A step‐by‐step protocol for isolation of murine nucleus pulposus cells

Abstract The intervertebral disc (IVD) is composed of three separate tissues with distinct origins and properties. Elucidating changes occurring in these tissues in response to injury or age is paramount to identify new therapies to better manage disc and spine degenerative conditions, including low back pain. Despite their small size and different mechanical load pattern compared to higher species, the use of mouse models represents a cost‐effective and powerful approach to better understand the formation, maintenance, and degeneration of the IVD. However, the isolation of the different compartments of the IVD is complicated by their diminutive size. Here, we describe a simple, step‐by‐step protocol for the isolation of the nucleus pulposus (NP) tissues that can then be processed for further analyses. Analysis from mouse NP tissues shows sufficient quantities of RNAs, purity of the NP fraction, and overall RNA quality for gene expression studies, and reveals no increase in expression of disc degeneration markers, including TNFa, IL1b, and Mmp1 up to 15 months of age in C57BL6 wildtype mice.

Healthy, nondegenerate human samples are limited in their ease of procurement and quantity. Therefore, animal models have been utilized to better characterize healthy cell behavior and characteristics, disc formation and the degenerative process, despite the recognized morphologic differences between IVDs from small animals and humans. 11 Studies in larger animals are aided in cell isolation by the gross morphological differences between the gelatinous NP and the fibrocartilaginous AF/CEP-these methods utilize gross dissection and further enzymatic digestion to ensure a high yield of cells. 12 Mice present an excellent opportunity for studying disc degeneration with the wide availability of transgenic models and low overall costs, but cell isolation from each of the IVD compartments is challenging, as the size of the mouse IVD and number of cells within each IVD compartment are far less compared to larger organisms. There is no consistent method for isolating cells of the different IVD compartments in mice for either cell culture or RNA expression (Table 1). Recently, efforts have focused on cell isolation without the use of enzymes, with concerns that enzymatic digestion can alter RNA and/or cell surface expression profiles, require longer expansion times in culture and necessitate multiple passages that leads to general dedifferentiation. 27 Due to the typical small NP cell yields in mice, many preclinical studies have also resolved on harvesting cells from multiple discs and pooling their contents, usually pooling discs from the same spine region, as differential composition or gene expression between IVDs from lumbar and sacral regions have been reported. [28][29][30][31] Lastly, because the NPs function in a unique, hypoxic, and hyperosmotic environment, gene expression is likely to change in response to enzymatic or nonenzymatic extraction due both to the time required for the procedure and the relatively dissimilar environment of cell culture conditions versus tissue. For studies focused on phenotyping IVD tissues, there is a need to prepare RNAs of the different IVD compartments in a consistent, fast, and reproductible fashion, without an extra step of ex vivo culture. Main goals of cell isolation from the mouse IVD are thus threefold: (a) preserve enough cells to allow experimental studies, (b) maintain the gene expression profile of IVD cells and limit variations and artifacts due to retrieval or culture methods, and (c) do so in an accessible and easily reproducible method with dependable results.
In this report, we describe a new method to isolate pure NP RNAs from murine lumbar and thoracic IVDs, based on a simple centrifugation step.

| Mice
Mice were euthanized by cervical dislocation under general anesthesia induced by isoflurane inhalation. All procedures were approved by the institutional IACUC committee.
Detailed step-by-step protocol for the isolation of NPs and CEPs/AFs: 1. Place the animal on its back. Douse the skin with 70% ethanol to disinfect and prevent hair from spreading during dissection ( Figure 1A). 3. Isolate the spinal column from the surrounding tissues-using scissors, remove the musculature and long bones, ribs, and pelvis to cut in a parallel direction to the exposed spine from the sacrum and up to the thoracic region ( Figure 1E,F). Complete the isolation of the spine by making a transverse cut at the level of the femurs and at the appropriate level in the lumbar/thoracic/ cervical region, depending on the region of interest ( Figure 1G).  Figure 2B). This will expose the IVDs ( Figure 2C, black arrows).
Continue to remove the lateral attachments of soft tissues ( Figure 2C, white arrow), as doing so will simplify the process of isolating the IVDs from the vertebral column ( Figure 2D).

