β2-microglobulin amyloid deposition and the RAGE-related inflammation pathway in ligamentum flavum thickening among patients undergoing hemodialysis: a comparative cross-sectional study from Japan
Article information
Abstract
Study Design
A comparative cross-sectional study using histological, biochemical, and molecular analyses of ligamentum flavum (LF).
Purpose
To investigate whether LF hypertrophy in hemodialysis (HD) patients is associated with β2-microglobulin (B2M) amyloid and advanced glycation end-product (AGE) deposition, along with the receptor for AGE (RAGE)–related inflammatory activation.
Overview of Literature
Lumbar spinal canal stenosis (LSCS) is common in HD patients. Although B2M amyloid is a hallmark of dialysis-related amyloidosis, the role of the AGE-RAGE axis in LF pathology remains unclear.
Methods
LF tissues from 33 patients with LSCS (HD, n=16; non-HD, n=17) were analyzed. Amyloid deposition was evaluated using direct fast scarlet (DFS) staining and B2M immunohistochemistry. Pentosidine and carboxymethyl-lysine (CML) levels were quantified by high-performance liquid chromatography in subsets (n=6 per group). Expression of RAGE-related inflammatory genes was measured using quantitative reverse transcription–polymerase chain reaction. Comparisons were made between the HD and non-HD groups, paired comparisons of DFS-stained areas on the dorsal and ventral sides of LF were performed within the HD group.
Results
B2M amyloid deposition was observed exclusively in LF of HD group and predoninated dorsally (29.7%±8.6% vs. 11.9%±6.9%, p=0.001). Pentosidine levels were significantly higher in HD group than in non-HD group (28.5±10.6 μg/g vs. 16.1±4.2 μg/g, p=0.009), whereas CML levels didn’t differ. On the dorsal side of LF, all examined RAGE-related inflammatory genes except toll-like receptor 4 was significantly upregulated in HD group; ventral expression showed no group differences. Immunohistochemical analysis demonstrated localized expression of RAGE, high-mobility group box 1, and nuclear factor-kappa B surrounding amyloid deposits.
Conclusions
LF in HD patients exhibits B2M amyloid and AGE accumulation with upregulation of RAGE-related inflammatory genes and proteins, especially on the dorsal side. LF hypertrophy in dialysis-related LSCS appears to represent an inflammatory–amyloid phenotype, indicating the AGE–B2M–RAGE pathway as a potential therapeutic target.
Introduction
Lumbar spinal canal stenosis (LSCS) is one of the most common spinal disorders in the elderly. Thickening of the ligamentum flavum (LF) is considered a major contributor to the development of LSCS through narrowing of the spinal canal [1,2]. Previous studies have demonstrated that degenerative changes, including loss of elastic fibers and fibrosis, occur in thickened LF, leading to tissue hypertrophy [3–5]. Mechanical loading on the ligaments during lumbar motion is thought to induce these changes, particularly on the dorsal side of the thickened LF [2,3,6]. In addition, angiogenesis and inflammation have been reported to play important roles in the pathophysiology of LF thickening, promoting fibrosis and tissue hypertrophy [7].
LSCS is also a common and significant problem in patients receiving hemodialysis (HD) [8]. Dialysis-related amyloidosis (DRA) develops in patients undergoing long-term HD and is characterized by extracellular deposition of β2-microglobulin (B2M) amyloid [9]. The most frequent DRA-related lesions include thickening of tendons and ligaments and bone erosions [10]. Previous studies have shown that B2M amyloid is deposited on or within the LF [10,11], suggesting an important role in the pathogenesis of LF thickening in HD patients; however, the underlying mechanisms in this population remain unclear.
