Conditional Deletion of LRP1 Leads to Progressive Loss of Recombined NG2-Expressing Oligodendrocyte Precursor Cells in a Novel Mouse Model

The low-density lipoprotein receptor-related protein 1 (LRP1) is a transmembrane receptor, mediating endocytosis and activating intracellular signaling cascades. LRP1 is highly expressed in the central nervous system (CNS), especially in oligodendrocyte precursor cells (OPCs). Previous studies have suggested LRP1 as a regulator in early oligodendrocyte development, repair of chemically induced white matter lesions, and cholesterol homeostasis. To circumvent embryonic lethality observed in the case of global LRP1 deletion, we generated a new inducible conditional knockout (KO) mouse model, which enabled an NG2-restricted LRP1 deficiency (NG2-CreERT2ct2/wtxR26eGFPflox/floxxLRP1flox/flox). When characterizing our triple transgenic mouse model, we noticed a substantial and progressive loss of recombined LRP1-deficient cells in the oligodendrocyte lineage. On the other hand, we found comparable distributions and fractions of oligodendroglia within the Corpus callosum of the KO and control animals, indicating a compensation of these deficits. An initial study on experimental autoimmune encephalomyelitis (EAE) was performed in triple transgenic and control mice and the cell biology of oligodendrocytes obtained from the animals was studied in an in vitro myelination assay. Differences could be observed in these assays, which, however, did not achieve statistical significance, presumably because the majority of recombined LRP1-deficient cells has been replaced by non-recombined cells. Thus, the analysis of the role of LRP1 in EAE will require the induction of acute recombination in the context of the disease process. As LRP1 is necessary for the survival of OPCs in vivo, we assume that it will play an important role in myelin repair.

Prior to plating and cultivating the OPCs on artificial fiber inserts composed of poly-L-lactic acid (Electrospinning Company, Didcot, UK) [47], the fibers were prepared and activated. After an initial incubation with 70% ethanol for 10-15min and three washes with sterile distilled water, 1h of PDL-coating (10µ g/ml, Sigma) at 37°C followed. Finally the inserts were washed with myelination medium (DMEM/Neurobasal medium (1:1, Gibco), 2% B27, 1% P/S, 1% SATO, 1% ITS supplement, 5µ g/ml N-acetyl-cystein, 10ng/ml D-Biotin (Sigma); SATO: DMEM (Gibco), 10mg/ml BSA fraction V, 6µ g/ml progesteron, 1.61mg/ml putrescine, 40µ g/ml T3, 40µ g/ml T4 (Sigma)). Now the OPCs were detached from the cell culture flasks. First, one wash with PBS was performed prior to addition of 5ml Trypsin/EDTA (Gibco). After incubation for 4min at 37°C an equal amount of ovomucoid was added to stop the dissociation and detachment of the cells. The cell suspension was then transferred into a 15ml reaction tube and centrifuged at 1,000xg for 5min. Supernatant was discarded, myelination medium was added and the sedimented cells were triturated with different tips (1000µ l, 200µ l) and cannulas (23G, 21G, B. Braun) three times each for improved dissociation. A number of 50,000 cells (in a volume of 230µ l myelination medium) was adjusted to fiber inserts (previously filled with 1.5ml myelination medium), placed in a 12-well-plate. The OPCs were allowed to differentiate and myelinate the coated fibers in the next 14 days at 7.5% CO2 with complete medium exchange every 2-3 days.

Immunocytochemical staining of myelinating oligodendrocytes on artificial fibers
The immunocytochemical staining was performed as previously described [48]. Culture medium was discarded first, followed by two washes with PBS. Next the oligodendrocytes were fixed with 4% PFA (Carl Roth) for 15min. After another two to three washes with PBS the cells were incubated with PBS + 0.1% Triton X-100 (AppliChem, Darmstadt, Germany) to permeabilize the cells for 10min at room temperature. Then the incubation with primary antibodies against MBP (1:250, rat, BioRad) and GFP (1:300, rabbit, Millipore) diluted in PBS was performed at 4°C over night. The next day, three washes with PBS for 5min each were done before the incubation with the secondary antibody against rat and rabbit (Jackson ImmunoResearch Laboratories Inc.) followed. Therefor the antibodies were diluted in PBS, adjusted to the cells and stored for 1h at room temperature impervious to light. After three final washes with PBS the inserts were carefully removed from the wells and fibers were covered with Fluoromount G (Southern Biotech, Birmingham, AL) and 13mm coverslips (Thermo Fisher Scientific). After 24-48h of hardening of the covering medium the cells were imaged.

