Investigation of alpl expression and Tnap-activity in zebrafish implies conserved functions during skeletal and neuronal development

Hypophosphatasia (HPP) is a rare genetic disease with diverse symptoms and a heterogeneous severity of onset with underlying mutations in the ALPL gene encoding the ectoenzyme Tissue-nonspecific alkaline phosphatase (TNAP). Considering the establishment of zebrafish (Danio rerio) as a new model organism for HPP, the aim of the study was the spatial and temporal analysis of alpl expression in embryos and adult brains. Additionally, we determined functional consequences of Tnap inhibition on neural and skeletal development in zebrafish. We show that expression of alpl is present during embryonic stages and in adult neuronal tissues. Analyses of enzyme function reveal zones of pronounced Tnap-activity within the telencephalon and the mesencephalon. Treatment of zebrafish embryos with chemical Tnap inhibitors followed by axonal and cartilage/mineralized tissue staining imply functional consequences of Tnap deficiency on neuronal and skeletal development. Based on the results from neuronal and skeletal tissue analyses, which demonstrate an evolutionary conserved role of this enzyme, we consider zebrafish as a promising species for modeling HPP in order to discover new potential therapy strategies in the long-term.


Materials and methods
Zebrafish maintenance. Zebrafish (Danio rerio) were bred and maintained as previously described [42][43][44] in the aquatic facilities of the Biocenter of the Julius-Maximilians-University Würzburg, Germany according to FELASA guidelines. In short, adult fish are kept at a mean temperature of 24-26 °C in 10 l glass and 2.5 l plastic tanks, while embryos younger than 120 h post-fertilization (hpf) are raised at a temperature of 28.5 °C in an incubator. A daily light cycle of 10 h dark/14 h light is maintained. Preconditioned reverse osmosis water with adjusted conductivity 800-1,100 µS/cm, pH7.0 and stable water hardness is used. A food combination of Artemia nauplii and GEMMA Micro Food (age dependent sizes; Skretting, USA) is standard.

RNA isolation, reverse transcription and quantitative real-time PCR (qPCR).
For qPCR experiments, RNA was extracted from adult zebrafish organs (age 3-6 month; AB/AB strain). For cDNA synthesis, 1 µg RNA was transcribed into cDNA and further analyzed in a ViiA7 Real-Time PCR System (Thermo Fisher Scientific, Waltham, USA). Each group and primer sample was analyzed via qPCR in triplicates on a single plate utilizing HOT FIREPol Eva Green Mix Plus (Solis BioDyne). Data analysis was performed via QuantStudio Real-Time PCS Software v1.1 by ΔΔCt method. Muscle tissue served as reference sample for relative comparison. Amplification of gapdh and eef1a1l1 was used as endogenous cDNA controls (housekeeping genes). Used primer pairs and targeted regions are noted in Table S1. Additional details on experimental set-up is given in the supplementary methods.
Synthesis of TNAP-specific riboprobes and in situ hybridization (ISH) to detect TNAP mRNA. ISH was performed according to standard protocols 45,46 . Proteinase K incubation and NBT/BCIP staining times were adjusted to age of embryos and to investigated tissues. RNA probes were synthesized from cloned partial mRNA sequences of alpl using the DIG or FLU RNA Labeling Kit (Roche, Basel, Switzerland). We used an alpl cDNA fragment of 434 bp size to synthesize specific anti-sense RNA probes. Sense probes were synthesized as negative control for each anti-sense probe and were used under the same reaction conditions. Additional details on experimental set-up is given in the supplementary methods.
Combined TNAP ISH and proliferating cell nuclear antigen (PCNA) immunohistochemistry on adult zebrafish brains. TNAP ISH was performed using the same TNAP-specific riboprobes as described for the embryonal stages subsequently followed by immunofluorescence (IF). PCNA-specific staining was performed using the primary antibody in a dilution of 1:200 (PCNA P8825, mouse monoclonal, Sigma-Aldrich, Antibody Registry nr. AB_477413) and anti-mouse IgG Alexa Fluor 488 (A-11001, Thermo Fisher Scientific, Antibody Registry nr. AB_2534069) secondary antibody in a dilution of 1:1,000. Finally, the brains were embedded using the JB-4 embedding kit (Polysciences, Warrington, PA, U.S.) according to manufacturer's instruction and the brain samples were sectioned in 4 µm slices. Additional details on experimental set-up is given in the supplementary methods. Brain regions were sketched, identified and named according to Wullimann et al. 47 .