Advanced glycation end products (AGEs), which are generated through non-enzymatic glycation and oxidation in the Maillard reaction, accumulate in various tissues with aging [12]. AGEs such as carboxymethyl lysine (CML) and pentosidine are uremic toxins that are elevated in patients with chronic kidney disease (CKD), particularly those undergoing HD, due to increased production and reduced clearance [13]. AGEs form cross-links between proteins in tissues and interact with specific AGE receptors, leading to systemic inflammation and progressive tissue damage [12,14]. In HD patients, AGE-modified B2M constitutes a major component of amyloid deposits and is involved in the pathogenesis of DRA [13].
AGEs bind to the receptors for AGEs (RAGE) [14]. High-mobility group box 1 (HMGB1) binds to RAGE in conjugation with Toll-like receptor 2 (TLR2) and 4 (TLR4) [14]. Binding of AGEs or HMGB1 to RAGE activates the nuclear factor-kappa B (NF-κB) signaling cascade [14,15]. Upon translocation of NF-κB to the nucleus, transcription of target genes is induced, resulting in the release of inflammatory mediators such as interleukin-6 (IL-6), interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), intercellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1) [14]. Genes encoding RAGE, HMGB-1, TLR2, TLR4, NF-κB, and these inflammatory cytokines are classified as components of the RAGE-dependent NF-κB inflammation pathway [16]. Because B2M amyloid is deposited on or within the LF [10,11], AGE-modified B2M amyloid may also accumulate in the LF of patients with CKD. Therefore, activation of the RAGE-related inflammatory cascade may present a pathological mechanism of DRA that contributes to LF thickening in HD patients. However, the involvement of the AGE-RAGE pathway in LF thickening in this population has not yet been elucidated.
In this study, we focused on B2M amyloids, AGEs, and the RAGE-related inflammatory cascade to investigate the pathogenesis of LF thickening in patients undergoing HD using immunohistochemistry, high-performance liquid chromatography (HPLC), and quantitative reverse transcription–polymerase chain reaction (qRT-PCR).
Materials and Methods
Ethics statement
This study was conducted in accordance with the principles of the Declaration of Helsinki. This study protocol was reviewed and approved by the Institutional Review Board of Tohoku University (approval number: 2022-1-136). Written informed consents were obtained from all participants.
Study design and population
This was a single-center, retrospective, cross-sectional study. We included 33 patients with LSCS who underwent decompressive surgery with ligamentum flavum resection between October 2017 and May 2020 at JCHO Sendai Hospital. Sixteen patients receiving maintenance hemodialysis were assigned to the HD group, and 17 patients not receiving hemodialysis were assigned to the non-HD group. The mean age was 67 years (range, 56–77 years; seven men and nine women) in the HD group and 75 years (range, 62–86 years; nine men and eight women) in the non-HD group. Preoperative magnetic resonance imaging was performed to measure LF thickness as previously described [5]. Clinical background information, including the presence of diabetes mellitus (DM) and serum creatinine levels, was reviewed from medical records to evaluate potential systemic comorbidities that may influence AGE accumulation or inflammation.
Tissue preparation
LF samples were obtained en bloc during surgery. The LF was sagittally sectioned at the thickest point, fixed in 10% paraformaldehyde for 24 hours, and embedded in paraffin blocks. Thin sections (5 μm) were prepared and deparaffinized using ethanol and xylene. Direct fast scarlet 4BS (DFS), which is useful for amyloid staining, and immunohistochemical staining were subsequently performed on all samples.
Amyloid detection: DFS staining
DFS staining visualized amyloid as orange deposits [17]. Sections from all 33 samples (HD group: n=16, non-HD group: n=17) were stained with DFS. In the HD group, regions of interest with an area of 0.25 mm2 were randomly selected from 10 locations per sample: five on the ventral side and five on the dorsal side of the LF, as previously described [6]. The orange-stained areas were quantified using ImageJ 1.53 software (National Institutes of Health, Bethesda, MD, USA), and the percentage area of amyloid deposition was calculated.
