The role of circulating galectin-1 in type 2 diabetes and chronic kidney disease: evidence from cross-sectional, longitudinal and Mendelian randomisation analyses

Aims/hypothesis Galectin-1 modulates inflammation and angiogenesis, and cross-sectional studies indicate that galectin-1 may be a uniting factor between obesity, type 2 diabetes and kidney function. We examined whether circulating galectin-1 can predict incidence of chronic kidney disease (CKD) and type 2 diabetes in a middle-aged population, and if Mendelian randomisation (MR) can provide evidence for causal direction of effects. Methods Participants (n = 4022; 58.6% women) in the Malmö Diet and Cancer Study–Cardiovascular Cohort enrolled between 1991 and 1994 (mean age 57.6 years) were examined. eGFR was calculated at baseline and after a mean follow-up of 16.6 ± 1.5 years. Diabetes status was ascertained through registry linkage (mean follow-up of 18.4 ± 6.1 years). The associations of baseline galectin-1 with incident CKD and type 2 diabetes were assessed with Cox regression, adjusting for established risk factors. In addition, a genome-wide association study on galectin-1 was performed to identify genetic instruments for two-sample MR analyses utilising the genetic associations obtained from the Chronic Kidney Disease Genetics (CKDGen) Consortium (41,395 cases and 439,303 controls) and the DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) consortium (74,124 cases and 824,006 controls). One genome-wide significant locus in the galectin-1 gene region was identified (sentinel SNP rs7285699; p = 2.4 × 10−11). The association between galectin-1 and eGFR was also examined in individuals with newly diagnosed diabetes from the All New Diabetics In Scania (ANDIS) cohort. Results Galectin-1 was strongly associated with lower eGFR at baseline (p = 2.3 × 10−89) but not with incident CKD. However, galectin-1 was associated with increased risk of type 2 diabetes (per SD increase, HR 1.12; 95% CI 1.02, 1.24). Two-sample MR analyses could not ascertain a causal effect of galectin-1 on CKD (OR 0.92; 95% CI 0.82, 1.02) or type 2 diabetes (OR 1.05; 95% CI 0.98, 1.14) in a general population. However, in individuals with type 2 diabetes from ANDIS who belonged to the severe insulin-resistant diabetes subgroup and were at high risk of diabetic nephropathy, genetically elevated galectin-1 was significantly associated with higher eGFR (p = 5.7 × 10−3). Conclusions/interpretation Galectin-1 is strongly associated with lower kidney function in cross-sectional analyses, and two-sample MR analyses suggest a causal protective effect on kidney function among individuals with type 2 diabetes at high risk of diabetic nephropathy. Future studies are needed to explore the mechanisms by which galectin-1 affects kidney function and whether it could be a useful target among individuals with type 2 diabetes for renal improvement. Graphical abstract Supplementary Information The online version contains peer-reviewed but unedited supplementary material available at 10.1007/s00125-021-05594-1.

Other registries used to identify diabetes cases included the National Patient Register, the Swedish Cause of Death Register (ICD-10 codes E10-E14 and O244-O249), and the Prescribed Drug Register (ATC code A10). The different sources of case ascertainment were overlapping. For analyses of incident diabetes, all subjects with prevalent diabetes mellitus (regardless of type) were excluded. Subjects with incident diabetes type specified as other than type 2 diabetes (type 1 diabetes, latent autoimmune diabetes in adults (LADA), secondary diabetes or other) were censored at their date of diagnosis and not included as incident type 2 diabetes cases.

Secondary outcomes
The study participants were followed for incident coronary artery disease (CAD) through record linkage using their Swedish personal identification number with the previously validated Swedish Hospital Discharge Register, the Swedish Cause of Death Register, and the Swedish Coronary Angiography and Angioplasty Registry. CAD was defined as coronary revascularisation, fatal or nonfatal myocardial infarction, or death due to ischaemic heart disease. Information on all-cause and cause-specific mortality was retrieved from the Swedish National Cause of Death Registry.

Genotyping quality control
The quality control (QC) procedures in the MDCS-CC have been presented in previously published papers using this data [1]. Individual level QC was performed by removing individuals with a call rate of <0.95, an inbreeding coefficient of >3 SD away from the mean, discordance between inferred and reported sex, duplicate samples, a second-degree relatedness or higher within the sample based on identify by decent sharing calculations, or deviating from the common population structure in the MDCS-CC (exceeding 8 σ on first two principal components). Marker level QC was performed by filtering out variants if they had a call rate <95%, minor allele frequency of <0.01, variants on sex chromosomes, mitochondrial DNA, and variants showing an extreme deviation from Hardy-Weinberg equilibrium (p<1 x 10 -6 ). After QC and imputation using the 1000 Genomes (1000G) Phase 1 version 3 reference panel 21,575,257 variants were left for genome-wide analysis in 4086 participants with measured serum galectin-1 levels.

Construction of the multi-SNP instrument.
In order to improve the statistical power of our genetic instrument we considered inclusion of additional cis variants in the genome-wide significant LGALS1 locus. The sentinel SNP was defined as the SNP with the lowest p-value (rs7285699). We used GCTA-COJO to perform conditional and joint multiple-SNP analysis of the genome-wide significant locus (+/-300 kb from LGALS1). A stepwise selection of additional variants was performed based on a conditional p value threshold of 0.01. A less stringent p-value threshold was selected to not miss informative tag SNPs due to their mild LD with selected variants. In total three variants were retained and the joint effect estimates, standard errors and p values were calculated and used in a fixed effect inverse variance weighted analysis to assess the causal effect of galectin-1 on CKD and T2D (ESM Table 1).  Table 3 -Baseline characteristics of ANDIS participants with measured galectin-1.
Mono-S was previously the standard method for HbA1c analysis in Sweden, normal range 3.9 -5.3%