LRRK2 kinase activity is necessary for development and regeneration in Nematostella vectensis

Background: The starlet sea anemone, Nematostella vectensis, is an emerging model organism with a high regenerative capacity, which was recently found to possess an orthologue to the human LRRK2 gene (nvLRRK2). The leucine rich repeat kinase 2 (LRRK2) gene, when mutated, is the most common cause of inherited Parkinson’s Disease (PD). Its protein product (LRRK2) has implications in a variety of cellular processes, however, the full function of LRRK2 is not well established. Current research is focusing on understanding the function of LRRK2, including both its physiological role as well as its pathobiological underpinnings. Methods: We used bioinformatics to determine the cross-species conservation of LRRK2, then applied drugs targeting the kinase activity of LRRK2 to examine its function in development, homeostasis and regeneration in Nematostella vectensis. Results: An in-silico characterization and phylogenetic analysis of nvLRRK2 comparing it to human LRRK2 highlighted key conserved motifs and residues. In vivo analyses inhibiting the kinase function of this enzyme demonstrated a role of nvLRRK2 in development and regeneration of N. vectensis. These findings implicate a developmental role of LRRK2 in Nematostella, adding to the expanding knowledge of its physiological function. Conclusions: Our work introduces a new model organism with which to study LRRK biology. We show a necessity for LRRK2 in development and regeneration. Given the short generation time, genetic trackability and in vivo imaging capabilities, this work introduces Nematostella vectensis as a new model in which to study genes linked to neurodegenerative diseases such as Parkinson’s.


Background
The starlet sea anemone Nematostella vectensis is a member of the Cnidarian family [1,2], and is an emerging model for studying development and regeneration due to its ease of maintenance in the lab, sequenced genome and genetic tractability [1,[3][4][5].Cnidaria are considered the sister group to the Bilateria, putting them in an excellent position to study the evolutionary trajectory of gene families.Animals from the Cnidarian phylum possess a strikingly similar gene content to humans compared to better studied ecdysozoan models (such as Drosophila melanogaster) and possess syntenic genetic loci [6][7][8][9].Of particular interest, N. vectensis possesses 4 LRRK genes, one of which being an orthologue to the human LRRK2 gene [10].
Being composed of two cellular layers, the ectoderm and the endoderm, Nematostella have a simple body plan and tissue organization [11][12][13][14].Additionally, Nematostella possess a nervous system that takes the form of a nerve net, composed of sensory cells, glandular cells, multipolar ganglion cells and cnidocytes (stinging cells) [4,15,16].
Unlike bilaterians, the nervous system of cnidarians is not centralized, and represents an ancestral, simple organization [17,18].Nonetheless, several conserved neurogenic pathways and proteins mirroring what is seen in bilaterians highlights their potential usefulness for studying the underlying neurobiology of neurodegenerative diseases like Parkinson's Disease (PD) [1].For example, a recent study by Steger et al [8] identi ed a key upstream regulator of Nematostella neuroglandular lineages (composed of neurons, cnidocytes and gland cells): SoxC, which has similar roles in bilaterians [19][20][21].Prior studies have also demonstrated homologous transcription factors (such as Hox genes) that are involved in neurogenesis, as well as numerous neuropeptides [22,23].It is thought that there are 32 potential neuronal cell types in the Nematostella, emphasizing the unusual complexity of the nervous system in this organism [24][25][26].
A key feature of N. vectensis is their remarkable ability to regenerate after injury via cellular proliferation [27][28][29][30][31]. Nematostella can be cut into multiple pieces, and each piece will give rise to a new animal, the exception being that if the foot alone is amputated it cannot regenerate [3,28,30].There has long been an interest in whether or not regeneration reuses the same genes that were developmentally deployed.
Recent work from the Rottinger lab has compared gene expression pro les during early development and regeneration and developed a user-friendly website that allows scientists to view the expression of their genes of interest during development versus regeneration in Nematostella vectensis [3].
LRRK2 is highly studied due to its aforementioned link to PD, with rare autosomal dominant coding mutations as well as more common non-coding variation at the LRRK2 locus linked to disease risk [32].The LRRK2 gene is located on human chromosome 12p11.2-q13.1, and encodes a large 2527 amino acid protein [33][34][35].Belonging to the Roco family of Ras-GTPases, LRRK2 is composed of multiple domains (Fig. 1) which give rise to multifunctional biology [36,37].
Mutations in LRRK2, including N1437H, R1441C, Y1699C, G2019S and I2020T, cluster in the enzymatic domains of the protein and increase LRRK2s kinase activity, with dysregulation of a subset of Rab GTPases a key consequence of this [38].How this alteration in enzymatic function leads to neurodegeneration is unclear, with studies implicating LRRK2 in a range of cellular processes including lysosomal function, mitochondrial biology and synaptic signaling.The tight association between LRRK2 and Parkinson's disease has led to a number of in-human clinical trials for both small molecule kinase inhibitors and antisense oligonucleotide gene therapy [39] .
Intriguingly, recent work has identi ed a role for LRRK proteins in stem cell proliferation and tissue remodeling in intestinal enterocytes during whole-body regeneration in the bilaterian freshwater planaria, the Schmidtea mediterranea [40] highlighting this as a potential phenotype of interest Nematostella vectensis.
In this study we have examined the cross-species conservation of LRRK2 and investigated the functional importance of LRRK kinase activity in Nematostella development and regeneration, demonstrating a key role for LRRK2 in these processes.