Immunohistochemistry
Endogenous peroxidase activity was blocked in all 33 sections by incubation with 30% hydrogen peroxide for 10 minutes. For RAGE, HMGB-1, and NF-kB staining, antigen retrieval was performed by autoclaving the slides at 121°C for 5 minutes. Nonspecific binding was blocked by incubation with goat serum for 30 minutes. Sections were incubated with primary antibodies against B2M (A0072, 1:2,500; DAKO, Carpinteria, CA, USA), RAGE (ab216329, 1:2,000; Abcam, Cambridge, UK), HMGB-1 (ab79823, 1:400; Abcam), and NF-κB (ab32536, 1:5,000; Abcam). After overnight incubation, sections for B2M staining were incubated with biotinylated secondary antibodies for 30 minutes, followed by streptavidin-peroxidase labeling for 30 minutes. Sections for RAGE, HMGB-1, and NF-kB were incubated with biotinylated secondary antibodies using EnVision FLEX-HRP (DAKO) for 30 minutes. Color development was performed using 3,3′-diaminobenzidine tetrachloride, and nuclei were counterstained with Mayer’s hematoxylin.
AGEs quantification: high-performance liquid chromatography
HPLC was performed to quantify CML and pentosidine according to the manufacturer’s protocols. Six samples from each of the HD and non-HD groups were dried using a vacuum dryer and placed in test tubes in which air was replaced with nitrogen. Samples were hydrolyzed with hydrochloric acid at 120°C for 16 hours and subsequently re-dried using vacuum. Samples and solvents were loaded onto octadecyl group-modified and benzenesulfonic acid group-modified columns and purified for CML or pentosidine analysis. All samples were fixed in microplates and incubated with anti-CML or anti-pentosidine monoclonal antibodies, followed by enzyme-labeled secondary antibodies. After washing, enzymatic activity was measured.
RNA extraction and purification
LF tissues from all patients were divided into ventral and dorsal portions, cut into small pieces, immediately placed in 3 mL of QIAzol (Qiagen, Hilden, Germany), and frozen. Samples were homogenized using a Polytron homogenizer (Kinematica AD, Luzern, Switzerland). Total RNA was extracted from the homogenate using the RNeasy Fibrous Tissue Mini Kit (Qiagen).
Gene expression analysis: quantitative reverse transcription–polymerase chain reaction
Complementary DNA was synthesized using the Cloned Avian Myeloblastosis Virus First Strand cDNA Synthesis Kit with a LightCycler system (Roche Diagnostics, Basel, Switzerland). PCR efficiencies and relative expression levels of RAGE-related factors were calculated relative to elongation factor 1α1 (eEF1α1) expression, as described previously [18]. Primer sequences are listed in Table 1.
Polymerase chain reaction primer sequences
Statistical analysis
Comparisons between the HD and non-HD groups were performed using the Mann-Whitney U test for HPLC and qRT-PCR data, the unpaired t-test for height, body weight, body mass index (BMI), age, and LF thickness, and the chi-square test for sex distribution. Differences between the ventral and dorsal sides of the LF within the HD group were analyzed using the Wilcoxon matched-pairs signed-rank test for the percentage of DFS-stained area. Data are presented as mean±standard deviation. All tests were two-sided, and a p-value <0.05 was considered statistically significant. Statistical analyses were performed using Prism ver. 10.0 for Mac (GraphPad Software, Boston, MA, USA).
Results
There were no significant differences between the two groups in terms of sex, height, BMI, or LF thickness. Body weight (non-HD group: 60.85±10.34 kg; HD-group: 54.43±7.97 kg; p=0.049) and age (non-HD group: 75±7 years; HD-group: 67±5 years; p<0.001) were significantly lower in the HD group (Table 2). DM was present in both groups (non-HD group: six patients; HD group: four patients). In the non-HD group, two patients had mildly elevated serum creatinine levels (1.1 and 1.9 mg/dL), whereas all other patients had levels ≤1.0 mg/dL.
