, Sony, Shiga, Japan) equipped with a 60-mm Nikkor objective and a Triplux extension tube (Nikon, Paris, France). Optical density (O.D.) was measured on digitized images by bilaterally delineating the hippocampal region on two sections at the same level of anteriority (AP=-3.14 and AP=-5.30 from bregma, respectively). For digitization, sections were placed on a Kaiser Prolite 5000 light box (Kaiser Fototechnik, Buchen, Germany). Magnification from section to computer screen was 2.5. The mean O.D. of the central part of the corpus callosum within the same brain section was taken as the background level and was subtracted from all measures before analysis. The investigator performing the O.D. measurements was blinded to the group assignment.

Choline acetyltransferase immunostaining in the septal region
Cryostat sections at 60 μm thickness were collected and kept free floating at -20 °C in cryopreservative medium until immunolabeling was performed. The sections were washed several times in 1 M PBS, pH= 7.4, then soaked for 1 h in 10% normal donkey serum (Dutscher, Brumath, France)
In other experiments, FTSC-labeled OTR4132 was injected I.M. (1.5 mg/kg) as well, to determine whether OTR4132 crossed the blood-brain barrier (BBB), penetrated, and remained within the lesion site.
Finally, to explore a possible neuroprotective/neurotrophic effect of I.M. injected OTR4132 in animals with larger lesions, rats were given a higher dose (0.60 µg/rat) of 192 IgG-saporin immediately followed by I.M. OTR4132 injection (1.5 mg/kg).

Other Treatments
For comparison, heparin (0.8 IU/ml ICV or 16 IU/ml I.M.) was injected using the same protocol as for OTR4132 treaments.
For competition between OTR4132 and the antibody 192 IgG for the p75-NTR receptor, a mixture of 5 µl OTR4132 (same dose as for ICV administration, i.e. 0.25 µg/site) and 2 µl of 192 IgG-Cy3 (0.37 µg/site) were injected intraseptally according to the above-described procedure for lesion induction.

Long-term OTR4132-mediated recovery
The effects of 192 IgG-saporin on septo-hippocampal cholinergic neurons and the potential OTR4132-mediated recovery were evaluated 1 month after lesion induction using neuropharmacological measurements of various cholinergic functions in the hippocampal structure as described hereinafter. One-quarter of the ventral left and right hippocampi parts were pooled and homogenated for measurement of cholinergic markers such as acethylcholinesterase and choline acetyltransferase activities. The remaining parts of hippocampi were used to prepare slices for in vitro evaluation of electrically evoked ACh-release in superfusion experiments. Controls for each experiment received sterile sodium salt solution (0.9% NaCl) instead of OTR4132.

Pharmacological approaches for testing ACh release: Tissue preparation
For all histological, histochemical, and immunohistochemical experiments, rats were given an overdose of sodium pentobarbital (100 mg/kg) and perfused transcardially with 60 ml of 4% phosphate-buffered formaldehyde (pH= 7.4; 4 °C) at the different DPL. The brains were then dissected,

Accumulation of [ 3 H]acetylcholine overflow
Accumulation and electrically-evoked release of [ 3 H]-acetylcholine overflow was assessed as previously described [47,48]. Briefly, coronal slices (300 μm thick) of the dorsal hippocampus were prepared using a McIlwain tissue chopper (Model TC752, Mickle Laboratory Engineering Co. Ltd., UK). The slices were distributed into Petri dishes containing 2 ml modified Krebs-Henseleit (KH) buffer , and 66 min (S 3 ) of superfusion. Drugs to be tested were added to the superfusion medium of some chambers from 8 min before S 2 and S 3 onward, in increasing concentrations (indicated on the abscisses in the corresponding figures) from S 2 to S 3 . Control slices were superfused throughout with drug-free KH buffer. At the end of the experiment (after 74 min of superfusion), radioactivity in the superfusate samples and slices (dissolved in 250 µl Solvable®, Perkin-Elmer, Rodgau, Germany) was determined by liquid scintillation counting. Accumulation of [ 3 H]-choline was defined as the tissue tritium content determined immediately before the beginning of fraction collection. Since hippocampal slices were very similar in size and thickness, these values provided an indirect indication of the density of the corresponding cholinergic axon terminals in the hippocampus. The fractional rate of tritium outflow was calculated as (pmoles tritium outflow per min) × 100 / 2 × (pmoles tritium in the slice at the start of the corresponding 2-min period). The stimulation-evoked overflow of tritium ([ 3 H]-overflow) was calculated by subtraction of the basal outflow and was expressed in PBS containing 0.5% Triton X-100 and 0.2% merthiolate. Sections were incubated overnight at room temperature with a goat polyclonal antibody against ChAT (1:200, Chemicon International, Temecula, CA), then incubated for 1 h with biotinylated secondary antibody with peroxidase against goat IgG (1:200; Chemicon International, Temecula, CA). Sections were soaked for 1 h in ABC Vectastain Elite kit (1:500; Vector Labs, Burlingame, CA) before using a diaminobenzidine kit (Vector Labs, Burlingame, CA) to reveal positive neurons. Brain sections were then mounted onto gelatin-coated slides, dried, dehydrated, and coverslipped. Omission of the primary antibody served as the negative control and resulted in no detectable staining.