Animal care
Nematostella vectensis were maintained at 17-20°C in Pyrex glass bowls kept in the dark in 15 parts per thousand Instant Ocean [41].Animals were fed 48-hour old artemia, ve times per week.Animals were cleaned a few hours after feeding.Spawning was induced by exposure to light and an increase in temperature to 23-25°C, and embryos were collected immediately after spawning.

Bioinformatics Evolutionary analysis
To determine the suitability of N. vectensis as a model for studying LRRK2-related PD, an evolutionary analysis of LRRK2 genes from different species was carried out.Multiple sequence alignment was carried out using Clustal Omega [42,43].The FASTA sequences obtained from National Centre for Biotechnology Information (NCBI) databases [44] were inputted into the program and a phylogenetic tree was generated.

Identifying conserved residues
The open reading frames (ORF) of nvLRRK2 and hsLRRK2 were aligned, and the Needleman-Wunsch algorithm was used via the BLAST tool [45,47] to carry out a pairwise comparison [48].This provided information including the similarity (amino acids with the same or slightly different side chains) and identity (amino acids in the sequence that are identical).The most common mutated residues in human LRRK2 pertaining to PD were distinguished [49][50][51], and the corresponding residues in nvLRRK2 were identi ed.Residues that are conserved have the same amino acid, and those that are not conserved do not.The same technique was used for identifying conservation of phosphorylated residues.

nvLRRK2 kinase inhibition
To obtain embryos, N. vectensis were spawned by exposure to increased temperature and light.The concentration of the kinase inhibitors used for the embryo experiments was 5µm, and for the adult experiments it was 10µm.Some experiments were carried out on Nv-LWamide transgenic animals that were expressing mCherry protein their neurons [18].Embryos were added to a 6-well plate and were incubated in a solution of 5ml sea water and 250µl of each inhibitor.The control embryos were incubated in seawater.For regeneration experiment wild type or transgenic adult Nematostella were relaxed in 7.4% MgCl2 (ThermoFisher) for 15 minutes and were subsequently amputated below the pharynx on a plastic petri dish using a sterile no.10 disposable scalpel (World Precision Instruments).The animals were monitored every 2 days and the solutions were changed.Whole animals were fed artemia to ensure any differences were not due to lack of food.
Upon completion of this experiment, the organisms were relaxed in 7.4% MgCl2 (ThermoFisher), and then in 10% MgCl2, 15% MgCl2 and 20% MgCl2.The animals were then xed in 4% paraformaldehyde (PFA) (ThermoFisher) at 4°C and were subsequently stored at 4°C until imaged.The transgenic animals were imaged using a Zeiss LSM780 microscope.