Patients’ parameters of two groups
β2-microglobulin amyloid deposition
Immunohistochemical analysis revealed positive DFS and B2M staining in all samples from the HD group, whereas no positive staining was observed in the non-HD group (Fig. 1). In the HD group, the DFS-stained area was significantly larger on the dorsal side of the LF (29.70%±8.61%) than on the ventral side (11.88%±6.87%, p=0.001) (Fig. 2).
β2-microglobulin amyloid deposition. Direct fast scarlet (DFS) staining of the ligamentum flavum in the non-hemodialysis (non-HD) (A) and hemodialysis (HD) groups (B). β2-microglobulin (B2M) immunostaining of the ligamentum flavum in the non-HD (C) and HD group (D). Scale bars=500 μm. DFS staining shows amyloid deposition. The HD group has a large area stained with DFS and B2M in the ligamentum flavum, while the non-HD group shows no DFS and B2M staining.
Amyloid deposition in the ventral and dorsal sides of ligamentum flavum in the hemodialysis (HD) group. Direct fast scarlet (DFS) staining of ventral side of the ligamentum flavum in the HD group (A); dorsal side in the HD group (B). Ratio of DFS-stained area in the ligamentum flavum in the HD group (C). Scale bars=100 μm. The dorsal side of the ligamentum flavum has a significantly larger DFS staining area than the ventral side in the HD group. (C) Dorsal side: 29.70±8.61; ventral side: 11.88±6.87 (p=0.001).
AGEs: evaluation of CML and pentosidine
HPLC demonstrated no significant difference in CML levels between the two groups (non-HD group: 227.0±36.73 μg/g; HD-group: 249.4±56.18 μg/g, p=0.699). In contrast, pentosidine levels were significantly higher in the HD group (28.49±10.55 μg/g) compared with the non-HD group (16.13±4.22 μg/g, p=0.009) (Fig. 3).
High-performance liquid chromatography (HPLC) of carboxymethyl lysine (CML) and pentosidine in ligamentum flavum. HPLC quantification of CML (A) and pentosidine (B) in the ligamentum flavum. There is no significant difference in CML between the two groups. (A) Non-hemodialysis (non-HD) group: 227.0±36.73 μg/g; hemodialysis (HD)-group: 249.4±56.18 μg/g (p=0.699). In contrast, pentosidine levels were significantly higher in the ligamentum flavum of the HD group. (B) Non-HD group: 16.13±4.22 μg/g; HD-group: 28.49±10.55 μg/g (p=0.009).
RAGE-dependent pathway in the LF
On the dorsal side of the LF, expression levels of multiple RAGE-related genes were significantly higher in the HD group than in the non-HD group (Table 3). For example, RAGE expression was 2.53±1.22 in the HD group and 1.24±0.89 in the non-HD group (p=0.003). HMGB1, TLR2, NF-κB, ICAM-1, VCAM-1, IL-6, IL-1β, and TNF-α showed similar upregulation (all p<0.05). In contrast, TLR4 expression did not differ significantly between the groups (p=0.403). On the ventral side of the LF, no significant differences were observed in the expression of any of these genes (Table 4). Immunohistochemical analysis demonstrated positive staining for RAGE, HMGB1, and NF-κB surrounding amyloid deposits, particularly on the dorsal side of the LF in HD patients (Fig. 4).
Gene expressions in the RAGE-dependent pathway in the dorsal side of LF
Gene expressions in the RAGE-dependent pathway in the ventral side of LF
Immunohistochemistry of receptor for advanced glycation end products (RAGE)-related molecules. RAGE immunostaining of the non-hemodialysis (non-HD) (A) and hemodialysis (HD) group (B). High-mobility group box 1 (HMGB1) of the non-HD (C) and HD group (D). Nuclear factor-kappa B (NF-κB) in the ligamentum flavum of the non-HD (E) and HD group (F). Scale bars=100 μm. Arrowheads indicate immunohistochemistry-stained areas. Brown-stained areas indicate positive immunoreactivity. Strong immunostaining of RAGE, HMGB-1, and NF-κB were observed in the HD group (B, D, F).