Determination of choline acetyltransferase-positive neuron counts
At the appropriate level of anteriority, the anterior commissure was used as an anatomical landmark to select, define, and standardize the location of counting frames of a predefined size in the MS. This frame was defined ventrally by an imaginary line through the center of the anterior commissure, and both dorsally and laterally by the distribution of ChAT-positive neurons, as described in detail elsewhere [51]. Three coronal sections per animal were selected: one section corresponding to the injection site (AP + 0.6 mm, as determined by the cannula track) and, in order to assess immunotoxin diffusion, two sections located within 0.4 mm from the injection site.

Determination of choline acetyltransferase and acetylcholinesterase activities
Rats used for superfusion experiments were sacrificed by decapitation using a custom-made guillotine, after carbon dioxide inhalation, one month after lesion induction. Quick removal of the brain was followed by the dissection of the hippocampus from each hemisphere. The dorsal threequarters of the hippocampus was used to prepare slices for release experiments (see below). The remaining ventral piece was homogenized in 0.5 ml of 0.32 M sucrose (in 2.5 mM HEPES, pH= 7.4) using a Potter Elvehjem glass/Teflon homogenizer (eight strokes at 500 rpm). From this crude homogenate, a 40-µl sample was taken, diluted with 360 µl of 0.1 N NaOH and used for protein assay. A 100-µl sample of the homogenate was stored at -80° C until determination of ChAT and AChE activity as previously described [47,48]. NaCl) as the factors. This analysis was performed separately from the analyses of the other three DPL, because at 4 DPL the damaging effects of 192 IgG-saporin were not yet at their maximum level [54]. O.D. and ChAT data from 10, 20, and 60 DPL were analyzed using three-way ANOVA with Group, Drug and DPL as the factors. As the variances of some measures did not satisfy the criterion of homogeneity, square root transformation of the data was performed prior to running the ANOVA, as described elsewhere [55], and post-hoc Tukey test performed. Type I error was set-up at 5% (i.e. α = 0.05). Data are represented as mean ± standard deviation (SD).

OTR4132 protects cholinergic neurons from partial lesion of hippocampus at the septal anterior and posterior levels, and cholinergic neurons from the medial septum
In order to detect the partial 192 IgG-saporin lesions and recovery after OTR4132 ICV administration (0.25 µg/5 µl/ rat in lateral ventricle), hippocampal slices were stained for AChE-positive fibers, and medial septal slices were immunostained for ChAT-positive neurons ( Fig. 2A). In the dorsal hippocampus (Fig. 2B), AChE staining was markedly decreased in the dentate gyrus and CA1 layer of the lesioned rats compared to the sham-operated rats ( . Moreover, we looked for an effect of side (right or left) that might produce asymmetry of the immunotoxic damage. Side had no significant effect. Therefore, the mean of the values on the two sides was used in subsequent analyses. Interestingly, after 4 DPL, ANOVA revealed a significant Group effect (Sham, Lesioned: F 1/20= 19.6, p-value< 0.01), but no significant either in absolute terms ('nCi') or in relative terms (percent of the tritium content of the slice at the onset of the relevant stimulation period). Effects of drugs added before S 2 (or S 3 when indicated) were determined as the ratio of the overflow evoked by the corresponding stimulation period (S n /S 1 ) and normalized according to the appropriate control ratio (100%, no drug addition before the corresponding stimulation). Values significantly greater than 100% indicated facilitation and values significantly smaller than 100% indicated inhibition of evoked ACh release.

Statistical analyses
O.D. data were first analyzed using one-way analysis of variance with side (left or right) as the factor. O.D. and ChAT data from 4 DPL were analyzed using two-way ANOVA with Group (Sham or Lesioned) and Drug (OTR4132 or medial septum and vertical limb of the diagonal band of Broca (Fig. 2D, right panel).
Taken together, the data showed that partial lesions of cholinergic neurons at the medial septum projecting to the hippocampus, induced by 192 IgG-saporin injection are significantly protected by ICV injection of OTR4132 (0.25 µg/5 µl/rat in lateral ventricle).

