Generation of an Lpar1-EGFP Fusion Knock-in Transgenic Mouse Line

Lysophosphatidic acid (LPA) is a lysophospholipid that acts as an extracellular signal through the activation of cognate G protein-coupled receptors (GPCRs). There are six known LPA receptors (LPA1–6). The first such receptor, LPA1, was identified in the embryonic brain and has been studied extensively for gene expression throughout the body, including through studies of receptor-null mice. However, identifying receptor protein expression in situ and in vivo within living cells and tissues has been difficult because of biologically low receptor expression and variable antibody specificity. To visualize native LPA1 receptor expression in situ, we generated a knock-in mouse produced by homologous recombination in murine embryonic stem (ES) cells to replace a wildtype Lpar1 allele with a mutant allele created by in-frame fusion of EGFP to the 4th exon of Lpar1 (Lpar1-EGFP knock-in allele). Homozygous knock-in mice appeared normal and the expected mendelian ratios of knock-in allele transmission were present in females and males. Histological assessments of the fetal and adult central nervous system (CNS) demonstrated expression patterns that were consistent with prior in situ hybridization studies. This new mouse line will be useful for studies of LPA1 in the developing and adult CNS, as well as other tissues, and for receptor assessments in living tissues and disease models.


Introduction
Molecular cloning of the first lysophospholipid receptor was reported 25 years ago from studies of the embryonic cerebral cortex [1], which identified a receptor now known as LPA 1 (gene name Lpar1 for mouse, LPAR1 for human [2]) that mediates the effects of lysophosphatidic acid (LPA) [3][4][5]. In the ensuing years, other G protein-coupled receptors (GPCRs) for LPA and related lysophospholipids, including sphingosine 1-phosphate (S1P) were identified, revealing a rich biology with true medical relevance [5][6][7].
Much of what is known about the physiology and pathophysiology of LPA 1 came from experimental studies of the mouse, through the targeted genetic deletion of Lpar1 [27,55], as well as by using derivatives of the original mutant mouse to improve viability in complex strain backgrounds (maLpar1 [61]). A particular area of ambiguity has been ascertaining the biologically relevant locations of the LPA 1 protein within the intact brain towards understanding cellular mechanisms. While gene expression studies by in situ hybridization (ISH) identified Lpar1 expression in NPCs [1] during embryonic brain development, and oligodendrocytes [29] in the adult brain, determining the location of receptor protein has been more challenging because of inconsistent antibody specificity and availability, low receptor expression, differences in tissue preparation, and variables related to the effects of mouse background strain that produced inconsistent, if not contradictory results. A possible solution to these challenges in mice is through gene knock-in strategies [62], wherein a wildtype allele is replaced by a modified version that enables tracking of a functional receptor tagged with an EGFP fusion protein. Here we report an Lpar1-EGFP fusion knock-in transgenic mouse line and its initial characterization within the developing and adult brain.

Mice
Animal husbandry for the Lpar1-EGFP mouse line was provided by the animal resource departments of The Scripps Research Institute (TSRI) and the Sanford Burnham Prebys Medical Discovery Institute (SBP). Embryonic and adult mice were euthanized via isoflurane overdose per approved protocols and institutional guidelines. All animal procedures were approved and conducted in accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines of TSRI and SBP.

ES Cell Transfection
Standard methodologies were used to establish Lpar1-EGFP knock-in ES cell lines [21,27,51,55,[63][64][65]. Briefly, the targeting vector (50 mg) was linearized and mixed with 1 × 10 7 R1 ES cells (generously provided by Dr. Andras Nagy) in a 0.4 cm electroporation cuvette and electroporated using a Bio-Rad Gene Pulser II (200 mVolts × 800 mF capacitance). The electroporated cells were allowed to recover on ice for 20 min and then plated on a feeder layer of neomycin-resistant mouse embryo fibroblasts. Twenty-four hours after plating, the ES cell media was replaced with media containing 150 mg/ml of geneticin (ThermoFisher Scientific), which was replaced daily for 7 days. After 7 days, 140 individual ES cell clones were picked and split into 96 well master plates that were frozen in fetal calf serum containing 10% DMSO and 24 well plates for DNA isolation. Appropriately recombined ES clones were isolated and injected into recipient blastocysts by the TSRI Murine Genetics Core. Southern blotting and PCR genotyping with targeted primers identified germline transmission of the knock-in transgene ( Fig. 2B: A1 EGFP KI Forward: 5′-GACAAAGAGATGAGCGC CAC-3′, A1 EGFP Wt Reverse: 5′-GAGTGTCCTCATC TCCTCTG-3′, and EGFP Internal Reverse: 5′-GTGG TGCAGATGAACTTCAGG-3′).