Discussion
The present study evaluated gene and protein expression in thickened LF tissue from patients undergoing HD and compared the findings with those of controls. The results demonstrated that B2M amyloid was present exclusively in the thickened LF of patients undergoing HD, mainly on the dorsal side, and that AGEs, RAGE, NF-κB, and related inflammatory cytokines were increased in the thickened LF of these patients. Deposition of B2M amyloids, including AGEs, within or on the LF, together with activation of the RAGE-related inflammatory cascade, may represent a crucial pathological mechanism underlying LF thickening in patients undergoing HD.
Consistent with previous reports, B2M amyloid deposition was observed in the LF of patients undergoing HD [11,19] but not in the LF of controls. Moreover, this study is the first to demonstrate that B2M amyloid deposition predominantly occurs on the dorsal side rather than the ventral side of the LF. Degeneration changes of the LF are known to occur predominantly on the dorsal side [2,3], where lumbar motion imposes greater mechanical stress than on the ventral side [2]. Such stress may induce microtissue injury followed by inflammation. Although injured tissue is expected to undergo repair, it may instead degenerate [7], resulting in scar formation and tissue thickening [3]. In addition, B2M amyloid tends to deposit at sites of inflammation, where it may further exacerbate inflammatory responses [10]. Therefore, the dorsal side of the LF, which is subjected to greater mechanical stress, may enter a vicious cycle of degeneration, inflammation, and B2M amyloid deposition, leading to dorsal-predominant thickening. DFS stain-positive specimens were also observed in the LF of control patients. Because transthyretin amyloid deposition has been reported in the LF of non-dialysis patients [20], the DFS-positive specimens in the control group may reflect non-B2M amyloid, such as transthyretin.
AGEs are known to accumulate in tissues and contribute to degeneration of ligaments and tendons [21], especially in patients with renal failure or those undergoing HD [22]. AGEs bind to proteins such as B2M, and AGE-modified B2M constitutes a major component of B2M amyloid [13]. Although few studies have focused on the spine, AGE deposition in epidural tissue and the LF of patients undergoing HD has been reported [8,23]. Nokura et al. [8] showed that AGEs exhibited a distribution pattern similar to that of B2M in cervical extradural thickened tissue from patients on dialysis, and Inatomi et al. [23] reported that partial colocalization of B2M and AGE immunostaining in amyloid deposition areas of the cervical LF in dialysis patients. However, no previous studies have compared AGE deposition in the LF between patients undergoing and not undergoing dialysis. The present study is the first to show increased pentosidine levels, an AGE, in the thickened LF of patients undergoing HD. Furthermore, because B2M amyloid deposition was observed exclusively in the LF of dialysis patients, these findings suggest an association between B2M amyloid and AGEs in LF thickening in this population.
The AGE-RAGE pathway is a key contributor to the pathogenesis of DRA [13]. AGEs also induce HMGB1, and the AGE-RAGE pathway is considered to include both the direct AGE-RAGE interaction and the HMGB1-RAGE pathway [24]. Interaction between AGE-modified B2M and RAGE is thought to initiate inflammatory responses and promote tissue damage, ultimately leading to bone and joint destruction in patients with CKD [13]. RAGE-mediated inflammation has also been reported in musculoskeletal disorders unrelated to CKD, including carpal tunnel syndrome, rotator cuff injury, and frozen shoulder [16,21]. In spinal tissues, several studies have showed involvement of the RAGE-related inflammatory cascade in nucleus pulposus degeneration [25,26]; however, its role in LF thickening has not been previously reported. This study demonstrated for the first time that expression levels of RAGE, HMGB-1, TLR2, NF-κB, and inflammatory cytokines such as IL-6, IL-1β, TNF-α, ICAM-1, and VCAM-1 were significantly higher in the thickened LF of patients undergoing HD compared with those not undergoing HD, indicating enhanced activation of the RAGE-related inflammatory cascade. Immunohistochemical analysis revealed prominent pericellular staining for RAGE, HMGB-1, and NF-κB. Because RAGE is expressed on the plasma membrane of vascular smooth muscle cells and inflammatory cells [27], and the LF contains relatively few cells, the observed staining likely reflects RAGE expression in infiltrating inflammatory cells. Overall, the present findings demonstrate B2M amyloid deposition, increased pentosidine levels, and accelerated RAGE-related inflammatory signaling in the thickened LF of patients undergoing HD, suggesting that AGE-B2M-RAGE-mediated inflammation is a plausible pathological mechanism of LF thickening in this population. AGEs are known to be reduced through dietary modification and exercise [28], and improvements in dialysis membranes have shown to reduce B2M accumulation [29]. Therefore, reduction of AGEs and B2M may help prevent LF thickening in patients undergoing HD.