Imaging
Immunocytochemical stainings of myelinating oligodendrocytes in the fiber assay were imaged via an LSM 510 Meta (Zeiss). Z-stacks with a distance of 0.37µ m were recorded until the complete myelin membrane of the object was scanned.

Quantification
Quantification of myelinated artificial fibers was done via ImageJ/FIJI. Completely myelinated sections were identified with the help of z-projections and the length of the internodes was measured via the segmented line tool and was documented in µ m. Additionally to the mean length, the number of internodes per oligodendrocyte was analyzed.

Statistics
For the analysis of myelination of artificial fibers GraphPad Prism was used. Experiments were performed with N=2 and n=4.

Artificial fiber assay and analysis of in vitro myelination of LRP1-deficient OPCs
To investigate myelination behavior of LRP1-deficient oligodendrocytes, we decided to perform an in vitro artificial fiber myelination assay. Therefore OPCs were isolated and cultivated via an immunopanning procedure. Afterwards the OPCs were plated on PDL-coated artificial fibers and were allowed to differentiate and to myelinate the fibers for 14 days in vitro (Figure S1 A+B). The myelination behavior of the cells was examined with focus on numbers of internodes per cell and lengths of internodes in mean or per single cell. C+D). In agreement with the interpretation of an increased fragility of LRP1-deficient OPCs only a minor population of recombined cells withstood the cultivation procedure. Consequently, less than 1% of the myelinating oligodendrocytes expressed GFP, which we did not consider sufficient for a rigorous statistical analysis. With this reserve in mind, we saw elongated myelin sheaths or rather internodes in the length plot, the log length plot and the mean internode length of the investigated recombined cells within the KO condition compared to the control ( Figure S1 E-G). However, the number of internodes was not impaired under these conditions (Figure S1 H). These findings suggested a modified myelination in LRP1-deficient cells, consistent with the assumption that LRP1 might be involved in oligodendrocytic functions in vitro.

Comparison of control and recombined tissue by immunohistochemistry with distinct markers
Recombined tissue displayed clearly a larger number of recombined GFP-expressing cells than the control ( Figure S2, GFP-labelling). The control experiment shows that secondary antibodies alone did not yield significant signals in the tissue sections ( Figure S2, secondary antibody labelling). Figure S2: Immunohistochemical stainings to detect LRP1 and GFP and to verify the staining specificity by secondary antibody controls.
Representative staining in P7 tissue with focus on Corpus callosum (within red lines). Cell nuclei were detected with Hoechst dye. LRP1 was labeled with Cy2-coupled donkey anti-rabbit antibody and GFP was identified by Cy3-coupled donkey anti-goat antibody. Channels are presented as single and merge images. Please note that cells that do not pertain to the NG2-lineage are not recombined and hence conserve LRP1-expression. Additionally the specificity of the staining was evaluated by immunohistochemical stainings with secondary antibodies only (merge images shown). Scale bar: 100µ m.

Cellular characterization of oligodendrocyte lineage-associated cells
We have scored the numbers of Olig2-or PDGFR-expressing cells with reference to the total number of cell nuclei detectable in the section, as revealed by staining with the Hoechst dye ( Figure 6). As an alternative approach, we have probed the number of labeled cells with reference to the surface area in mm 2 . Differences that emerged when using the number of nuclei was used as basis ( Figure 6) did not substantiate when the latter parameter was applied. However, this approach does not consider variations in cell density. Figure S3: Exemplary analysis of cell numbers per area to characterize oligodendrocyte-specific lineage markers alternatively.
A Number of Olig2-positive cells normalized to the analyzed area in mm 2 in the KO at P7, differening between the rostral and caudal Corpus callosum. B Number of Olig2-positive cells normalized to the analyzed area in mm 2 in the KO at P21, separating between the rostral and the caudal part of the Corpus callosum. C Number of PDGFRα-positive cells normalized to the analyzed area in mm 2 at P56. Rostral parts of the Corpus callosum were compared. Data are expressed as mean ± SEM. N=3-4, n=9-12 per rostral and caudal part. At least 200-1200 cells per area were counted. Depending on normally or not normally distributed data, Student's t-test or Mann-Whitney U-test were used for the evaluation within the individual ages.