Clinical characteristics and APOE variant spectrum in Chinese patients with lipoprotein glomerulopathy

To the Editor: Lipoprotein glomerulopathy (LPG) is a rare renal disorder characterized by lipoprotein thrombi in the glomerular capillaries. Clinically, it presents with proteinuria and frequently progresses to progressive renal insufficiency.[1] Although identified in various populations worldwide, LPG is predominantly reported in Japan and China. Etiologically, the disease is primarily associated with variants in the apolipoprotein E (APOE) gene.[2] Variants in the APOE gene disrupt its normal function, accumulating pathological lipid deposits in glomeruli. Several APOE variants are linked to LPG, including the Tokyo-Maebashi (p.L162_K164del) and Kyoto (p.R43C).[2] These variants exhibit distinct geographical distributions and founder effects, suggesting that genetic drift and population-specific factors may influence the prevalence and clinical presentation.[3,4] With the advancement in molecular diagnostics, the number of identified patients is progressively increasing, necessitating a comprehensive re-evaluation of the spectrum and distribution of variants associated with LPG. Emerging evidence suggests that variants specific to certain gene regions exhibit phenotypic heterogeneity,[2] thereby warranting an analysis of genotype–phenotype correlation in risk stratification and health care optimization. This study evaluated the demographic, genotypic, and clinical characteristics of 27 LPG patients in Northern China. Specifically, we aimed to identify APOE variants, assess clinical features, and estimate progression risks to end-stage kidney disease (ESKD). Furthermore, we investigated genotype–phenotype correlations and potential founder effects, utilizing three-dimensional protein modeling to elucidate functional impacts and characterize a novel pathogenic variant. This retrospective study included 27 LPG patients from 24 families, diagnosed at Peking University First Hospital between January 2000 and August 2023. Diagnostic confirmation was undertaken using kidney biopsy. Whole-exome sequencing (WES) was used to perform genetic analysis. Ethical approval was obtained from the Ethics Committee of Peking University First Hospital (IRB No. 2018Y99), and all participants provided informed consent. For normally distributed variables, t-tests were used; while non-parametric tests, such as the Mann–Whitney U test, were employed for non-normally distributed variables. A linear mixed-effects model was used to construct estimated glomerular filtration rate (eGFR) trajectories to evaluate the progression of renal function over time. Pearson’s and Spearman’s methods were used to calculate the correlation coefficients for normally distributed and non-normally distributed data, respectively. The Kruskal–Wallis test was employed for analyses that involved multiple groups. A significance threshold of P <0.05 was adopted. Detailed methodologies are provided in the Supplementary Materials, https://links.lww.com/CM9/C764. Among the identified variants, the Tokyo-Maebashi genotype was predominant. The variants spectrum included the Tokyo-Maebashi (15 cases), followed by the Kyoto (8 cases) and Chengdu (2 cases). Notably, two patients exhibited typical clinical features of LPG but lacked the common pathogenic variants. Of these, one harbored a novel APOE variant (p.A210P), while the other remained without a molecular diagnosis. The mean age at diagnosis was 34 years. Geographically, the patients were predominantly distributed in North China, with a notable concentration in central and southern Hebei as well as northern Henan [Supplementary Figure 1A, https://links.lww.com/CM9/C764]. Family history data were available from 14 patients, with 50% (7/14) reporting a clear family history. The pedigree charts of six families [Supplementary File, https://links.lww.com/CM9/C764] revealed patterns consistent with autosomal dominant inheritance. Detailed clinical baseline data are summarized in Supplementary Table 1, https://links.lww.com/CM9/C764. The mean eGFR was 79 mL·min−1·1.73 m−2, with 63% (17/27) classified as chronic kidney disease (CKD) stages 1–2. The urinary protein excretion ranged from 0.5 g/24 h to 14.7 g/24 h and the mean value was 3.7 g/24 h. Thirty percent (8/27) of the patients exhibited nephrotic syndrome while 52% (14/27) presented with nephrotic-range proteinuria. Correlation analysis revealed that eGFR was negatively