OTR4132 I.M. administration protects cholinergic neurons from partial septal lesions
Based on the data hereinbefore, to further develop potential therapeutic approaches, OTR4132 was administrated I.M. at 1.5 mg/kg in rats to test its protection on cholinergic neurons from partial lesions induced by 192 IgG-saporin (Fig. 3A). As in the previous experiment, hippocampal slices were stained for AChE-positive fibers and medial septal slices for ChAT-positive neurons. The overall histological findings were similar to those observed after OTR4132 ICV injection, in terms of both the density of AChE reaction products in the hippocampus and the number of ChAT-positive neurons in the MS. Thus, I.M. administration of OTR4132 Drug effect (NaCl, OTR4132: F 1/20< 1.0) and no significant interaction between the two factors (F 1/20< 1.0), as shown in Fig. 2C on the top right panel. The Group effect reflected significantly lower AChE O.D. values in the lesioned animals compared to their sham-operated counterparts (approx. -30%). At the three other times after lesion, ANOVA showed a significant Group effect (F 1/59= 122.5, p-value< 0.001), a significant Drug effect (F 1/59= 8.1, p-value< 0.05), and a significant Group-Drug interaction (F 1/59= 3.9, p-value= 0.05). There was no significant DPL effect (F 2/59< 1.0), and none of the other interactions was significant. Then, the Group-by-Drug interaction was ascribable to significantly higher AChE O.D. values in OTR4132-treated lesioned rats compared to their saline-treated counterparts (ANOVA test; p-value< 0.001), whereas the difference between OTR4132 and saline treatment in Sham rats was not significant.
Since OTR4132 significantly protect cholinergic neurons at the septal posterior level from partial hippocampus lesions, we investigated if the OTR4132 protective effect could also be observed at the septal posterior level (Fig. 2C, bottom panels). The results were very similar to those obtained at the anterior level (Fig. 2C, top panels). Indeed, at 4 DPL, two-way ANOVA showed a significant Group effect (F 1/20= 9.5, p-value< 0.01, ANOVA-test) but no significant Drug effect (F 1/20< 1.0), and no significant interaction between the two factors (F 1/20 <1.0). At the other DPLs, significant Group (F 1/59= 127.6, p-value< 0.001) and Drug (F 1/59= 6.8, p-value< 0.05, ANOVA-test) effects were found, as well as a significant Group-Drug interaction (F 1/59= 4.2, p-value< 0.05, ANOVA-test), with no significant overall DPL effect (F 2/59= 1.4). Multiple comparisons showed that these effects (Fig. 2C, bottom panels) were ascribable to the same differences as for the anterior level (Fig. 2C, top panels).
These results indicated that OTR4132 ICV administrated (0.25 µg/5 µl/rat in lateral ventricle) significantly protected hippocampus cholinergic neurons at the septal posterior level.
Because the majority of cholinergic projections target the hippocampus and the entorhinal cortex in the basal forebrain complex, specifically the medial septum (MS),In order to evaluate 192 IgG-saporin diffusion and the effect of OTR4132 ICV injection (0.25 µg/5 µl/rat in lateral ventricle) on these cholinergic neurons, ChAT-immunostaining in the MS and vertical limb of the diagonal band of broca was performed, and the number of ChAT-positive neurons was measured at the injection site (AP= + 0.6 mm; Fig. 2D). The data showed that OTR4132 significantly protects cholinergic neurons from partial lesion of hippocampus at the Rats were first either injected with aCSF or 192 IgG saporin (0.37 µg, ICV), and then injected with OTR4132 (0.25 µg, ICV). Brains were harvested 4, 10, 20 and 60 days post-lesion (DPL) and ChAT-and AChE-immunostainings were performed at the medial septum (+ 0.6 mm) and hippocampus sections of each brain (-3.14 mm and -5.30 mm) according to the brain rat stereotaxic coordinates [36]. (B) Representative photomicrographs of cryostat sections through the dorsal hippocampus (left panels) or medial septum (right panels) from rats subjected to a sham operation (top panels), or after intra-septal injection with 192 IgG-saporin inducing immunolesion followed immediately by intracerebroventricular (ICV) NaCl 0.9% injection (middle panels), or by ICV OTR4132 injection (bottom panels). Black scale bar= 2.0 mm; white scale bar= 50 μm. (C) Quantification of photomicrographs from (B, dorsal hippocampus) by optical density (O.D.) assessment of acetylcholinesterase (AChE)-positive neurons in coronal brain sections through the dorsal hippocampus at an anterior (AP= -3.14 mm; top panels) or posterior (AP= -5.30 mm; bottom panels) levels in the four groups of rats including rats with sham surgery injected with NaCl or OTR4132, or rats with 192 IgG-saporin lesion followed immediately by ICV injection of NaCl 0.9% solution or OTR4132 solution. Top panels represent the percentages of the mean O.D. compared to Sham + NaCl group of the data collected at days 4, 10, 20, and 60 post-lesion. Right panels represent the data from left panels were collapsed for each group at either the anterior (AP= -3.14 mm) or the posterior (AP= -5.  (p-value= 0.058, ANOVA-test followed by post-hoc Tukey test; Fig. 3C, right panel).
Taken together, these results indicate that, after I.M. injection, OTR4132 is able to significantly protect cholinergic neurons at the medial septum projecting to the hippocampus from partial lesions induced by specific 192 IgG saporin, and thus OTR4132 is able to reach the injured area of the brain.