This study has several limitations. In addition to the relatively small sample size, significant differences in mean age and body weight were observed between the HD and non-HD groups. Patients in the HD group were significantly younger than those in the non-HD group, suggesting that amyloid deposition may accelerate LF thickening and contribute to earlier onset of LSCS in HD patients. The mean body weight was significantly lower in the HD group, likely reflecting undernutrition and protein-energy wasting, commonly associated with HD [30]. In addition, DM was present in both groups, and two patients in the non-HD group exhibited mildly elevated serum creatinine levels. These variations in metabolic background—including age, body weight, diabetes status, and renal function—may have influenced RAGE-related gene expression to some extent. Although gene expression on the ventral side of the LF did not differ significantly between groups, a trend toward higher expression in the HD group was observed, which may be attributable to the limited sample size. Finally, this study did not demonstrate a direct causal relationship between activation of the RAGE-related inflammatory cascade and LF thickening. Further studies are therefore required to clarify this issue.
Conclusions
In patients undergoing hemodialysis with lumbar spinal canal stenosis, β2-microglobulin amyloid deposition together with advanced glycation end products activates RAGE-related inflammatory cascades, predominantly on the dorsal side of the LF. These findings suggest that the AGE-B2M-RAGE pathway may drive LF thickening and represent a potential therapeutic target.
Key Points
Lumbar spinal canal stenosis frequently occurs in hemodialysis patients, and β2-microglobulin (B2M) amyloid deposition in the ligamentum flavum (LF) has been reported.
This study demonstrated that B2M amyloid deposition is more extensive on the dorsal side of the LF in hemodialysis patients, where mechanical stress and degeneration are more pronounced.
Pentosidine accumulation and marked upregulation of receptor for advanced glycation end products (RAGE)-related inflammatory genes and proteins were identified in the LF of hemodialysis patients, particularly on the dorsal side, linking advanced glycation end products with amyloid-driven inflammation.
These findings suggest that LF hypertrophy in dialysis-related stenosis represents an inflammatory–amyloid phenotype, highlighting AGERAGE signaling as a potential therapeutic target.
Notes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
The authors thank Dr. Daisuke Kurosawa and Dr. Eiichi Murakami of the Department of Orthopedic Surgery/Low Back Pain and Sacroiliac Joint Center, JCHO Sendai Hospital, Sendai, Japan, for their excellent assistance.
Funding
This study was supported by a Grant-in-Aid for Scientific Research (C; 21K09291) from the Japan Society for the Promotion of Science.
Author Contributions
Conceptualization: KI, YY. Methodology: KI, YY. Investigation / Data curation: KI, YY, TO. Formal analysis: KI, YY, YO. Visualization: KI, YO. Writing–original draft: KI, YY. Writing–review & editing: KI, YY, YO, TO, KT, KH, TA. Funding acquisition: YY. Project administration: YY, TA. Supervision: YY, TA. Final approval of the manuscript: all authors.