qRT-PCR analysis
Embryos were ash frozen on liquid nitrogen, approximately 150 embryos were used for each sample.
RNA isolation was carried out following Invitrogen phenol-chloroform RNA extraction protocol.cDNA was then synthesized using iScript™ cDNA Synthesis Kit (BioRad).qRT-PCR reactions were prepared using SsoAdvanced Universal SYBR® Green Supermix (BioRad) and carried out on Real-Time PCR (CFX Opus 96) (BioRad).Nematostella were collected at different time points after fertilization, xed in 4%PFA overnight at 4C.Embryos or young larval animals were dehydrated in methanol series and stored at -20C.Before starting the in-situ protocol animals were rehydrated again through a methanol series back into phosphate buffered saline (PBS).In situs were carried out using published protocols [52].In summary, embryos were incubated in 1:1 PBST:Hyb for 30 minutes and pre-hybridized for 30 minutes.NvLRRK2 probe was diluted into hybridization buffer and slides were allowed to hybridize overnight at 55°C.The following day the animals were washed 3 times in Wash Buffer (50% Formamide, 5x SSC and 0.1% Tween) for 30 minutes each, once in 1:1 Wash Buffer:PBST for 30 minutes and once in PBST at 55°C.Samples were rinsed in room temperature PBST 3 times for 5 minutes each before blocking buffer was added (2% goat serum, 2% BST in PBSTx) for 1 hour.Anti-DIG F AB (Roche) was diluted 1:1000 in blocking buffer and slides were incubated for at least 1 hour.Samples were washed 3 times for 10 minutes each before addition of fresh AP Buffer for at least 10 minutes.Finally, samples were incubated in BM Purple (Roche) until colored reaction was observed.The reaction was stopped by several quick rinses in PBS and were xed in 4% PFA for 10 minutes.Samples were mounted in 80% glycerol and images were taken with a Zeiss Discovery V8 microscope using Zen software.

Statistical analysis
Statistical tests were performed using the Graphpad Prism 9 software.A D'Agostino and Pearson test was carried out to test for normality.Upon nding a non-normal distribution, a non-parametric Kruskal Wallis test was used to test for statistical signi cance between the medians of each measurement in the different experimental groups.A Dunn's post-hoc multiple comparison test was used to compare the medians of the experimental groups against the control group
The animalia display varying numbers of LRRK genes, ranging from 1 to 4 copies (Fig. 1A).In order to de ne the relationship between Nematostella nvLRRK2 and hsLRRK2, an evolutionary analysis into conservation of this gene was carried out.A phylogenetic tree using data from NCBI [44] and Clustal Omega [42,43] was generated (Fig. 1B) to assess the relationship between representative LRRK genes across the animalia.
We then went on to compare the domain structure of N. vectensis and human LRRK2 to further determine the suitability of the Nematostella as a model for LRRK2-related PD (Fig. 1C), using publicly available data sets from NCBI [44].Consistent with conservation of LRRK proteins across evolution, these orthologues display a similar domain organisation, besides nvLRRK2 being a larger protein and lacking N-terminal armadillo repeats [62].A Needleman-Wunsch algorithm was used to align the protein sequences to observe any biological similarities or differences [47].This generated a score of 29% identity and 48% similarity across the open reading frame.
Parkinson's Disease-causing mutations in hsLRRK2 are localized to the ROC, COR and kinase domains of the protein.As such, these domains in nvLRRK2 were compared to the human equivalent under the same global sequence alignment tool, the results of which are shown in Fig. 1D.The scaffolding domains, Ankyrin and LRR, were also compared for a thorough comparison.The kinase domains were highly conserved, including the key kinase DYGI motif [63].In contrast, the scaffolding and ROC/COR domains are less well conserved.

Mutated residues are conserved in N. vectensis
To assess functional conservation across hsLRRK2 and nvLRRK2, residues that are mutated in human LRRK2 contributing to PD development were identi ed.These include R1441C/G, and N1437H (localized to the Roc domain), Y1699C (in the COR domain), G2019S and I2020T (in the kinase domain) [64-69].Residues that are functionally signi cant and have been evolutionary conserved, are thought to have more of an impact in terms of disease status [70].Figure 1E demonstrates that shows the most common PD associated LRRK2 mutation, G2019S, as well as the I2020T mutation residing in the same motif, is conserved in nvLRRK2.
Several of the amino acids surrounding these residues are also conserved, suggesting that these regions have not diverged throughout evolution despite selection pressures because of functional signi cance.The kinase domain appears to be the most well-conserved region of the protein, and the DYGI motif where these mutations occur are conserved across kinases, highlighting this as a critical region for kinase function [71].In contrast, mutations in the ROC domain are less well-conserved, which mirrors the earlier analysis of a lower similarity score between nvLRRK2 and hsLRRK2 ROC and COR domains.