ELF 97 staining (AP-activity staining). ELF 97 Phosphatase Substrate (Thermo Fisher Scientific), which is converted into a fluorescence precipitate by active AP enzymes, was used on adult brain cryosections of AB/ AB wildtype fish, according to the manufacturer's instructions with certain modifications. Specificity control of ELF staining was performed by TNAP inhibitor application and is shown in Fig. S15. Additional details on experimental set-up is given in the supplementary methods.
Incubation of zebrafish embryos with TNAP inhibitors. The embryos were collected directly after fertilization, the chorion was punctured using a glass needle, and treatment with the respective inhibitors was performed in a final concentration of 0

Results
The ALPL gene encoding the ectoenzyme TNAP (tissue-nonspecific alkaline phosphatase) is widely conserved across a wide number of animal species (Treefam family: TF323513; ENSEMBL gene tree: ENSGT00950000183063). For providing consistent differentiation between species as well as between gene and protein, we used the respective conventional denominations in our manuscript. The gene name is ALPL in human, Alpl in mouse, and alpl in zebrafish, whereas the protein is called TNAP in human and mouse but Tnap in zebrafish. Detailed information on the human, murine, and zebrafish orthologues are listed in Fig. 1A. Transcript alignment of the zebrafish alpl gene with the human and mouse sequences revealed an identity of 61% and 63%, respectively. In contrast to many other genes, the alpl gene is not duplicated within the zebrafish genome. Interestingly, the peptide alignment resulted in 67% overall identity to mouse and humans. Moreover, the alignment showed a remarkable high conservation level of 79% identity and 86% similarity in amino acid sequences of the zebrafish and the human alkaline phosphatase domain. Additional synteny analysis further revealed that genes located in close proximity to the respective ALPL, Alpl, and alpl locus were showing a high level of genetic conservation between human, mouse, and zebrafish on the respective chromosomes 1, 4, or 11 ( Fig. 1B; Genomicus v98.01 database). Taken together, this in silico data shows genetic conservation between vertebrate species and implies retained Tnap/TNAP enzyme activity.
Determination of the spatio-temporal expression pattern of alpl in zebrafish embryos. In order to determine the spatio-temporal expression pattern of alpl, initial whole mount ISH with alpl-specific probes was performed on zebrafish embryos at different developmental stages (from 3 somites to 48 hpf; Fig. 2). www.nature.com/scientificreports/ ISH showed the first alpl expression at early somitogenesis stages (11 hpf) within the midbrain-hindbrain boundary (MHB) region of the neural keel ( Fig. 2A). During subsequent stages of zebrafish development, alpl expression was detected in more distinct areas: the MHB, the mesencephalon, the eyes (including retina and lens), the pectoral fin, the heart field, and the tail bud (Fig. 2B). The alpl expression in somites at later stages between 22 and 48 hpf was restricted to posterior parts of the trunk and was temporally tightly regulated ( Fig. 2B; sense controls are shown in Fig. S2). For getting a more detailed impression of alpl expression within different tissues, sections of 48 hpf zebrafish embryos were analyzed (Fig. 2C). Low expression signals were detected in the developing neural tissues: diencephalon, optic tectum, as well as at the MHB. However, strong alpl expression was prevalent in the dorsal aortic root, the pectoral fin, the nephros, the retina, and the lens. In summary, expression analysis in the developing zebrafish embryo indicates tight regulation of alpl expression during embryonic stages and aligns with previously published expression patterns in other vertebrates (see Discussion section, Table 1).

Analysis of endogenous alpl expression and Tnap-activity in adult zebrafish tissues.
To determine the endogenous alpl expression and Tnap-activity in adult zebrafish, different organs of adult zebrafish were analyzed using qPCR (Fig. 3A) and AP-activity assay (CSPD-assay; Fig. 3B).
