Bipolar disorder is a serious mental disorder, characterised by recurrent episodes of depression and mania, with a lifetime risk around 1%. The pathoaetiology has not been clarified but heritability estimates of up to 90% signify the importance of genetic factors. Commonly occurring single-nucleotide polymorphisms (SNPs) in the CACNA1C gene have been associated with bipolar disorder in several genome-wide association studies.1 CACNA1C encodes the alpha 1C subunit of Lc-type voltage-gated calcium (Ca2+) channels, which lead to an influx of extracellular Ca2+ into a cell upon depolarisation. There are several common variants in the CACNA1C gene, but the strongest association with bipolar disorder has been seen for the SNP rs1006737 in the third intron.Reference Ferreira, O'Donovan, Meng, Jones, Ruderfer and Jones2 Genetic association results need to be explored by studying allelic variation in relation to biomarkers and clinical phenotypes. Risk-associated CACNA1C SNPs predicted higher scores on depression and anxiety questionnaires in healthy individuals,Reference Erk, Meyer-Lindenberg, Schnell, von Boberfeld C, Esslinger and Kirsch3,Reference Roussos, Giakoumaki, Georgakopoulos, Robakis and Bitsios4 and decreased the hyperphosphorylated tau/total tau ratio in cerebral spinal fluid (CSF) from patients with bipolar disorder.Reference Jakobsson, Palsson, Sellgren, Rydberg, Ekman and Zetterberg5
Here, we investigate associations between CACNA1C polymorphisms and CSF or serum biomarkers in patients with bipolar disorder and controls. We report that CACNA1C polymorphisms are associated with serum levels of brain-derived neurotrophic factor (BDNF) and the enzyme converting pro-BDNF to mature BDNF, tissue plasminogen activator (tPA).
Method
The sample includes 282 patients with bipolar disorder and 90 healthy controls (see Supplementary Table 1 available at https://doi.org/10.1192/bjp.2017.173 for details). Patients and controls were recruited from the St. Göran Bipolar Project, enrolling patients from the bipolar unit at the Northern Stockholm Psychiatric Clinic, Stockholm, Sweden.Reference Rolstad, Jakobsson, Sellgren, Ekman, Blennow and Zetterberg6 The study was approved by the Regional Ethics Committee in Stockholm (approval number Dnr 2005/554-31/3) and conducted in accordance with the Helsinki Protocol. After a complete description of the study, all enrolled patients and controls consented orally and in writing to participate in the study. In brief, the Swedish version of the Affective Disorder EvaluationReference Sachs, Thase, Otto, Bauer, Miklowitz and Wisniewski7 was the key clinical assessment method. Bipolar diagnoses were made according to DSM-IV criteria as per the Structured Clinical Interview for DSM-IV.Reference Kübler, Gellman and Turner8
Participants were subjected to serum and CSF sampling. Lumbar puncture and serum sampling was performed on the same occasion when patients were in a stable mood phase, as described previously.Reference Jakobsson, Palsson, Sellgren, Rydberg, Ekman and Zetterberg5 Assessment of depressive and manic symptoms are presented in Supplementary Table 7. Participants were genotyped for selected SNPs with KASP technology for Mac OS X (KBioscience, Hoddesdon, UK; see https://www.biosearchtech.com/products/pcr-kits-and-reagents/genotyping-assays/kasp-genotyping-chemistry). MATLAB (Mathworks version R2017b) and SPSS (IBM version 24), both for Mac OS X, were used for statistical calculations. The false discovery rate with threshold 0.1 was used to correct for multiple comparisons.
In total, 39 molecular species (including ratios) were measured from either CSF or serum, using a variety of analytic methods. Analytes included mature BDNF, pro-BDNF and the ratio. See Supplementary Materials and Methods for more details.
Results
Six SNPs for CACNA1C were studied, but only risk alleles for SNP rs1006737 were significantly more common in patients (64 v. 46%, P < 0.01, see Supplementary Tables 2 and 3). In total, 39 molecules were measured in either CSF or serum. Here, we analyse these biomarkers in relation to CACNA1C SNPs in patients with bipolar disorder and controls (see Supplementary Table 4 and Supplementary File 1 for complete results). The only significant difference that withstood correction for multiple comparisons was serum levels of BDNF variants. For patients with the rs2370411 risk allele, the mature BDNF/pro-BDNF ratio was 46% higher than in patients without the risk allele (mean: 1.74 (95% CI 1.46–2.02) v. 1.20 (95% CI 1.00–1.39)). Although there was no significant difference in either mature BDNF or pro-BDNF itself, the former was numerically higher and the latter lower, which is consistent with the ratio being higher. For controls with risk allele rs1006737, the serum level of mature BDNF was 18% lower than in those without the risk allele (27.15 ng/mL (95% CI 25.6–28.7) v. 33.15 ng/mL (95% CI 30.0–36.3)). There were no significant differences in mature BDNF/pro-BDNF ratio or in pro-BDNF in controls, although the first was higher and the second lower in those with the risk allele. As the serine protease tPA is known to convert pro-BDNF to mature BDNF,Reference Pang, Teng, Zaitsev, Woo, Sakata and Zhen9 serum concentrations of tPA were analysed post hoc in relation to CACNA1C genotype. For controls with the rs2370411 risk allele (dominant model), the mature BDNF/pro-BDNF ratio was significantly correlated with tPA (r = 0.56, P < 0.001, see Supplementary Table 5 for complete information). There were no differences between risk allele carriers and those with wild type concerning age, gender, smoking status, body mass index or medication (Supplementary Tables 6 and 7).