correlated with age and urinary protein, and positively correlated with kidney size [Supplementary Figure 1B, C, https://links.lww.com/CM9/C764]. Microscopic hematuria was observed in 67% (18/27) of the patients, with a median value of 8.0 red blood cells per high power field (RBC/HPF). Regarding treatment history, 52% (14/27) had received glucocorticoid therapy before diagnosis, while 22% (6/27) had been treated with a combination of glucocorticoids and immunosuppressants (e.g., cyclophosphamide and mycophenolate mofetil). One patient developed Pneumocystis pneumonia while undergoing combined glucocorticoid and cyclophosphamide therapy before enrollment. Rectifying the diagnosis significantly changed the treatment strategy for all these patients. Disease progression was monitored over an average duration of 6.5 years (11 patients with a follow-up time range of 2–16 years). During this period, two patients progressed to ESKD. All patients received fenofibrate lipid-lowering therapy and renin-angiotensin-aldosterone system inhibitors following their diagnosis. Although renal function remained stable during the initial years, a gradual decline in eGFR was evident with extended follow-up, highlighting an unmet need for novel or additional therapy. Two of the patients exhibited a significantly faster decline in eGFR, with mean rates of 18 mL·min−1·1.73 m−2 per year (patient 8 and 11). Patient 8 had concurrent hepatitis B virus (HBV) infection while patient 11 experienced fungus Pneumocystis jirovecii infection as aforementioned. Upon excluding these two patients, the estimated average annual decline in eGFR was 4 mL·min−1·1.73 m−2. This result suggests a possible lifetime risk of LPG progression to ESKD by the time the patients reach around 50 years of age [Supplementary Figure 1D, https://links.lww.com/CM9/C764]. Histologically, renal biopsies were performed on 24 patients. It consistently revealed key pathological features of glomerular capillary dilation and the presence of oil-red O-positive lipid thrombi in capillary lumens. Nineteen patients (79%) exhibited varying degrees of glomerulosclerosis, with an average sclerosis rate of 14% (range: 3–41%). The proliferation of segmental endothelial cells, mesangial cells, and the mesangial matrix was also observed. Other findings included mesangial interposition (33%, 8/24), mesangiolysis (38%, 9/24), and basement membrane thickening with double contours (46%, 11/26). Positive immunofluorescence was observed in 54% (13/24) of the patients, with segmental granular deposits of immunoglobulin M (IgM) (42%, 10/24) and C3 (13%, 3/24) in the capillary walls and mesangial areas. Three patients presented with distinct immunofluorescent patterns: IgM, C1q, and fibrinogen-fibrin-related antigen (FRA) deposits in one patient; and IgM and C3 deposits in two patients. According to electron microscopy results, podocyte foot process effacement was noted in all patients, with 52% (13/24) of them showing glomerular basement membrane thickening. Mesangial electron-dense deposits were observed in four patients. In this study, three known APOE variants were identified: the Kyoto (p.R43C), Tokyo-Maebashi (p.L162_K164del), and Chengdu (p.L173P) variants. Structurally, these variants are located within the N-terminal α-helix of the APOE protein.[2] The allele frequencies of the Kyoto variant in the Chinamap, Westlake, and gnomAD databases range between 8/106 and 1/104. This suggests that the prevalence of LPG may have been underestimated, considering that only less than 300 cases have been reported globally.[2] Conservation analysis indicates that the sites for the Tokyo and Chengdu variants are evolutionarily conserved across species. In contrast, the arginine at position 43 (Kyoto) appears evolutionarily recent, being present in higher primates whereas proline is found in other species [Figure 1A–C].Figure 1: Evolutionary conservation and population genetic analysis of APOE variants in lipoprotein glomerulopathy. (A) Phylogenetic tree of the APOE gene. The clade highlighted in blue includes humans and two closely related higher primates, which possess arginine at position 43 of the APOE protein, in contrast to the proline present in the remaining species. (B) Distribution of three variants in the APOE protein within our patients. (C) Amino acid sites involved in the Tokyo and Chengdu variants are conserved across different species. APOE: Apolipoprotein E; ARG: Arginine; LEU: Leucine; PRO: