The correlation between imaging expression of P16 and S100 in hypertrophic ligamentum flavum

The research program was approved by the Institutional Review Committee of Tianjin Union Medical Center, and all procedures are based on the Helsinki Declaration. When patients underwent posterior lumbar surgery, the full thickness of the LF was removed from L4/5 segments. (p > 0.05, Table 3). There was no significant difference in baseline data between groups (p > 0.05, Table 1). After conservative treatment for at least three months, no obvious symptoms improved in all patients. All patients had no ossification of LF, secondary adhesive arachnoiditis, polyneuritis, no history of lumbar surgery, history of intraspinal invasive treatment such as epidural, etc. Considering the influence of diabetes and hypertension on the hypertrophy of the LF, none of the selected patients had a history of diabetes, hypertension, and underlying diseases that could have a potential impact on LF. We divided LF tissue into three groups: simple lumbar disc herniation (LDH), single-segment spinal stenosis (SLSS), double-segment spinal stenosis (DLSS). The dorsal and dural LF were detected for each group. Western blot and qPCR were used to detect S100 and P16 and their expression products.

The measurement method of LF was done as in our previous study [8]. We consider the ligamentum flavum> 3 mm as hypertrophy of ligamentum flavum in imaging measurement. The thickness of the LF was measured from the mid-point of the LF to the ventral side of the inner rim. The lumbar spinal canal oblique diameter is measured from the midpoint of the dorsal side of the ligamentum flavum to the midpoint of the posterior margin of the vertebral body. Relative thickness of LF (RT)(%) was calculated as the percentage of the thickness of LF to the oblique diameter of lumbar spinal canal. The average of three independent measurements of three surgeons were taken to determine the relative thickness of a single sample.

The dorsal and ventral LF of LDH, SLSS and DLSS were ground in liquid nitrogen. Extraction of total protein using RIPA lysate containing PMSF (Solarbio, USA). The CBA protein quantitative kit (Biosciences, USA) was diluted with 5 × buffer at 1:5 and denatured at 95 °C for 10 min To prepare SDS-PAGE gel (biorbyt, UK), 20 μL of total protein was taken, and the protein band was separated by 120 V electrophoresis for 1 h, then 100 mA transferred it to PVDF membrane (Millipore, USA). After 5% skimmed milk powder was sealed for 3 h, the PVDF membrane and the first antibody were incubated overnight at 4 °C. After washing the membrane 3 times with tris-buffered saline with Tween 20 (TBST, Solarbio, USA), the secondary antibodies labeled by Horseradish peroxidase (HRP, Solarbio, USA) were incubated at 37 °C for 1 h. The ChemiDocMP chemiluminescence imaging system (Bio-rad, USA) was exposed to observe the bands on the PVDF film.

Total RNA was extracted using Trizol reagent (Takara, Japan). SuperScriptIIIRT kit (Gibco, USA) is used for reverse transcription. Sybr qPCR mix (TOYOBO, Japan) was used for quantitative PCR analysis. PCR amplification conditions: 95 °C, 2 min, 94 °C, 20 s, 60 °C, 20 s, 72 °C, 30 s, a total of 40 cycles. Quantitative analysis was performed on ABI7900PCR instrument. Using Actin as the control, the primer sequences of quantitative mRNA, are provided in Table 2. The gene expression level was calculated using the 2^-ΔΔCt method relative to the actin gene. 2^-ΔΔCt > 2 or < 1/2 is considered statistically significant.

The thickness of LF was measured and analyzed for a total of 120 times. Its absolute and relative thickness is shown in Table 3. Compared with SLSS and LDH group, the absolute and relative thickness of LF in DLSS group was larger (p <  0.05). The average thickness of DLSS group was 5.821 mm (rt = 44.501), while that of SLSS group and LDH group was 4.958 mm (rt = 37.420) and 2.811 mm (rt = 21.510), respectively.

Western-blot analysis was adopted to evaluate the differential expression of P16 and S100 proteins among all three groups. As displayed in Fig. 1c, the protein expressions of S100 in both the dorsal and dural side of the LF among all the three groups showed no significant differences, consistent with the mRNA expression. However, as shown in the Fig. 1d, the expression level of P16 protein in the LDH group was much lower than that in the SLSS and DLSS group (p < 0.01), on the dorsal side or the dural side. In both sides of the LF, the P16 protein expression in the DLSS group was higher than that in the SLSS group, regardless of there being no statistical significance between them. There was a significant correlation between absolute LF thickness, relative LF thickness and the protein expression of P16 (r = 0.671, p = 0.001 and r = 0.732, p = 0.000, respectively; Tables 4 and 5). No significant correlation was existed between absolute LF thickness, relative LF thickness and the protein expression of S100.

As illustrated in Fig. 2a, the mRNA expression of S100 obtained from the LF cells among the three different groups of LDH, SLSS and DLSS showed no significant differences, on the dorsal and dural side of the LF. However, the P16 mRNA expression in the group of SLSS was dramatically higher than that in the LDH (p < 0.01), and DLSS (p < 0.01) groups. Moreover, in the dorsal side of the LF, the expression level of P16 mRNA in the DLSS group was much higher than that in the SLSS group, and showed significant differences (p < 0.05, Fig. 2b) Conversely, in the dural side of the LF showed no significant differences between the SLSS and DLSS group. There was a significant correlation between absolute LF thickness, relative LF thickness and the protein expression of P16 (r = 0.633, p = 0.003 and r = 0.590, p = 0.006, respectively; Table 4 and 5). No significant correlation was noted between absolute LF thickness, relative LF thickness and the protein expression of S100.