The qPCR data was complemented with measurements of enzymatic Tnap-activity within several organs isolated from adult zebrafish (normalized to the activity levels in the non-treated muscle samples; Fig. 3B). The highest relative AP-activity (Fig. 3B, black bar, w/o levamisole) was detected in the skin (3.57) and the eyes (3.55). Additionally, a medium level of relative AP-activity was observed in the heart (1.91), in the nephros (1.19), and in the gills (1.44), directly followed by the intestine (1.12) and the liver (1.03). Low relative AP values were observed in whole brain samples (0.53) of adult zebrafish. In order to reveal the portion of Tnap-activity included in the total AP-activity, protein lysates were additionally treated with different concentrations of the Tnap-specific inhibitor levamisole, in a range between 1 and 10 mM. Interestingly, levamisole treatment was able to significantly decrease the AP-activity in all samples depending on the applied concentration, except for intestine and liver samples (Fig. 3B, dark to light grey bars). Comparing the mRNA expression level with the detected enzymatic activity, low levels of alpl mRNA but high Tnap-activity values were detected in skin samples, while high mRNA expression but low Tnap-activity were detected in whole brain samples. Apart from these, the alpl expression and Tnap-activity were corresponding to each other.  www.nature.com/scientificreports/ Basic alpl expression and Tnap-activity analysis in adult zebrafish brain. Besides embryonic stages and whole brain samples, we analyzed the alpl expression pattern in distinct adult zebrafish brain regions by using qPCRs (Fig. 4B) and ISH (Fig. 4C).
For a detailed analysis of alpl expression, adult brains were dissected into four different regions, namely telencephalon (region 1: olfactory bulbs, telencephalon and otic nerve), cortex (region 2: tectum opticum and torus longitudinalis), ventral mesencephalon (region 3: tegmentum, diencephalon, hypothalamus), and rhombencephalon (region 4: medulla oblongata, cerebelum) as schematically depicted in Fig. 4A. The results of the respective qPCR analysis (RQ calculated in comparison to the expression levels of whole brain samples; cDNA levels normalized to the expression of gapdh and eef1a1l1; Fig. 4B) revealed the highest expression level of alpl in the telencephalon (RQ: 10.04), followed by the cortex (RQ: 7.20), the mesencephalon (RQ: 4.48), and finally the rhombencephalon (RQ: 1.38). Additionally, samples from eyes were analyzed as a positive control. Subsequently, ISH was performed on whole adult zebrafish brains (Fig. 4C). The alpl signals showed restricted expression in the telencephalon, the optic tectum, and the hypothalamus. Control samples treated with an alpl sense probe, did not reveal any specific signals within these regions. Aberrant signal was visible in the brain treated with the sense control, but these signals were strongly differing from the actual ISH signals and can therefore be considered as nonspecific and not interfering with the results of the ISH. Vibratome sections of the stained brains further clarified alpl expression in the dorsal diencephalon, the lateral recess, the ventral anterior thalamic nucleus, the hypothalamus, and the peripheral growth zone.
