This was a prospective, single-center study to detect whether monoclonal immunoglobulins were bound to RBCs. Forty-nine patients (age ≥ 18 years) who were newly diagnosed with monoclonal gammopathy or MGRS in the Nephrology Department of Shandong Provincial Hospital affiliated with Shandong First Medical University from April 2023 to December 2024 were included in this study. The clinical data, including age, sex, albumin (ALB), serum creatinine (SCR), hemoglobin, estimated glomerular filtration rate (eGFR), 24h urinary protein, as well as the results of renal biopsy and bone marrow biopsy, were collected. We used the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula to estimate the eGFR.

We enrolled 40 age- and gender-matched healthy individuals as normal controls, along with five patients with lupus nephritis (LN) and five patients with membranous nephropathy (MN) who were positive for anti-phospholipase A2 receptor (PLA2R) antibodies as disease controls.

Written informed consent was obtained from each patient. The research was conducted in compliance with the Declaration of Helsinki. The study was approved by the local ethics committees of Shandong Provincial Hospital, affiliated with Shandong First Medical University (SZRJJ: NO. 2021-145).

In the local clinical practice, patients admitted to the Nephrology Department with kidney diseases routinely received sTLCs assays, sFLCs assays, and capillary electrophoresis and immunosubtraction (CE/IS) for monoclonal gammopathy screening. sTLCs were analyzed by the N Antisera to Human Immunoglobulin/L-chains kit, using a nephelometric analyzer Dade Behring BN II (Siemens, Germany). Concentrations of κFLC and λFLC were assessed using the Siemens N-Latex FLC kappa (κFLC) and lambda (λFLC) assays (Siemens Healthineers, Germany) according to the manufacturer’s instructions. CE/IS was performed using CAPILLARYS 2 (Capillary System) according to the manufacturer’s instruction. The reference ranges for κFLC and λFLC are 6.7–22.4 and 8.3–27 mg/L, respectively, and the κFLC/λFLC is 0.31–1.56. The reference ranges for κTLC and λTLC are 1.7–3.7 g/L and 0.9–2.1 g/L, respectively, and the κTLC/λTLC is 1.35–2.65.

RBCs were isolated from EDTA-anticoagulated blood immediately after blood collection by centrifugation at 956 g for 10 min at 4°C. After centrifugation, the supernatant (plasma and buffy coat) was carefully aspirated to obtain packed RBCs, which were then washed with phosphate-buffered saline (PBS). Then pre-chilled 10 mmol/L hypotonic Tris-HCl buffer (pH 7.4) was addedto the washed RBCs at a volume ratio of 1:40 and then incubated at 4°C for 2h. This step was followed by centrifugation at 4°C and 15,294 g for 15 min to pellet the erythrocyte membranes. RIPA lysis buffer (Solarbio, Beijing, China) was then introduced to dissolve erythrocyte membranes, and the mixture was incubated on ice for 30 minutes. Thereafter, protein samples were extracted from the supernatant following centrifugation. The protein concentration was determined using a BCA protein analysis kit (Epizyme, Shanghai, China).

Erythrocyte membrane proteins were denatured in protein sample loading buffer (denaturing, reducing, 5×) and boiled for 10 min. Subsequently the proteins were electrophoresed with 10% SDS-PAGE and then were transferred to a polyvinyl fluoride membrane (PVDF) membrane (Millipore). Dilutions of the following antibodies were made and cocultured with the membrane overnight at 4°C: rabbit monoclonal anti-human lambda antibody (diluted 1:5000, ab124719, Abcam), rabbit monoclonal anti-human kappa antibody (diluted 1:1000, ab124727, Abcam), rabbit monoclonal anti-human IgG gamma chain antibody (diluted 1:5000, ab133470, Abcam), rabbit monoclonal anti-human IgM heavy chain antibody (diluted 1:1000, ab212201, Abcam), rabbit monoclonal anti-human IgA heavy chain antibody (diluted 1:4000, ab133660, Abcam), and rabbit monoclonal anti-human apolipoprotein E (ApoE) antibody (diluted 1:100, ab52607, Abcam). Then incubate the membrane with HRP-conjugated goat anti-rabbit IgG (diluted 1:10000, SA00001-2, Proteintech) for 1h at room temperature. The ECL system and Bio-Rad electrophoretic image analyzer were used to observe the immune reaction zone after the developer was added.