| Statistical analyses
Statistical tests are specified in the legend of each figure. When data were analyzed by two-way ANOVA, tissue compartment (NP vs AF/CEP) and age were the main effects (independent variables).
Levene's tests were used to confirm that the data met the equal variance assumption. Where the assumption was violated (Mmp13), the data were log transformed to correct the violation. Where the age effect was significant (and the tissue compartment by interaction was not), the age effect was further examined with trend analysis, testing for significant linear, quadratic, and cubic trends to determine if and how gene expression changed over time. Significance was defined as P < .05.  Figure 4A,B). Significantly more RNA was procured from lumbar NP (P = .007) and AF/CEPs (P = .006) when compared to thoracic counterparts ( Figure 4C). The RIN range was 4.7 to 5.8. There was no correlation between RIN number and age of the mouse.
T A B L E 2 qPCR primer set sequences  Figure 5A,B). In contrast, Bra and Cd24 expression was not different between NP and AF/CEP fractions, thus indicating significant contamination of NP RNA in AF/CEP samples, as these latter genes are not expressed in AF and CEPs ( Figure 5C, D). 34 This result was confirmed histologically on IVD cryosections showing remnants of NP cells following centrifugation ( Figure S1A). Compared to AF/CEP fractions, the NP fractions also showed very low (6-to 60-fold weaker) expression of marker genes for mature chondrocyte (Rankl, The NP has been implicated as an indicator and driver of IVD degeneration. 35 The ability to procure a pure fraction of NP tissue allows precise investigation of molecular phenotypes, response to treatments, and the degenerative process separate from the AF/CEP.
This prerequisite information can serve as a foundation for future celland stem cell-based therapies.
A caveat of the method described herein is the low RIN numbers observed in samples examined for RNA integrity. Although inadequate sample handling or storage may cause RNA degradation, the low RIN values obtained likely result from partial RNA degradation that may occur postmortem during the dissection process, before samples are F I G U R E 6 Expression of inflammatory cytokine and metalloproteinase genes. Gene expression (qPCR) for (A) Tnfa, (B) Mmp9, (C) Mmp13, and (D) Il1b (n = 5 per age, two-way ANOVA). Interaction was nonsignificant for all genes tested. Expression differed significantly between NP and AF/CEP for each gene measured (A: P = .03, B: P = .01, C: P < .001, D: P < .001). Age was a significant factor for both Mmp13 and Il1b with P < .001 and P = .04, respectively. Changing expression over time analyzed by orthogonal polynomial contrasts, with a significant linear decrease only for Mmp13 (P < .05), with a nonsignificant linear increase in Il1b (P = .06). Post-hoc test with Sidak's multiple comparison test was performed for analysis of NP and AF/CEP expression at each age (*P < .05, **P < .01, ***P < .001, ****P < .0001) snap-frozen or homogenized in guanidine-containing solutions, as aver- Mechanical disruption of the tissues while submerged in liquid nitrogen follows this principle, as disruption with a bead mill or a mechanical homogenizer exposes the samples to higher temperatures.
By comparing the expression of tissue specific transcription factors, we were able to show that the isolated NP is relatively free of AF/CEP and marrow cell contamination, but the above-described method is insufficient to result in an AF/CEP fraction free of NP cell contamination, as shown by the detection of NP-specific gene (Bra, Cd24) in AF/CEP fractions and the presence of NP remnants in centrifuged IVDs by histology. Therefore, the gelatinous property of the NP helps this tissue to be extracted by simple centrifugation from the IVDs, but peripheral NP cells remaining within the intervertebral discs during centrifugation contaminate AF/CEP samples.
There have been efforts made to differentiate changes due to the normal physiologic process of aging and the pathologic degeneration seen in degenerative disc disease. For instance, some studies have shown that IL1β and TNFα were key inflammatory cytokines in disc degeneration, released by IVD cells in addition to infiltrating inflammatory cells and able to induce catabolic and anti-anabolic shifts in bovine and human IVDs. [38][39][40] We report here no increase in expression of these key inflammatory genes, in the absence of diseasecausing conditions or degeneration-susceptible gene knockout, up to 15 months of age in the mouse NP. The only significant change in expression with age was a decrease in Mmp13 expression, which may reflect reduced ECM remodeling. To our knowledge, this is one of the first reports where these genes were examined in the context of aging and nondegenerative state in in vivo mouse IVDs and in different IVD compartments, establishing an important foundation for future studies focused on the correlation between aging and inflammatory gene expression. Our data show that although MRI, radiographic, and histologic evidence of IVD degeneration may occur at earlier ages, 41-44 no significant changes in gene expression for inflammatory markers occur up to 15 months of age in mice. These data are in line with a recent study comparing global changes in gene expression during aging in the mouse NPs and AFs by microarray analysis, which did not reveal any genes related to inflammation in the top most upregulated genes. 26

| CONCLUSION
Mice are important animal models for investigating disease processes and genetic influences in spine development, aging, and degeneration.
We present a reliable and reproducible method for isolating murine NP tissue from IVDs. We also show that there is no evidence of inflammation of the mouse NP up to 15 months of age at the gene expression level.