Proline.To investigate the ancestral origin of the Tokyo variant, eight tag single nucleotide polymorphisms were selected and a haplotype was constructed within a 0.55 Mb region surrounding the APOE gene [Supplementary Figure 2A, https://links.lww.com/CM9/C764]. The GCGATCTA haplotype was the most prevalent (36%, 4/11) among 11 unrelated patients who carried the Tokyo variant. This haplotype was present in only 1.2% of individuals from the Han Chinese population in the 1000 Genomes Project. It suggested a founder effect of Tokyo variant in northern Chinese. DMLE+ (University of Alberta, Edmonton, Alberta, Canada) was used to estimate that this variant most likely originated 203 generations ago, which is extrapolated to approximately 4060 years ago [Supplementary Figure 2B, https://links.lww.com/CM9/C764]. Genotype–phenotype correlation analysis [Supplementary Figure 3 and Supplementary Table 2, https://links.lww.com/CM9/C764] indicated that patients harboring the Tokyo variant exhibited a relatively milder clinical phenotype. To elucidate the underlying molecular mechanisms, we investigated the structural impact of the Tokyo and Kyoto variants using 3D protein modeling [Supplementary Figure 4A–F, https://links.lww.com/CM9/C764]. For the Kyoto variant, the amino acid change alters local surface hydrophilicity and electronic charge. AggreProt (Masaryk University, Brno, Czech Republic) supported that protein aggregation peaks at the variant site. This phenomenon is likely attributable to cysteine’s propensity to form disulfide bonds, as illustrated by the protein dimer model. Conversely, the Tokyo variant is a deletion variant that does not cause significant changes in the local hydrophobicity or surface charge. Given that this variant site is located within the low-density lipoprotein receptor (LDLR) binding domain, its interaction with different receptors was examined. Structural analysis confirmed that the LRKLRK sequence affected by the Tokyo variant might play a crucial role in receptor binding [Supplementary Figure 4G–J, https://links.lww.com/CM9/C764]. Additionally, the variant modifies the ionic associations in the APOE–LDLR complex from Aspartate–Arginine to Aspartate–Lysine. The guanidinium group of arginine possesses a concentrated positive charge, facilitating a stronger ionic bond than the amino group of lysine, which has a less focused charge distribution. Consequently, the electrostatic interaction between arginine and the negatively charged aspartate is more robust compared to the interaction involving lysine. In addition to these established genotypes, we identified a novel variant associated with LPG in a 34-year-old female, who carried a heterozygous variant (c.628G >C, p.A210P) in the APOE gene. According to American College of Medical Genetics and Genomics guidelines, this variant is classified as likely pathogenic (LP) based on the criteria PM1, PM2, PP2, PP3, and PP4. Finally, to explore potential biomarkers, we analyzed plasma and urinary APOE levels. We observed that these levels were significantly elevated in LPG patients compared to healthy controls (HCs) but showed no significant difference versus IgA nephropathy (IgAN) patients. However, after normalizing for urinary protein excretion, the median urinary APOE-to-protein ratio was significantly lower in LPG patients (0.06) compared to the IgAN (0.13) and HC (0.17) groups [Supplementary Table 3, https://links.lww.com/CM9/C764]. The limitations of this study include the relatively small sample size and single-center retrospective design, which are inherent to the rarity of LPG. Additionally, while structural modeling provided mechanistic insights, further in vitro and in vivo studies are warranted to fully validate the functional impact of these variants. In conclusion, despite the high phenotypic variability characteristic of APOE-associated LPG, the APOE Tokyo-Maebashi variant emerges as the predominant variant in Northern China, associated with relatively milder clinical manifestations. Continued investigation into pathogenesis, potential biomarkers, and therapeutic strategies is warranted. Acknowledgments The authors thank all the participants in the study. Funding This study was supported by grants from the National Science Foundation of China (Nos. 82370709 and 82570833); and National High Level Hospital Clinical Research Funding (Interdisciplinary Clinical Research Project of Peking University First Hospital, Nos. 2024IR35 and 2022CR41).