Statistical software SPSS version 22.0 (SPSS, Chicago, IL) was employed for statistical analysis. Quantitative data were expressed as mean ± SD, median and 25th–75th percentile, median and range, or number (%).

Based on the results of CE/IS, sTLC or sFLC ratios, and renal biopsy, 49 patients were enrolled in this study, including 48 with abnormal CE/IS results and/or sTLC and/or sFLC ratios, and 1 with MGRS lacking a circulating M protein. The general characteristics of the patients are shown in Table 1 . These patients were all newly diagnosed patients with monoclonal gammopathy who had not received systematic treatment. Among the 49 patients, 31 were diagnosed with multiple myeloma (MM), of whom 9 underwent renal biopsies. The renal pathological findings included three cases of light chain amyloidosis, two cases of cast nephropathy, three cases of light chain deposition disease (LCDD), and one case of cryoglobulinemia. Two of the 49 patients were diagnosed with Waldenström’s macroglobulinemia (WM) without renal biopsies. Twelve of the 49 patients were diagnosed with MGRS, which included seven cases of light chain amyloidosis, two cases of LCDD, and one case of light/heavy chain deposition disease where proliferative glomerulonephritis with monoclonal IgG deposits (PGNMID) could not be definitively excluded, as well as two confirmed cases of PGNMID. The remaining four patients were considered to have MGUS. Among them, one patient had abnormal CE/IS results (IgG-kappa) and sTLC ratios without any signs of organ involvement. One patient had abnormal CE/IS results (IgG-lambda) and sTLC ratios with minimal change disease (MCD) diagnosed by renal biopsy. One patient had abnormal CE/IS results (IgM-lambda) but normal ratios of sTLC and sFLC with focal segmental glomerular sclerosis (FSGS) diagnosed by renal biopsy. One patient had abnormal CE/IS results (IgG-lambda) along with abnormal sTLC and sFLC ratios. A renal biopsy was not performed due to her advanced age. However, she was positive for anti-PLA2R antibodies, and her proteinuria responded favorably to rituximab therapy, so a diagnosis of MGUS combined with PLA2R-associated MN was considered.

Previous studies have documented small amounts of immunoglobulins (IgG, IgA, or IgM) coated RBCs in healthy subjects as determined by some measurement methods, such as the agglutination-based direct antiglobulin test (DAT), radioimmunoassay (RIA), and enzyme-linked immunosorbent assay (ELISA) (16–18). In this study, we employed Western blot analysis to identify the immunoproteins, including IgG, IgM, IgA, kappa, and lambda, presented on the surface of RBCs in 40 healthy controls, 5 LN patients, 5 MN patients, and 49 patients with monoclonal gammopathy.

Among the 40 healthy controls, 4 exhibited very weak bands corresponding to lambda, kappa, IgM, or IgG on RBCs. In five patients with LN, all the tested immunoproteins were present on the RBCs’ surface. That was immunoglobulins binding to RBCs in patients with LN exhibiting polyclonal characteristics, rather than being of monoclonal origin. In five patients with MN who were positive for anti-PLA2R antibodies, no immunoglobulin was detected on RBCs. Among the 49 patients with monoclonal gammopathy enrolled in this study, substantial increases in the amounts of erythrocyte membrane-bound immunoglobulins, characterized by monoclonal properties, were observed in 45 (91.8%) patients. Not only intact monoclonal immunoglobulins but also solely monoclonal immunoglobulin fractions were detected on RBCs. Among the remaining four patients, one patient with light chain amyloidosis had all the tested immunoglobulins detected on RBCs. No immunoglobulin was detected on RBCs in one patient with LCDD, one patient with MGUS (renal pathological type was MCD), and one patient with MM. The representative and detailed results of erythrocyte electrophoresis were respectively illustrated in Figure 1 and Supplementary Figures 1 and 2 , respectively.