Residues involved in post-translational modi cations are not well conserved
To assess conservation of post-translational regulation between human and Nematostella LRRKs, residues previously reported to be involved in signal transduction were compared.Phosphorylation sites on LRRK2 are involved in regulating function and downstream signaling events [72].Characterized phosphorylation sites in hsLRRK2 include S910, S935, S973, S955 located between the Ank and LRR domain, S1292 between LRR and Roc, and T2031, S2032 and T2035 in the kinase domain [72].
Unlike the mutated residues, the phosphorylated residues were not well conserved in both the LRR domain and the kinase domain.Figure 1F shows the sequence alignment demonstrating non-conserved residues as well as a conserved site.Only the S973 residue in the LRR domain and T2035 in the kinase domain were conserved in nvLRRK2, suggesting that the mechanisms involved in regulating signal transduction have diverged between these two proteins over evolution.
Scaffolding and structural domains in LRRK2 are less well conserved throughout species due to the lack of dependence for the functioning of the protein.Thus, is it unsurprising that these residues were not well conserved in nvLRRK2, and it suggests that post-translational modi cations have perhaps drifted throughout evolution to suit higher or lower species depending on their functional requirements.This mirrors other studies into global phosphoproteomics that suggest phosphorylation events evolve individually between different species [73].

NvLRRK2 kinase activity is necessary for faithful embryonic development
To investigate a role of the kinase activity of NvLrrk2 in regulating the development of N.vectensis, anemones were treated with structurally distinct validated inhibitors of LRRK2 kinase activity GNE-0877 and MLi-2.Freshly laid embryos were incubated in 5µM of respective inhibitor diluted in Nematostella water, while control embryos were grown in just water.The concentration of inhibitor to use was determined empirically by rst testing concentrations in the range of 1-20µM, 5µM was chosen as for both inhibitors most of the embryos survived and showed a phenotype, while in higher concentrations a large percentage of the embryos died within a few days.Embryos incubated at 5µM developed through the rst phases of development and reached the motile stage at relatively the same frequency as the control embryos.However, by 1 week post fertilization clear differences were observed, usually at this timepoint the embryos have developed tentacles and will commence feeding.Animals where the kinase activity of LRRK2 has been inhibited were overall smaller in length (Fig. 2) and most showed no development of tentacles or stunted growth of tentacles (Fig. 2) in comparison to controls that have much longer bodies and clear tentacles.Overall, the inhibitor treated animals display stunted inhibited growth from around 4 days post-fertilization onwards making it very di cult to determine if they developed normal organs like the pharynx and mesenteries.However, as all LRRK kinase inhibitor treated animals displayed stunted growth of tentacles or no tentacle growth we next examined expression of genes involved in specifying the oral region of the animal, Wnt4.Wnt4 is expressed early in development from the planula stages and is important for induction of genes involved in specifying the oral region of the embryo [31,[74][75][76][77].Here we nd that at 48 hours post fertilization (hpf) in embryos exposed to the LRRK inhibitor that levels of Wnt4 are signi cantly decreased in comparison to the control embryos, suggesting that the embryos are not inducing the genes necessary to specify head and to ultimately direct the embryos towards making tentacles (Fig. 3A).As the tentacles are formed early in development and are highly innervated, we also examined induction and formation of neurons.NvSoxB is wellcharacterized to be expressed early in development in the cells that will form neurons and nematocysts [78,79].Quantitative analysis of NvSoxB levels in control embryos versus the LRRK2 kinase inhibitor treated animal discovered a signi cant lack of expression of NvSoxB in inhibitor treated animals (Fig. 3B).We next examined the presence of development in the Nematostella embryos taking advantage of the NvLWamide-like::mCherry transgenic reporter line [18].By one week post fertilization young animals have differentiated neurons with a complex network of axons in the body and tentacles as seen in Fig. 3C.In comparison in animals exposed to the LRRK kinase inhibitor far fewer neurons are present, and they fail to extend axons and form networks (Fig. 3D, E) suggesting that LRRK2 kinase activity is necessary for speci cation and proper differentiation of neurons during embryonic development.To determine if NvLRRK2 kinase activity is also necessary for maintenance of neurons in adult animals we placed 3 month and 6-month adult transgenic animals in LRRK2 kinase inhibitors and imaged their neurons after 4 days.We observed in all cases that the axons appeared to degenerate, and no axonal processes could be observed in the inhibitor treated animals versus control animals (Fig. 2S), suggesting that the kinase activity of LRRK2 is necessary to maintain healthy connected axons The LRRK2 kinase activity is necessary for regeneration in Nematostella Nematostella are well-characterized to be capable of regeneration throughout life [28][29][30][80][81][82].The animal can be cut into multiple pieces and each fragment is capable of forming a whole new animal, the only exceptions to this is that a piece of foot alone or amputated tenacles are unable to facilitate full body regeneration.To assess whether LRRK2 has a role in regeneration in N. vectensis, the kinase activity of LRRK2 was inhibited after amputation below the pharynx.In adult animals' pharynx and tentacle regeneration is complete within 7 days as seen here in control animals (Fig. 4A).However, when exposed to the LRRK2 kinase inhibitor animals appeared to heal the wound but failed to regenerate signi cant tentacle, often both the pharynx and tentacles were missing.(Fig. 4).Additionally, there was a difference in body length for those treated with the kinase inhibitor, suggesting that like in the homeostatic conditions lack of a functional LRRK2 kinase leads to degeneration of the axons and overall shrinking of the body axis (Fig. 2S).Taken together these data suggest that nvLRRK2 plays an important role both in the development, homeostasis, and regeneration of neurons.