Furthermore, we performed fluorescence analysis on JB-4 embedded plastic sections in order to increase ISH specificity and resolution of alpl expression analyses within the adult brain. Conventional NBT/BCIP alpl www.nature.com/scientificreports/ ISH signals were visualized by its near infrared fluorescence 50 and sections were counterstained with DAPI as a nuclear marker (Fig. 5A). In sagittal sections, alpl expression was detectable in the dorsolateral zone of the telencephalon (Fig. 5A, magnification A1). The transverse sections confirmed alpl expression in the lateral zone of the dorsal telencephali (Dl; Fig. 5A, magnifications A2, A3 and A5) and in the ventricular zones, e.g. the telencephalic ventricle (TelV; Fig. 5A, magnification A4). To correlate alpl expression and the enzymatic function of Tnap within the adult zebrafish brain, we performed ELF T 97 staining on corresponding telencephalon cryosections (Fig. 5B). Strong Tnap-activity could be visualized broadly within the telencephalon in the central, dorsal and lateral dorsal zones (Dc/Dd/Dl). Subsequently, adult zebrafish brains marked for alpl expression were co-stained for PCNA via ISH/IF. Transverse sections were prepared at three different levels within the telencephalon and revealed alpl expression in close proximity or co-localized to PCNA positive cells (Fig. 5C, magnifications C1 to C3, green arrowheads). Moreover, double positive cells were detected at the border of the medial zone of the telencephalon (yellow arrowheads). Additional JB-4 sections were prepared at different levels of the mesencephalon, a second brain region showing high alpl expression (Fig. 6). Six different cutting planes in the area, including the mesencephalon, diencephalon, and hypothalamus were initially investigated (Fig. 6A) and implied that alpl expression is predominantly prevalent in the ventricular areas (DiV, TeV; Fig. 6A, magnifications A2, A3, A5 and A9), the valvula cerebelli/ torus longitudinalis (Vam/TL, Fig. 6A, magnifications A4, A6 and A7), the dorsomedial optic tract (DOT; Fig. 6A, magnification A1), the caudal hypothalamus (HC; section 5), and in blood vessels (Bv, Fig. 6A, magnification A8). ELF 97 staining was again performed to visualize different areas of AP activity (Fig. 6B), most prominently within the torus longitudinalis (TL), the hypothalamus (TH/IL), and in single regions of the medulla oblongata   www.nature.com/scientificreports/ (MO). Combined alpl ISH and PCNA IF confirmed that alpl was expressed in cells of the ventricular regions, either within or surrounding PCNA positive cells ( Fig. 6C; magnifications 3A-3F). In the DiV, the PCNA expressing cells were located very closely to the alpl positive cells (Fig. 6C, magnifications 3A and 3D) and sometimes showed overlapping patterns. More cells that expressed both, alpl and PCNA, could be detected besides the DiV also in the area between the tectum opticum (TeO) and the commissura tecti (Ctec; Fig. 6C, magnifications 3B and 3E). A high number of PCNA positive cells were located in the lateral recess (LR), a ventricle embedded in the hypothalamus (TH/IL; Fig. 6C, magnifications 3C and 3F). Close to this specific region, further alpl positive cells were detectable. In addition, analysis of the rhombencephalon was analogously performed, but only weak alpl expression was detected by ISH in this area (data not shown).
In summary, the analysis of the alpl expression and Tnap-activity pattern in the adult zebrafish brain revealed, that the highest prevalence of Tnap is localized to the telencephalon, here especially to the dorsal and lateral regions. Additionally, alpl expression is, among others, prevalent in the hypothalamus, the thalamus, and interestingly also in cells of the ventricular regions, which are important zones for adult neurogenesis. Our own, novel data have been compared to previously published expression patterns in primates, mice and zebrafish in Table 1 in the discussion section.
Relevance of Tnap in the early development of the nervous system and the skeletal structure of the skull. In order to investigate a possible influence of Tnap inhibition on the early development of the nervous system and the skeleton in zebrafish, embryos were treated with the Tnap-specific inhibitor levamisole at different concentrations (0.5, 1 and 10 mM). After morphological analysis of the embryos within the first 48 h (Fig. 7A), functional consequences of decreased Tnap-activity on early nervous system development were analyzed. Axonal outgrowth within the optic tectum was visualized by IF staining for acetylated tubulin at 48 hpf and investigated with a confocal microscope (Fig. 7B). Moreover, the craniofacial cartilage and mineralization progress of levamisole treated embryos was analyzed at 120 hpf, using combined cartilage and mineralized tissue staining (Alcian-blue and Alizarin-red; Fig. 7C).
Levamisole at low doses (0.5 and 1 mM) had no or only a slight impact on the regular morphological development of the treated zebrafish embryos at 24 and 48 hpf. Abnormal developmental effects, aberrations to untreated/ DMSO control embryos and increased lethality were concentration dependent and increased in intensity with higher levamisole concentration ( Fig. 7A; data not shown). At 10 mM, altered rates of blood flow, enlarged pericardial sacs, impaired normal heart, and vascular development were observed in the majority of 48 hpf embryos investigated (red arrow in Fig. 7A). The analysis of axonal growth after levamisole treatment (48 hpf) indicated aberrant axonal outgrowth of neurites in the optic tectum even at a low concentration of 1 mM levamisole (marked with red arrows; Fig. 7B). In General, increasing the levamisole concentrations led to more pronounced effects on the axonal outgrowth.