I.M. administration of fluorescent OTR4132 reaches hippocampal injuries
To check that I.M. injection of OTR4132 indeed can cross the BBB and penetrate into the brain to specifically reach the partial lesions of cholinergic neurons (Fig. 4), rats received intracerebral injection of aCSF (Fig. 4A) or 192 IgG-saporin (Fig. 4C), and immediately after, received ICV (Fig. 4C) or I.M. injections (Fig. 4E) of FTSC-labeled OTR4132. Furthermore, to determine the duration of OTR4132 persistence at the lesion site, rats injected with FTSC-OTR4132 were killed 4, 10, and 20 DPL to be analyzed.
As shown in Fig. 4, a fluorescent signal from FTSClabeled OTR4132 was identified in the MS at all the DPLs studied. Interestingly, this signal was detected after both ICV ( Fig. 4B and D) or I.M. injection (Fig. 4F) of OTR4132. Also, the fluorescent signal was confined to the lesion site and no fluorescence was noticed in the hippocampus or at 1.5 mg/kg induced a significant 20 to 30% increase in the O.D. of AChE reaction products (measured at both hippocampal levels i.e. AP= -3.14 mm and AP= -5.30 mm; Fig. 3B) in lesioned rats compared to untreated lesioned rats (i.e. 192 IgG saporin + NaCl 0.9%). At the anterior level (AP= -3.14; Fig. 3B, top panels), from 10 DPL onward, ANOVA showed significant effects of Group (F 1/56= 85.7, p-value< 0.001, ANOVA-test) and Drug (F 1/56= 4.14, p-value< 0.05, ANOVA-test), as well as a significant Group-Drug interaction (F 1/56= 6.4, p-value< 0.05, ANOVA-test). This interaction was ascribable to a significantly higher O.D. value in the OTR4132-treated lesioned rats compared to their saline-treated lesioned counterparts. No such difference was found in the sham-operated animals. There was no significant DPL effect (F 2/56= 1.3) and no other significant interaction was found (F 2/56< 1.0).
Moreover, at a more posterior hippocampal level (AP= -5. Similarly, OTR4132 I.M. injection significantly increased ChAT-positive neuron counts in the septal area after lesion induction at the various anteriority levels, compared to saline injection (Fig. 3C, left panel, DPL= 20). However, this increase was not but close to significancy The data showed that ICV or I.M. injection of heparin ( Fig. 5A and D, respectively) has significant protective effects, similar to those observed with OTR4132 in the hippocampal anterior level (AP= -3.14 mm), but not at the posterior (AP= -5.30 mm) level (Fig. 5B, left and right panels, respectively). At the anterior level (AP= -3.14 mm, Fig. 5B and E, left panel), acetylcholinesterase staining in hippocampal slices showed a significant Group effect (ICV: Our data show that the Group-by-Drug interaction was ascribable to significantly higher O.D. values in heparin-treated lesioned rats compared to their saline-treated counterparts (p-value< 0.001, ANOVA-test). The difference between heparin and saline-treatment was not significant in sham-operated rats (data are summarized in Fig. 5B and E).
These results indicate that OTR4132 is able to reach and stay only in the injured area of the brain even when I.M. injected.
Taken together with the results from Fig. 3, the data indicate that OTR4132 injected I.M. is able to reach the injured area of the brain, in agreement with the fluorescence observed in the lesioned region of the brain after injection of OTR4132-FTSC.

Effects of heparin administration after partial lesion to the septal cholinergic neurons
RGTA® such as OTR4132 have been specifically designed to be structurally similar to heparin and heparan sulfates to potentially replace them, as future treatments after brain lesions such as in AD [23,24,56]. Thus, effects of heparin administrated by ICV or I.M. injections have been tested in the model of partial lesions of cholinergic neurons induced by intracerebral injection of 192 IgG-saporin (Fig. 5). ChAT-and AChE-immunostainings were performed at the medial septum (MS, + 0.6 mm) and hippocampus sections of each brain (-3.14 mm and -5.30 mm) according to the brain rat stereotaxic coordinates [36]. Quantification of brainimmunostaining by optical density (O.D.) assessment of (B) AChE-positive neurons in coronal brain sections through the dorsal hippocampus at an anterior (AP= -3.14 mm) or posterior (AP= -5. (Bregma +0.6 mm; Fig. 5C and F) had no effect on the survival of ChAT-positive.
Taken together, these results indicate that heparin is less efficient than OTR4132 in protecting cholinergic neurons from 192 IgG-saporin induced partial injury.
Then, the effects of ICV or I.M. injection of heparin in ChAT-positive neurons by immunostaining in the medial septum and vertical limb of the diagonal band of Broca ChAT-and AChE-immunostainings were performed at the medial septum (+ 0.6 mm) and hippocampus sections of each brain (-3.14 mm) according to the brain rat stereotaxic coordinates [36]. Representative photomicrographs of coronal cryostat sections through the (B) dorsal hippocampus or (C) medial septum of rats subjected to a sham surgery (top panels), or injected with high-dose (0.60 µg) of 192 IgG-saporin followed immediately by I.M. with saline (middle panels), or by I.M. with OTR4132 (bottom panels). Number of rats per group was three (n=3). Black scale bar= 2.0 mm; and white scale bar= 50 μm. DH: dorsal hippocampus; MS: medial septum.