Discussion
In this study, we have introduced a novel organism for studying LRRK biology, and demonstrated that its LRRK2 orthologue, nvLRRK2, has great similarity to hsLRRK2.It has previously been noted that Nematostella are more similar to vertebrates on a genomic scale, than more well-used organisms such as the Drosophila are [75].This study expands on this and demonstrates its great potential for providing information regarding the function of LRRK2 in a disease context.
Here we have taken advantage of the availability of large number of Nematostella embryos and of commercially available validated LRRK2 kinase inhibitors to determine the function of nvLRRK2.We have uncovered a key role for the kinase activity of this gene in promoting normal embryonic growth, in speci cation of the oral region of the animal and in neuronal differentiation.Additionally, we have demonstrated that the kinase activity of LRRK2 is necessary to maintain a functional radial network of nerves in the Nematostella and to promote head regeneration where a complex set of nerves must be regenerated.This study indicates conservation of a gene which has been primarily studied in vertebrates and illustrates its high conservation in invertebrates.In previous studies, LRRK2 has been identi ed as making an important contribution to regeneration in a large-scale transcriptional pro ling approach during planaria regeneration, where the authors suggest it may play a role in activated of the neoblast populations that are essential for planaria regeneration [40].It will be important to determine in the future the identity of the effector pathway used by nvLRRK2 to regulate regeneration and development, providing mechanistic insight into this process.In mammals it is well-established that a subset of Rab proteins -including Rab10 -are LRRK2 substrates [83], undergoing phosphorylation by LRRK2 [84].We were able to con rm that Nematostella has a Rab10 ortholog, but in the absence of speci c tool antibodies to investigate this were unable to test for direct phosphorylation by nvLRRK2.Likewise, the conservation of key residues that are mutated in human disease (G2019S and I2020T) with de ned biochemical consequences provides an opportunity to assess the implications of gain of LRRK2 function on regeneration and development in the Nematostella system.
Our results open up a new avenue of investigation into the physiological role and importance of LRRK genes and provides a novel model platform to test the consequences of modulating LRRK2 functionwith the potential to provide insights relevant to human disease.
Primers used: 18SF: CGG CTT AAT TTG ACT CAA CAC G 18SR: TTA GCA TGC CAG AGT CTC GTT C Wnt4F: CGC CTA ACT ACT GCC ACA AA Wnt4R: CCT CGC CCA CAA CAA AGA TA SoxB1F: GTT GAC GGC TGA AGA GAA GG SoxB1R: AGA ATT TGT CAA CCG CCA TC LRRK2F: CCC ATA CCT CAC AGC TAC TTT AC LRRK2R: CGA ATT CCG CCC TGT GTA TAA In situ hybridization A 300bp fragment of the.NvLRRK2 coding region was cloned by PCR.The resulting PCR product was used to synthesize in situ probe by the addition of DIG-labeled UTP (Roche) plus the appropriate RNA Polymerase T7 or Sp6 (NEB).Probes were puri ed with RNA Clean Up Kit (Qiagen) and resuspended in 100uL of hybridization buffer.
Abbreviations days post fertilization