Southern Blotting and DNA Hybridization
ES cells with homologous recombination of the Lpar1 genomic locus were identified by Southern blot hybridization [51,55,66,67] of HindIII digested genomic DNA that was separated on a 0.08% 1 × TAE agarose gel and transferred to Hybond TM -N membranes (Amersham), UVcrosslinked, and probed with 25 ng of a 32 P-labeled 1.2 kb DNA fragment from a region of the Lpar1 locus external to 3′ end of the Lpar1-EGFP targeting construct. Blots were pre-hybridized for one hour at 42°C in hybridization solution (0.5 M phosphate buffer pH 7.4, 50% formamide, 5 × SSPE, 1 × Denhardt's solution, 1% SDS, and 0.1% denatured salmon sperm) followed by the addition of the denatured radiolabeled DNA probe prepared using an Agilent Random Primer Labeling Kit. Following overnight hybridization (42°C), the blots were washed with 2 × SSPE 0.1% SDS and 0.2 × SSPE 0.1% SDS, then visualized with a GE Typhoon Phosphorimager. ES cell clones that did not have proper recombination of the Lpar1 locus were identified by a 6.7 kb band, and ES cell clones with homologous recombination events at Lpar1 were identified by the presence of a 6.7 kb wildtype band and 4.4 kb recombined band ( Fig. 2A).

Tissue Preparation and Sectioning
Mutant mice were identified and validated by standard genotyping procedures. Forward 5′-ACATGGT CCTGCT GGAGTTC-3′ plus Lpar1 Reverse 5′-GAGTGTCCTC ATCTCCTCTG-3′) of the neomycin. Cryostat sections were prepared from homozygote, heterozygote, and wildtype. Expression of the Lpar1-EGFP fusion allele was assessed by isolating embryonic day 13.5 (E13.5) heads and adult brain and lumbar spinal cord from wildtype and Lpar1-EGFP knock-in age-matched animals, which were processed for standard cryostat sectioning, fixation, and fluorescence microscopy [68][69][70][71][72][73]. In brief, samples were mounted in prechilled molds containing Neg-50 TM (Thermo Scientific) frozen section media and then frozen on dry ice. Tissue sections were cut at 20 μm using a Leica cryostat, fixed with cold 4% paraformaldehyde in phosphate-buffered saline (PBS) for 5 min, washed twice with PBS, mounted, and coverslipped using VECTASHIELD ® HardSet TM antifade mounting medium with DAPI (Vector Laboratories, Burlingame, CA). Tissue section, staining, and mounting was performed at the SBP histology core facility. Endogenous EGFP fluorescence in all tissues was evaluated using a KEYENCE BZ-X810 all-in-one fluorescence microscope. RNAscope [74] against mouse Lpar1 utilized commercially available probes and protocols (Advanced Cell Diagnostics, Newark, CA).

Construction of an Lpar1-EGFP Fusion Knock-in Construct
Mouse Lpar1 genomic DNA fragments were amplified from a bacterial artificial chromosome (BAC) template containing the Lpar1 genomic locus (BAC RP23-149020, Children's Hospital Oakland Research Institute) using Pfx50 TM DNA polymerase (Invitrogen). An amplified fragment containing Lpar1 exon 4 including a portion of the 3′ untranslated region (UTR) was used to fuse EGFP inframe to the carboxy terminus of Lpar1 using overlap PCR. An introduced BamHI restriction enzyme site positioned between the EGFP stop codon and the start of the Lpar1 3′ UTR was used to insert a loxP flanked neomycin cassette under the control of the phosphoglycerate kinase promoter (PGK-neo). A HindIII restriction enzyme at the 5′ end of the PGK-neo cassette was inserted to allow for the identification of homologously recombined ES cell clones. Amplified Lpar1 genomic fragments were assembled in the pGEM ® -T Easy Vector System (Promega) to produce an Lpar1 knockin targeting vector with 8.2 and 1.7 kb of homologous flanking sequence on the 5′ and 3′ ends of the EGFP and PGK-neo DNA sequences.