Intraseptal co-injection of OTR4132 and fluorescent 192 IgG
To evaluate whether nonspecific binding of OTR4132 to p75-NTR receptor might prevent the binding and internalization of 192 IgG-saporin, thus explaining the decrease in lesion severity after OTR4132 treatment, we assessed the internalization of fluorescent 192 Cy3-labeled IgG by confocal microscopy after intraseptal co-injection or not of OTR4132 (Fig. 7A). Results in Fig. 7B and C showed that no difference in fluorescence staining was observed between rats that received or not OTR4132 (Fig. 7B and C), both at a dorsal (top panels) and at a ventral level of the medial septum (bottom panels). Accordingly, the quantitative fluorescence analyses of the Cy3 positive neurons did not show significant differences between the group of rats receiving 192 IgG-Cy3 alone (Fig. 7B) and those co-injected with 192 IgG-Cy3 + OTR4132 (Fig. 7C).
These results show that OTR4132 does not prevent the internalization of fluorescent 192 Cy3-labeled IgG, and thus OTR4132 decreases the severity of the MS lesions of 192 IgG-saporin by a mechanism different from competing with the mechanisms usually involved in internalization of 192 IgG-saporin into the neurons, which is based on IgG specific interaction with target protein on the cell surface [57]. Also, 192 IgG-saporin is a protein composed of around 250-290 amino acids, corresponding to a molecular weight of OTR4132 I.M. administration does not protect cholinergic neurons from larger septal lesions In Fig. 3, data showed that I.M. injection of OTR4132 significantly protects cholinergic neurons from partial septal lesion. To go further with OTR4132 neuroprotection effects, we explored a possible neuroprotective/neurotrophic effect of I.M. injected OTR4132 in animals with larger lesions in cholinergic neurons. Rats were given a higher dose (0.6 µg) of 192 IgG-saporin, immediately followed by I.M. OTR4132 injection (1.5 mg/kg; n= 3 under each lesion and drug condition). As illustrated in Fig. 6A, at 10 DPL, no ChAT-positive neurons in the septal (Fig. 6B) or AChE reaction products in the hippocampus (Fig. 6C) area were detected following the administration of the higher dose of 192 IgG-saporin (0.60 µg). Under these conditions, no beneficial effect of OTR4132 were detectable.
Taken together, these results show that OTR4132 protects cholinergic neurons of the medial septum projecting to the hippocampus from partial, but not extensive, lesions. These data suggest that OTR4132 treament has to be administered early during the development of cholinergic neuron degenerescence to be effective.  (DPL) and immunoflurorescenvce performed at the medial septum (+ 0.6 mm) according to the brain rat stereotaxic coordinates [36]. Representative immunofluorescence photomicrographs of cryostat sections of intraseptal injection of (B) 192 IgG-Cy3 alone or (C) co-injected with OTR4132 through the dorsal (top panels) or the ventral (bottom panels) medial septum in rats (n= 3). MS: medial septum.
To go further on understanding OTR4132 protecting effects and potential physiological effects on cholinergic neurons affected by 192 IgG-saporin, pharmacological approaches have been used (Fig. 9A). To assess if OTR4132 modified evoked overflow of [ 3 H]-ACh, physostigmine was used alone or associated with atropine. Previous studies have shown that physostigmine alone can reduce the evoked [³H] overflow [33,58]. Accordingly, in this study, the results showed that physostigmine significantly reduced by about 28.6 kDa with alpha-folds, it does not contain an heparin binding-site nor a cluster of basic amino acids (D,R,E, K and/or H). Thus, it is unlikely that OTR4132 can directly bind to 192 IgG-saporin.
To get more mechanistic insights, the effects of the muscarinic agonist oxotremorine on electrically-evoked overflow of [ 3 H] in hippocampal slices preincubated with [³H]-choline was assessed in animals treated or not with OTR4132 (Fig. 9D). The data showed that increasing oxotremorine doses significantly inhibited the electrically-evoked [³H] 45% the [³H] overflow in hippocampal slices preincubated with [³H]-ACh (Fig. 9B), its inhibitory effects being slightly (but not significantly) less pronounced in lesioned animal treated with NaCl or with OTR4132 (35% and 30%, respectively, Fig. 9B).
Moreover, the simultaneous presence of atropine not only abolished this reduction, but also increased the evoked [³H]-ACh overflow by up to 33% (Fig. 9C). Statistical analyses revealed a significant Drug effect (F 1/18= 210.2, p-value< 0.001, ANOVA-test) and a significant "Group x Drug" interaction (F2/18= 1.78: p-value< 0.05, ANOVA-test), but not significant Group effect (F2/18= 3.51; p-value= 0.52, ANOVA-test). The 16% increase in the evoked overflow stimulation period (S 1 ) was slightly but not significantly improved by OTR4132, the mimetic was able to attenuate the lesion effects by acting on the responsiveness of inhibitory muscarinic autoreceptors to oxotremorine without altering the presynaptic serotoninergic modulation of [ 3 H]-ACh release.