Furthermore, cartilage and mineralized tissue staining of levamisole treated embryos revealed aberrant craniofacial development (see Fig. 7C). Interestingly, especially mineralized structures were deteriorated in embryos treated with 1 mM and were completely absent in embryos treated with 10 mM levamisole (120 hpf; Fig. 7C, red arrows). The respective control embryos and the embryos treated with the lowest inhibitor concentration of 0.5 mM levamisole developed normal mineralized structures at this developmental stage. Simultaneous cartilage staining with Alcian-blue revealed morphological changes of the developing skull structure, which were statistically evaluated for preparations of the viscero-and neurocranium (Figs. S5 and S6; n = 10 for each group). Most significant changes were detected in measurements of branchial arch and anterior-posterior length within embryos treated with 10 mM levamisole. Treatment of zebrafish embryos with the alternative Tnap inhibitor MLS-0038949 showed very similar results for the development of the nervous system and the cartilage/mineralized tissue, what further confirmed the levamisole results ( Fig. S13 and Fig. S14).
Zebrafish embryos were additionally microinjected with alpl Splice and ATG Morpholinos for knockdown experiments and for recapitulating the indicated influence of Tnap during the development of the nervous system and mineralization of the zebrafish (Fig. S3 and Fig. S4). Both Morpholinos led, analogous to the effects of the inhibitors, to distinct morphological changes depending on the respective concentration, e.g. altered rates of blood flow and an enlarged pericardial sac (Fig. S4A). Acetylated tubulin IF staining confirmed aberrant axonal growth in the optic tectum of alpl morphants at 48 hpf (Fig. S4B). The analysis of the craniofacial structures at 120 hpf confirmed morphological changes of cartilage structures and a missing development of mineralized structures after alpl knockdown (Figs. S4C). The changes of the cartilage structures were statistically evaluated, analogous to the analysis of levamisole treatment (Fig. S7, S8, S9, and S10; n = 10 for each group) and were compared to control Morpholino injected embryos (Figs. S11 and S12). Most significant changes were again detectable in measurements focusing on branchial arch and anterior-posterior length, but also measurements of the skull width were significantly altered. Additionally, mineralization of teeth, operculum, and cleithrum was limited in embryos treated with Tnap inhibitors and Morpholinos at 5 dpf (Fig. S14).
Taken together, our data reveals that zebrafish alpl expression and Tnap-activity can be detected during zebrafish embryonic development, in different adult organs, and in distinct adult brain regions. Tnap inhibition and alpl knockdown result in specific and comparable developmental defects, e.g. interference with axonal growth and mineralization. Moreover, the presented data is in accordance with descriptive and functional observations made in other vertebrates and therefore implies conservation of Tnap function also in zebrafish.