Isolation of Recombined ES Cell Clones and Lpar1-EGFP Knock-in Mice
ES cell clones positive for homologous recombination were identified by Southern blotting using the external probe ( Fig. 2A). Positive clones were expanded for confirmation of the recombination event, analyzed to ensure that only single integration events were present, and pathogen tested (IDEXX BioAnalytics). A single ES cell clone was selected for injection into C57BL/6 J mouse blastocysts to produce chimeric mice. Chimeric offspring were bred to C57BL/6 J animals to assay for germline transmission. Homologous recombination events were assessed by Southern blot hybridization of HindIII digested genomic DNA isolated from the offspring of chimeric mice using the external probe ( Fig. 1) and by PCR genotyping.
To delete the loxP flanked PGK-neo cassette, male mice heterozygous for the Lpar1-EGFP allele were crossed to female EIIa-cre transgenic mice that are routinely used for germline deletion of loxP-flanked DNA sequences [75][76][77][78]. Mice hemizygous for the Lpar1-EGFP allele and devoid of the neomycin cassette were identified by genomic Southern blotting ( Fig. 2A) and PCR (Fig. 2B, C). Mice with appropriate genotypes were used to propagate the Lpar1-EGFP line and were bred to homozygosity. Homozygous animals appeared phenotypically indistinguishable from littermate controls, showed normal Mendelian female-tomale ratios, and were fertile, allowing colony establishment with homozygous Lpar1-EGFP genotype-confirmed animals. Timed-breedings to assess embryonic day 13.5 (E13.5) brains, an age at which the ventricular zone of the embryonic cerebral cortex is prominent [1,79,80], were complemented by live births to produce adults for analyses.

Histological Analyses Reveal a Concordance between Lpar1-EGFP and RNAscope
Wildtype controls were compared to Lpar1-EGFP animals, wherein cryostat sections were mounted, fixed, counterstained with 4,6-diamidino 2-phenylindole (DAPI), and observed using fluorescence microscopy (Fig. 3A, B). Specific EGFP endogenous fluorescence was observed only within mutants. In embryonic specimens, EGFP was detected in the cerebral cortical ventricular zone (VZ), leptomeninges (LH), and in non-brain compartments. Compared to Lpar1 RNAscope in situ hybridization (Fig.  3C), the patterns of EGFP labeling extensively overlapped, including a relative absence of signal in the ganglionic eminence (ge), as previously observed [1,16]. These results support the presence of higher receptor density on the cell membranes of soma expressing LPA 1 . Within the adult brain, strong EGFP labeling was observed within most, if not all, myelinated fiber tracts and prominently within the corpus callosum (CC), internal capsule (ic) fimbria (fi), and medial and lateral habenulae (MH, LH) (Fig. 3D-F). Notably, robust EGFP was not observed in the hippocampal formation (H) that included the dentate gyrus (dg). Further analyses supported EGFP localization surrounding cell nuclei (Fig. 4A-C), which is consistent with the membrane localization of LPA 1 . The spinal cord (lumbar region) was also examined (Fig. 4D, E), which revealed more diffuse EGFP signals within known myelinated fiber tracts.