Discussion
During the last decades, it has been shown that basal forebrain cholinergic neurons undergo degeneration in some brain diseases, such as Alzheimer's disease, stroke, and traumatic brain injury [59][60][61]. Therefore, compounds that protect neurons from degeneration are candidates for potential therapeutic intervention [62]. In this study, by using a model of partial immunotoxic lesion of septal cholinergic neurons induced by 192 IgG-sapo injection, we assessed the capacity of the glycosaminoglycan mimetic named OTR4132 to protect cholinergic neurons from degeneration. We first showed that OTR4132 administered within the lateral ventricles (i.e. ICV) or by I.M. injection significantly protect MS cholinergic neurons and their projection fibers in the hippocampus. Accordingly, OTR4132 significantly restored the function of hippocampal muscarinic autoreceptors on hippocampal cholinergic terminals in the mice model. By using a fluorescent OTR4132, we showed that OTR4132-FTSC administered I.M. could accumulate in the injured brain region, suggesting that, in 192 IgG-saporin-induced brain injury, the compound could cross the BBB. In the first part of our study, we validated the model using small amounts (0.37 µg) of 192 IgG-saporin to induce a selective but partial lesion of the septohippocampal cholinergic pathways, as previously described [31,47,48]. Accordingly, septal GABAergic neurons expressing parvalbumin were relatively preserved. The model was characterized by the loss of about 50% of septal cholinergic neurons near the immunotoxin injection sites (i.e. at the level of the cannula tracks). In the OTR4132 non-treated injured dorsal hippocampus, the density of AChE reaction products decreased by about 45-50%, at both the anterior and the posterior levels. It is well known that the intensity of AChE staining in the hippocampus reflects the density of cholinergic septohippocampal inputs [63,64]. Accordingly, the lesion in our model probably deprived the hippocampus of approximately 50% of its cholinergic inputs. Interestingly, the toxic effect of 192 IgG-saporin was the greatest at about 10 days after the neurotoxin injection, and remained stable thereafter until our latest time point (i.e.60 DPL). Previous reports found that ChAT activity depletion in brain regions innervated by basal forebrain cholinergic neurons was the greatest at about seven days after immunotoxin injection with overflow (dose-effect: F(3/42)= 235.93; p-value< 0.001, ANOVA-test). We additionally observed a significant Group effect (F(2/14)= 7.71; p-value< 0.05, ANOVA-test), with a significantly smaller decrease of the release in lesioned rats given saline (192 IgG saporin + NaCl) compared to sham-operated rats (Sham + NaCl) or OTR4132-treated lesioned rats (192 IgG saporin + OTR4132; p-value< 0.01, ANOVA-test followed by pos-hoc tukey test). The effects of oxotremorine on the evoked [³H] overflow in sham-operated rats (Sham + NaCl) were not different from those in OTR4132-treated lesioned rats (192 IgG saporin + OTR4132; p-value= 0.53, Fig. 9D). The results also showed a significant interaction between Dose and Group comparison (F(6/42)= 14.61; p-value< 0.001; ANOVAtets), especially with the highest dose of oxotremorine (10 µM), which induced a larger decrease in the evoked release in the sham-operated (Sham + NaCl, -75%) and lesioned rats treated with OTR4132 (192 IgG saporin + OTR4132, -79%) than in the lesioned rats given saline (192 IgG saporin + NaCl, -57%; p-value< 0.001; ANOVA-test followed by pos-hoc tukey test, Fig. 9D). These results suggest that OTR4132 failed to reverse the decrease in electricallyevoked [ 3 H]-ACh release during the first stimulation period (S 1 ). However, OTR4132 attenuated the lesion effects on the responsiveness of inhibitory muscarinic autoreceptors to oxotremorine.
Taken together, these data showed that I.M. administration of OTR4132 attenuated the decrease of muscarinic auto-receptor sensitivity after 192 IgG-saporin lesions, but such lesions, whether combined with OTR4132 treatment or not, did not produce changes in serotoninergic heteroreceptor sensitivity. This suggests that, although the decrease in electrically-evoked [ 3 H]-ACh release during the first experiment, no evidence of Evans blue extravasation following intraseptal injection of 192 IgG-saporin was found [2]. Therefore, under our working conditions, the presence of OTR4132 within the lesion reflected the BBB permeability to OTR4132. Selective concentration of RGTA® in and near the lesion has been reported after systemic injection of tritiated RGTA® in a muscle lesion model [71]. RGTA® selectively targeted the muscular lesion and persisted in the lesion for several weeks. As suggested by Meddahi et al., RGTA® would bind specifically to the multiple heparin-binding sites of matrix proteins that become available after lesion-induced degradation of heparan sulfates and chondroitin sulfates [72]. As OTR4132 was administered immediately after 192 IgG-saporin injection, its protective effects may be due to indirect mechanisms. OTR4132 may reduce the internalization of 192 IgG-saporin, and therefore its toxic effects, either by decreasing the penetration of 192 IgG-saporin within cells or by competing with the p75 NTR receptors. However, our results obtained with intraseptal co-injection of Cy3-labeled 192 IgG and OTR4132 argue against this possibility. The septal cellular fluorescent signal accounting for saporin incorporation was not different between the rats receiving OTR4132 and their saline-treated counterparts, arguing against milder toxin-induced injury as the mechanism underlying cholinergic neuron protection by OTR4132, but in favor of a direct protective effect of OTR4132 on the extracellular matrix injury induced by 192 IgG-saporin.
Our data suggest that OTR4132 may exert beneficial effects on basal forebrain cholinergic neurons. Although mechanisms involving interaction with growth factors were not studied here, for decades, it was showed that RGTA® interact, protect, and increase the bioavailability of various heparin binding growth factors [9,21,71,73], thus increasing their half-life [9,21,[74][75][76]. Conceivably, RGTA® may act via mechanisms involving the NGF and/ or BDNF-mediated pathways. Accordingly, evidence for neuroprotective effects of growth factors after septohippocampal pathway injury has been reported [12,77]. Thus, the neuroprotective effects of RGTA® may involve an increase in endogenous neurotrophic factor levels, promoting neuroprotection and repair. This putative mechanism is consistent with the ability of NGF and BDNF to restore impaired cholinergic function in several lesion models [18,78].A transient lesion-induced increase in NGF alone does not seem sufficient to promote survival and/or recovery in lesioned animals treated with saline only. Treatment with the NGFprotecting OTR4132 may have prolonged the beneficial effects of endogenous NGF, thus supporting the restoration of cholinergic functions.
Previous experiments have shown that highly sulfated glycosaminoglycans, such as chondroitin sulfate E, possess non spontaneous recovery [65,66]. After assessing the toxicity of 192 IgG-saporin in forebrain cholinergic neurons, we assessed the neuroprotective effect of OTR4132 administrated ICV or I.M. OTR4132 attenuated the effect of 192 IgG-saporin on septohippocampal cholinergic neurons. This finding is consistent with the reported cholinergic neuroprotective effects of a previously tested glycosaminoglycan analog called GAG-C3 that was not further clinically developed, possibly because of its structural heterogeneity [67,68]. GAG-C3 was used daily for seven days after cholinergic neurons damage induced by ICV administration of the fairly selective cholinergic toxin named ethylcholine aziridinium (AF64A) [3,67,68]. Interestingly, the neuroprotective effects of OTR4132 in the present study occurred after a single OTR4132 injection performed immediately after the immunotoxin injection. In addition, in this study, OTR4132 was found in the lesion site for 20 days after a single administration (as discussed hereinafter). Accordingly, in a recent study, we also showed that OTR4132 injected I.V. protected and repaired the brain after a stroke, favoring functional recovery [23]. Here, we found that OTR4132 not only crossed the BBB, as it was found within the brain after intramuscular injection, but also specifically targeted the lesion site, where it was detectable for at least 20 DPL. Mechanical BBB injury