Discussion
Although the molecular mechanisms of HPP-related deficits in bone formation and mineralization offer plausible explanations for the predominant skeletal phenotype, the molecular reasons behind a considerable number of HPP symptoms are to date not fully understood. These include the development of craniosynostosis in an Scientific RepoRtS | (2020) 10:13321 | https://doi.org/10.1038/s41598-020-70152-5 www.nature.com/scientificreports/ actually repressive environment for mineralization processes and the increased risk for depression and/or anxiety disorders in HPP patients 51,52 . The high level of genetic conservation of the zebrafish alpl compared to other vertebrate orthologues, like human ALPL, implies that zebrafish might be suitable for functional analyses within bone and nervous system development 20,21,26 . Our presented zebrafish study now provides initial data on alpl expression patterns, describes functional consequences of Tnap inhibition during embryogenesis, and sets the scene for the establishment of a novel animal model in order to study HPP. Initially, we investigated the endogenous expression of the alpl gene and of Tnap's biochemical function using qPCR, AP-activity assays as well as whole mount ISH in embryonal stages and in adult zebrafish. During early segmentation stages of embryo development, alpl mRNA is predominantly detected in the MHB, the precursor  www.nature.com/scientificreports/ structures of the eyes, and in the tail bud. In subsequent developmental stages until 48 dpf, alpl is expressed in the MHB, the diencephalon, the mesencephalon, the lens, the retina, the heart, the nephros, the tail bud, the somites, and the pectoral fin. Apparent expression pattern changes can be observed within somites over time, which implies tight regulation of alpl expression during embryogenesis. Interestingly, the pectoral fin is one of the earliest developing skeletal structures expressing alpl and precedes many other bony structures. Therefore, the detection of the alpl-specific signals in this precursor tissue might foreshadow its conserved role in regulation of the phosphate metabolism during mineralization in extremities 7 . Furthermore, our experiments affirm that alpl mRNA and Tnap-activity are detectable in almost all analyzed adult tissues, apart from the intestine and the liver, which do not respond to the Tnap-specific inhibitor levamisole. The lack of responsiveness to levamisole leads to the assumption that those organs express an intestinal (alpi.1 or alpi.2) instead of the tissue-nonspecific isoform of the AP (alpl). Additionally, the highest alpl expression levels along with strong AP-activity can be observed in the eyes, more precisely in the lens and the retina, which seems to be conserved between different vertebrate species and is in line with previously published literature 15 . However, no specific symptoms concerning the visual performance of HPP patients have been described so far and the exact role of TNAP/Tnap in the visual system and potential compensatory mechanisms remain to be unraveled in future experiments.
For detailed investigation of the nervous system related symptoms of HPP and for confirmation of qPCR and CSPD-assay results, the endogenous expression of alpl was additionally analyzed in dissected and in sections of adult zebrafish brains. The results show that alpl expression is prevalent in the telencephalon (especially in the dorsal and lateral telencephalic area), the tectum opticum, and in the hypothalamus (lateral recess, ventral anterior thalamic nucleus, peripheral growth zone). Moreover, we can frequently detect alpl expression in ventricular zones of the brain (e.g. TeIV and TeV). The alpl expression pattern implies restricted Tnap-activity in the zebrafish brain, but staining for ELF 97 shows rather wide areas of Tnap-activity in the telencephalon, in the hypothalamus, and in the torus longitudinalis. Proteolytic shedding processes might lead to cleavage and spreading of active Tnap enzyme within brain compartments and subsequently result in tissue-wide activity. Release of TNAP from the cellular membrane via degradation of its phospholipase mediated GPI-anchor has already been reported and might be one molecular mechanism for this observation 53,54 . Nevertheless, it is unclear to what extent those processes might lead to the observed discrepancy between the different staining methods. Another interesting observation are strong ISH signals within the dorsomedial optic tract (DOT), a white matter structure comprising predominantly of retinal ganglion cell axons. Localization of TNAP to axons of cultured hippocampal neurons has been previously reported and TNAP-activity is crucial for correct axonal growth by ATP degradation 10 . Another explanation for the discrepancy between ISH and ELF 97 data may further result from different sensitivity thresholds of the respective staining method.
The proliferation marker PCNA is partly co-localized or localized in close proximity to the alpl expressing domains in the brain sections. In total, 16 proliferative zones have already been described in the adult zebrafish brain, which are predominantly located at the ventricular surface, and incorporate neuronal stem cells 55 . Compared to zebrafish, the mammalian brain shows only limited capability of adult neurogenesis. Novel formation of neurons in adult brains is only possible in the subventricular zone (SVZ) of the lateral ventricles and the subgranular zone of the hippocampal dentate gyrus in humans 56 . Interestingly, Tnap-activity is indeed located in the ventricular and SVZ of mice including zones of adult neurogenesis 57 .