Discussion
Knowledge of LPA 1 protein localization under basal conditions of normal development and adult life will enable more accurate mechanisms to be identified for biological, pathophysiological, and drug discovery efforts that target LPA 1 . The Lpar1-EGFP knock-in mice described here express from the endogenous promoter and enhancer elements, and show no evidence of LPA 1 functional loss or deficiency compared to phenotypes from constitutive [21,27,61] and conditional [55] Lpar1 knock-out studies. These results support the overall functionality of LPA 1 -EGFP fusion proteins, which is consistent with other functional LPA 1 fusion proteins, including their use to determine the LPA 1 crystal structure (using the modified apocytochrome c, "bRil") [81]. The presence of EGFP in non-brain compartments (Fig. 3B) indicates that this mutant will likely be useful for many, if not all, cell types that express LPA 1 . Formal examination of the other cell types and tissues known to express LPA 1 [12] remains an area for future work.
The high concordance between EGFP signals and ISH labeling supports proximity between Lpar1 mRNA and receptor protein, at least at the level of signal that can be identified by the employed techniques. This was particularly true in the embryonic brain where immature cells have relatively little cytoplasm compared to their nuclei [82], revealing EGFP signals surrounding nuclei (Fig. 4B, C). However, it is certain that minute levels of receptor protein expression, not reported by direct Lpar1-EGFP expression, are present, particularly in developing neurons and comparatively rare, but existent, neuronal expression of LPA 1 that has been detected by single-cell Drop-seq transcriptomic studies [39]. Functional support for active receptors, at distal cellular locations of processes and growth cones on developing NPCs and neurons, can be seen in the effects of exogenous LPA exposure that results in process and growth cone retraction [18,20,22,25,83,84], indicating that receptors are present in these highly polarized cells. Amplification techniques such as anti-EGFP antibodies and structured illumination microscopy [85] could better resolve the presence of low-abundance receptor signals marked by the expressed receptor fusion proteins. In the adult brain, oligodendrocytes were by far the most prominent cell type identified by early ISH studies [29].  [39]. The robust labeling of myelinated fiber tracts is consistent with prior ISH and single-cell transcriptome studies that identified receptor expression in oligodendrocytes, while the current images support high-density LPA 1 localization within the cell membrane layers of the myelin sheaths [26,29,86,87]. An area of technical and biological ambiguity that may be approached using the newly created Lpar1-EGFP knockin mice relates to the hippocampus and LPA 1 's roles in learning and memory. Multiple reports identified hippocampal phenotypes in mice null for Lpar1 [45,[88][89][90][91]. However, the cell types responsible for these phenotypes are ambiguous. One attractive cell population for future studies is the mouse adult hippocampal precursor cell [92][93][94] that was reported to show high levels of LPA 1 expression based primarily upon the EGFP BAC mouse line created by the NIH GENSAT [95,96] project. This mouse showed high EGFP expression within white matter tracts like the Lpa 1 -EGFP mice reported here, but also showed high EGFP expression within the dentate gyrus [94] (and other locales such as the cerebral cortex) that did not overlap with the Lpa 1 -EGFP expression pattern. This raises the question on which mouse model provides a more accurate description of LPA 1 location.
Multiple, independent ISH analyses have uniformly been devoid of robust Lpar1 signals in the dentate gyrus including the original GENSAT ISH screen, the Allen Brain Atlas, the current report (Fig. 3), and past reports [29], as well as Drop-seq data [39] that did report a dentate gyrus neuron subcluster (Neuron_dentate_C1qI2[#4](HC)), but with low expression levels (~2 Lpar1 transcripts per million (TPM)) compared with the top brain subcluster (oligoden-drocyte_Mbp[#9](TH);~1600 Lpar1 TPM). A putative hippocampal progenitor subcluster that is not limited to the dentate gyrus (Neurogenesis_Sox4 [#13](HC)) reported somewhat higher expression (~11 Lpar1 TPM). Primary cell culture studies also reported an absence of LPA 1 expression in hippocampal neurons [84]. Evidence for low LPA 1 expression in the dentate gyrus vs. high EGFP expression in the BAC mice-as well as other discordant EGFP patterns-could be explained in multiple ways. A partial list includes the precise ages of brain analyses, tissue preparation, different EGFP half-life, differing mouse background strains, and mouse husbandry conditions, as well as a BAC-transgene that has unknown copy number, orientation, intactness, and integration sites. These issues might affect BAC-transgene EGFP expression under basal and challenged (e.g., cell culture, disease models) conditions. As notable, BAC-mouse lines were purposefully selected for optimal EGFP expression during line curation, which therefore selected for discrepant patterns compared to standard approaches [96] and might have forgone lesser signals more akin to those seen in the current Lpar1-EGFP mouse that employs wildtype gene regulatory elements. Use of these new Lpar1-EGFP knock-in mice should enable future clarification of hippocampal phenotypes and related issues, while providing a new experimental tool to assess LPA 1 protein localization within intact tissues.
Acknowledgements We thank Dr. Andras Nagy for providing the ES cell line and Dr. Gwendolyn Kaeser for editorial assistance. Federal and non-federal support was provided by the Department of Defense W81XWH-17-1-0455 (J.C.), the National Institute of Neurological Disorders and Stroke of the National Institutes of Health under award number R01NS084398 (J.C.), and the Clause Scholarship Program in Neurodegeneration and Aging (P.S.P.). We dedicate this paper to the good health and happy retirement of Dr. Viswanathan Natarajan. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Compliance with Ethical Standards
Conflict of Interest The authors declare no competing interests.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
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 license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license 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 license, visit http://creativecommons.