caused by the needle was not responsible for penetration of OTR4132 through the BBB, since no FTSClabeled OTR4132 was found in the brain of saline-infused control rats. However, we cannot rule out that the immunotoxin itself induced BBB alterations allowing OTR4132 penetration. Roher et al. suggested that injection of 192 IgG saporin into the nucleus basalis of rabbits might result in breaching of the BBB, probably via alterations in vascular chemistry [69]. FTSC-labeled OTR4132 was detected in the vicinity of the 192 IgG-saporin-induced lesions. Conceivably, the mechanical injury induced by cannula insertion may result in a milder degree of inflammation when saline is injected, whereas immunotoxin injection may boost the inflammatory response, concentrating specific neurotrophic factors near the lesion site [70]. Thus, the kinetics of local heparanase activity induction may differ between saline and immunotoxin injection, and the number of available chondroitin-and heparan-binding sites in the lesion immediately after the injection may also differ. As FTSC-OTR4132 was injected rapidly after intraseptal insertion of the cannula, there were probably not enough GAG-binding sites available to allow detection in sham animals. Importantly, fluorescent OTR4132 was not identified in the cortex of rats injected with 192 IgG-saporin, although the cannula altered this tissue. Moreover, OTR4132 appeared to selectively target the focus of inflammation produced by the immunotoxin. This site selectivity may limit possible adverse effects on adjacent healthy tissue. Additionally, in a preliminary obtained with OTR4132. Also, in a lesion model characterized by a stronger distal inflammatory response compared to our study model, OTR4132 might display a more extensive neurotrophic effect in the CNS compared to heparin. This hypothesis is supported by the recent publication on OTR4132 protection of the brain tissue following stroke injuries [23].
In the present study, some experiments were designed to examine lesion and OTR4132 effects on hippocampal muscarinic autoreceptor functions. Indeed, we assessed the effects of 192 IgG-saporin and OTR4132 on the modulation of electrically-evoked [ 3 H]-ACh release in hippocampal slices. The electrically-evoked overflow of [³H]-ACh from hippocampal slices preincubated with [³H]choline reflects the action potential-induced exocytotic release of [³H]-ACh, as shown previously [47,48]. A major advantage of the slice superfusion technique is that it enables a focused and reliable investigation of modulation by axon terminals ex vivo, using a small piece of brain tissue [47]. Previous studies of the cholinergic neurons originating in the septal area and projecting to the hippocampus showed that acetylcholine release was under inhibitory control involving, among other receptors, muscarinic autoreceptors of the M 2 type and serotonergic heteroreceptors of the 5-HT 1B type [82,83]. In the present study, OTR4132, failed to induce recovery of the partial lesion-induced decrease in ChAT and AChE activities in the ventral part of the hippocampus (Fig. 8). However, the decrease in electrically-evoked [ 3 H]-ACh release during the first stimulation period (S 1 ) was slightly but not significantly improved by OTR4132. In addition, OTR4132 attenuated the lesion effects on the responsiveness of inhibitory muscarinic autoreceptors to oxotremorine (Fig. 9D), without altering the presynaptic serotonergic modulation of [ 3 H]-ACh release (Fig. 9E). The level of release in hippocampal slices of saline-treated lesioned rats revealed a decrease in the sensitivity of muscarinic autoreceptors toward both the exogenous agonist oxotremorine and the endogenous agonist ACh. This lesion-induced downregulation of muscarinic autoreceptor sensitivity may ensure sufficient ACh release in the hippocampus by the cholinergic axon terminals spared by the lesion. Similar downregulation of muscarinic autoreceptor sensitivity was, however, not apparent in the rats treated with OTR4132. Thus, administration of the OTR4132 seems either to promote restoration of sensitivity of autoreceptors-mediated inhibition of ACh release in the terminals spared by the lesion or to prevent lesion-induced downregulation of these receptors. Such downregulation was demonstrated in experiments involving more extensive but less selective damage to the cholinergic septohippocampal pathways [82,83]. Neural growth factors may contribute to the changes in receptor sensitivity. For instance, FGF2 infusion restored neural function in a model direct neurite growth-promoting activity and reduce the amounts of growth factor needed to promote this activity [79,80]. Our evaluation of cholinergic marker kinetics suggested that ICV or I.M. injection of OTR4132 may have neuroprotective rather than neurotrophic effects. ChAT immunostaining and histochemically-documented AChE activity, which were similarly reduced at 4 DPL in all lesioned animals, decreased further over time in salinetreated rats but remained stable in their OTR4132-treated counterparts. These findings suggest that OTR4132 slowed and attenuated the lesion-induced decreases in cholinergic markers, rather than promoted restoration after an initially larger reduction. It is noteworthy, however, that the protective effect of OTR4132 varied according to lesion size as it was undetectable in animals with large immunotoxic lesions. The marked septal tissue necrosis seen with the higher toxin amount (Fig. 6B) may have been associated with severe disorganization of the extracellular matrix scaffold, preventing OTR4132 from attaching to the lesion site. Furthermore, the higher toxin concentration may have caused the death of a larger number of cholinergic neurons compared to partial lesions, which may have overwhelmed the rescue/protection capabilities of locally synthesized neurotrophic factors (Fig. 6B).
RGTA® are synthetic heparin analogs that mimic some of the properties of heparin while exhibiting two favorable characteristics, (1) they are not degraded by heparanases, hyaluronidases, or other glycanases, and (2) their anticoagulant activity is only a fraction of that of heparin [22,37,77]. Therefore, it was of interest to compare heparin and OTR4132 in our model. In this study, ICV or I.M. injection of heparin had no beneficial effect on the number of cholinergic cell bodies in the MS but attenuated the lesion-induced decrease in hippocampal AChE staining. Thus, unlike OTR4132, heparin failed to protect the septal cholinergic cell bodies but possibly stimulated the sprouting of remaining cholinergic fibers, thereby increasing the AChE activity in the hippocampus. Marked increases in mRNA of heparin-binding growth-associated molecule (HB-GAM) have been reported after hippocampal neuronal injury. HB-GAM protein binds strongly to heparin and enhances neurite outgrowth in vitro [79][80][81]. The difference in effects between heparin and OTR4132 may be related in part to a higher level of enzyme activity at the site of toxin injection compared to within the hippocampus, leading to a faster degradation of heparin than of OTR4132 in the MS. This fast degradation would limit the effects of heparin to the hippocampal target region, and therefore to neurite outgrowth from the terminals of surviving cells, and would preclude an effect on cell survival. However, we cannot exclude that repeated administrations of heparin with low anticoagulant activity would eventually produce the same effects as those Conflicts of interest/Competing interests Pr. Denis Barritault is co-inventor and co-owner in the patents describing the RGTA® technology and president of the company OTR3, manufacturer of RGTA®-based products. Dr. Franck Chiappini is an employee of OTR3. RGTA®is a registered trademark of OTR3. None of the other authors have commercial interests and they state that they have no conflict of interest.
Ethics approval All animal procedures have been performed in compliance with 3R's ethical rules on animal experimentation and in accordance with the European Directive on animal experimentation (2010/63/EU). The protocol was submitted to ethic approval of the Ministry of Higher Education and Research to obtain a project authorization for using animals for scientific purposes (Animal protocol numbers: 67-215 and 67-101 for J-CC and SS, respectively under their authorities). Every effort was made to minimize suffering, and the number of animals was as small as possible given the experimental and statistical constraints.