Previous studies have reported that RBCs carry some amyloid substances, such as amyloid β in Alzheimer’s disease and amylin in type 2 diabetes (13–15). Given that ApoE is a protein usually co-deposited in light chain amyloidosis (19, 20), we performed Western blot analysis to assess the presence of ApoE on RBCs in 10 patients with light chain amyloidosis. Our findings revealed an absence of ApoE on RBCs in all subjects examined (data not shown).

Subsequently, we explored whether the deposition of M proteins on RBCs was determined by the quantity of serum immunoglobulins. As shown in Figure 2 , in most patients with M proteins presented on RBCs, the serum levels of the corresponding immunoglobulins were found to be elevated. Nevertheless, it was observed that in a subset of patients with normal serum levels of immunoproteins, the corresponding immunoglobulins were detected on RBCs, and conversely, patients with undetectable RBC M proteins exhibited elevated levels of the corresponding immunoproteins in serum.

Among the 49 patients with monoclonal gammopathy enrolled in this study, 45 (91.8%) had M proteins detected on RBCs, 38 (77.5%) had abnormal CE/IS results, 36 (73.5%) had abnormal sFLC ratios, and 32 (65.3%) had abnormal sTLC ratios. When considering the combination of tests for identifying M proteins, the combination of serum tests, including CE/IS, sTLC, and sFLC ratios, could diagnose 98.0% of cases, while the combination of serum tests with erythrocyte electrophoresis could diagnose 100.0% of cases in this study. Furthermore, due to the relatively clean background, the results of erythrocyte electrophoresis appeared to be more recognizable than those of CE/IS. The findings from erythrocyte electrophoresis and serum CE/IS for all patients included in the study are presented in Supplementary Figure 2 , with representative examples illustrated in Figure 3 .

As presented in Supplementary Table 1 , among the 49 patients, the findings from erythrocyte Western blot, serum CE/IS, sFLC, renal biopsy, and bone marrow flow cytometry were concordant in 26 (53.1%) cases (patients 1–26). Eleven (22.4%) of the 49 patients (patients 27–37) demonstrated negative results in CE/IS testing, while erythrocyte Western blot analysis disclosed the presence of extrinsic monoclonal immunoprotein bands, which were consistent with the results of sFLC, bone marrow flow cytometry, or renal histopathology. Patient 32 merits particular attention. She presented with a normal sFLC ratio and negative CE/IS results. Immunofluorescence analysis of the renal biopsy specimen demonstrated deposits of IgG 3+, kappa 3+, lambda weakly positive (±), C3 2+, and C1q 3+, with IgG3 1+ and IgG1/2/4 negative. Light microscopic examination revealed histological features consistent with membranoproliferative and nodular glomerular alterations. Electron microscopic analysis uncovered fine granular electron-dense deposits on the endothelial aspect of the glomerular basement membrane. The final diagnosis considered MGRS (light/heavy chain deposition disease), with the possibility of PGNMID not being definitively excluded. Interestingly, erythrocyte Western blot analysis revealed the presence of extrinsic IgG bands in the absence of kappa or lambda chains, suggesting the deposition of IgG heavy chains on erythrocytes. Eight (16.3%) of the 49 patients (patients 38–45) presented abnormal yet discordant results in serum CE/IS and erythrocyte electrophoresis. Patient 39 merits particular attention. She had normal sFLC ratios; the serum CE/IS result indicated the presence of IgM-lambda, yet erythrocyte electrophoresis analysis revealed distinct and intense bands for IgM in the absence of either kappa or lambda chains. Immunofluorescence analysis of the renal biopsy specimen demonstrated deposits of IgG 1+, IgM 1+-2+, kappa 1+, and lambda 1+. On light microscopic examination, the histological features were indicative of FSGS. Unfortunately, no glomeruli were available for electron microscopy examination, precluding a definitive diagnosis of heavy chain deposition disease. Four (8.2%) of the 49 patients had no or polyclonal immunoproteins on RBCs.