Prominent ALPL signals can be verified in specific brain regions of different mammals and in layers across the neocortex (occipital-, frontal-, and temporal lobe), for example in layer 4 of the primary visual and somatosensory cortices. Our zebrafish results are comparable to expression patterns observed in mammals and indicate that alpl transcripts can not only be detected during embryogenesis, but also in specific brain structures of adult animals. Generally, ALPL expression levels are higher during vertebrate development compared to mature organs, tissues (e.g. cerebral neuronal networks), or the complete animal 12,13,18,58 . The results of our own study and literature data on the ALPL, Alpl, and alpl expression in adult brains are summarized in the Table 1. www.nature.com/scientificreports/ The expression of alpl and inhibition of TNAP-activity at very early embryonic stages of the zebrafish implies that the enzyme plays an important role during initial steps of development. This assumption is further promoted by patients suffering from genetically determined severe TNAP deficiency and by homozygous Alpl knockout mice, which often die very shortly after birth 65,66 . The inhibition of Tnap-activity, either with chemical components (levamisole and MLS-0038949) or by knockdown of alpl expression using Morpholinos, impairs the development of cartilage/bone as well as the nervous system in zebrafish 67 . Considering the known side effects of levamisole, like ion channel inhibition 68 , the investigation of MLS-0038949 and alpl knockdown was indispensable in order to confirm the results. Our studies with the alternative inhibitor MLS-0038949 and with two alpl Morpholinos affirm the levamisole results to the greatest extent. However, we observe differences between the two inhibitors, as MLS-0038949 is effective at much lower doses and displays a stronger effect on axonal growth even at 10 µM. Levamisole needs to be prevalent in higher doses for effective blocking and shows more prominent effects on bone development over time. Our results clearly show that both inhibitors and Morpholino knockdown are able to repress Tnap-activity, but differ in their influence on neuronal and skeletal development. These results imply different modes of TNAP functional reduction and might point to different biochemical consequences. TNAP/Tnap's promoting role in hydroxyapatite crystallization is mediated by diminishing the amount of prevalent mineralization inhibitor PP i in the extracellular space during in vitro osteogenic differentiation 69 . While axonal outgrowth is initiated by preventing the activation of the purinergic receptor of P 2 X 7 7,10 . Overall, the observed effects of Tnap inhibition on early neuronal and bone development indicate a rather broad spectrum of Tnap efficacy, including effects on dephosphorylation of general substrates, such as PPi, ATP/ADP/AMP, and phosphorylated osteopontin 1,9,70 . Despite the low level of interference of the Splice Morpholino with the wildtype gene product and general concerns with Morpholino usage, the data are supporting our results gained with the Tnap inhibitors levamisole and MLS-0038949. However, we have to point out that, deviant of the Stanier et al. guidelines, no mutant lines have been generated yet, which would be essential for a complete characterization of the phenotype. The establishment of transgenic zebrafish lines has to be done in future research projects. Nevertheless, we think that we can trust the inhibitor and Morpholino data as they are confirming an influence of decreased Tnap-activity on mineralization processes and neuronal development, which is additionally going along with the phenotype of HPP patients.
In summary, our study identifies the zebrafish as a suitable model for the rare disease HPP. The analysis of endogenous alpl expression in zebrafish as well as first experiments on functional consequences of decreased alpl expression/Tnap inhibition on the embryonic development of bones and the nervous system pave the way for consecutive analysis of transgenic zebrafish lines in the future. These will help to further elucidate so far unsolved molecular mechanisms that are underlying certain HPP symptoms. One unsolved symptom to be investigated is how the paradox situation of premature suture fusion in the developing skull can be reconciled with an environment of hypomineralization in the rest of the body. Moreover, future studies have to investigate how neurological symptoms, like depression or anxiety disorders, can indeed robustly be linked to alpl deficiency in single brain regions or to specific neurons 51,71,72 . An important role of TNAP/Tnap during the processes of synaptogenesis, myelination, neurotransmitter metabolism, and axonal outgrowth has already been proclaimed, but has not been investigated in zebrafish 8,10,12,13 . Therefore, those cellular developmental functions definitely have to be considered for establishing a complete picture of TNAP's function in the nervous system of zebrafish by applying cell-type-specific methods. Generally, our results of the embryonic and adult alpl expression analysis in zebrafish suggest a possible correlation to the pathology of the rare disease HPP and open up new perspectives concerning the neurological consequences of TNAP/Tnap dysfunctions.