Consent for publication
The authors listed in this manuscript consented to publish this manuscript.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. of cognitive dysfunction in rat injury induced with Kainic acid, possibly due to mitogenic effect on dentate gyrus neurogenic area leading to generation and migration of newer cholinergic neurons [84]. Under our conditions, OTR4132 treatment may potentiate the beneficial effects of endogenous neurotrophic factors on maintenance of pre-synaptic cholinergic neuronal function, thus preserving/restoring the auto-inhibitory processes impaired by the lesion and the pre-synaptic autoreceptor-mediated regulatory activities. We are currently investigating these hypotheses.
The structure of OTR4132 mimics the highly sulfated glycosaminoglycans such as heparin, heparan sulfates and chondroitin sulfate E, known for their neuroprotective properties. It differs from the structure of the more abundant, weakly sulfated chondroitin sulfates A and B, which inhibit new axonal growth after injury [85]. The current study suggests that potentiation by OTR4132 of the protective effects exerted by growth factors normally bound to extracellular matrix proteoglycans may reduce cholinergic deficits in the degenerating brain. This hypothesis is further supported by studies on the role of altered heparan sulfate structures and functions in the mechanisms leading to Alzheimer's disease-related [56,86,87].
In conclusion, our study shows that OTR4132, a dextran-based polymer, engineered to mimic highly sulfated glycosaminoglycans, may attenuate the extent and functional damages of partial immunotoxic cholinergic lesions. Memory impairments were previously shown to be mild or absent after partial 192 IgG saporin-induced lesions of basal forebrain cholinergic neurons [84,88,89]. Therefore, we did not assess effects of OTR4132 on cognition. However, cognitive evaluations, perhaps after the induction of lesions with greater effects on behavior, would be of interest. Our data also show that OTR4132 crossed the likely disrupted BBB, penetrated within lesioned tissues, and persisted in or near the lesion locus. Although the mechanisms of this neuroprotection remain to be elucidated, our findings suggest that OTR4132 may hold promises for the treatment of neurodegenerative disease and perhaps even invasive CNS damage. These CNS protective effects of OTR4132 using multiple injections (e.g. every 2 or 3 days) should be investigated, especially for stroke or AD treatments.