Discussion
Here, we screened for differences in biomarkers in both serum and CSF dependent on CACNA1C polymorphisms in patients with bipolar disorder and healthy controls. The main finding was that patients with the CACNA1C risk allele rs2370411 showed a 46% higher mature BDNF/pro-BDNF ratio in serum than patients without the risk allele. Further, in controls, the CACNA1C risk allele rs1006737 was associated with 18% lower mature BDNF serum levels.
The neurotrophic growth factor BDNF promotes the survival of existing neurons and induces differentiation and proliferation of developing neural precursors. By activating the extracellular protease plasmin, tPA regulates the cleavage of the immature form pro-BDNF to mature BDNF.Reference Pang, Teng, Zaitsev, Woo, Sakata and Zhen9 Importantly, pro-BDNF and mature BDNF impose opposite effects on several biological processes via different receptors.Reference Chao and Bothwell10 Multiple studies support the role of BDNF in bipolar disorder.Reference Sodersten, Palsson, Ishima, Funa, Landen and Hashimoto11 In one study, the authors analysed protein levels of different forms of BDNF in post mortem material of a group of patients with major depression, schizophrenia and bipolar disorder.Reference Yang, Ren, Zhang, Chen and Hashimoto12 In the parietal cortex, the concentration of the mature form was significantly lower in patients compared with healthy controls, whereas the concentration of the precursor form was higher. Also, all of tPA, mature BDNF and mature BDNF/pro-BDNF were found to be lower in patients with depression compared with controls.Reference Jiang, Chen, Li, Lu, Yue and Yin13 Our results suggest that BDNF levels in serum are associated with CACNA1C polymorphisms, possibly via tPA. Interestingly, tPA levels were positively correlated with mature BDNF/pro-BDNF in controls, but not in patients, for those with the risk allele rs2370411. The reasons for this discrepancy is unknown. A potential confounder is that lithium has a neuroprotective function by activating BDNF signaling.Reference Hashimoto, Takei, Shimazu, Christ, Lu and Chuang14 However, there was no difference in the proportion of patients treated with lithium in the two groups.
It remains a question for further research to unravel the mechanism by which CACNA1C genetic variants lead to changes in BDNF expression. It is known that BDNF expression is regulated by Ca2+ signalling via activation of cAMP response element by CaM kinase IV.Reference Shieh, Hu, Bobb, Timmusk and Ghosh15 The Ca2+ channels that are encoded by CACNA1C are furthermore known to efficiently regulate transcription via cAMP response element.Reference Zhang, Fu, Altier, Platzer, Surmeier and Bezprozvanny16 One may therefore hypothesise that risk variants of CACNA1C lead to changes in Ca2+ signalling, and thus altered BDNF-dependent transcription. Although the CACNA1C risk allele rs1006737 has gained widest interest in relation to bipolar disorder,Reference Ferreira, O'Donovan, Meng, Jones, Ruderfer and Jones2 rs2370411 has also been reported to be involved in mood-related behaviour in humans and to interact with gender.Reference Dao, Mahon, Cai, Kovacsics, Blackwell and Arad17
To conclude, we report that different risk alleles of CACNA1C are associated with serum levels of BDNF in patients with bipolar disorder and controls. This sheds light on possible links between genetic variants and pathophysiological mechanisms in bipolar disorder, and opens up new lines of research.
Supplementary material
Supplementary material is available online at https://doi.org/10.1192/bjp.2019.173.
Acknowledgements
We thank all of the participants who have kindly given their time to participate in our research. We also thank the staff at the St. Göran bipolar affective disorder unit, including study nurses Martina Wennberg, Lena Lundberg, and data manager Mathias Kardell. Yngve Hallström and Auris Pelanis are acknowledged for performing lumbar punctures. We finally thank the BBMRI.se and KI Biobank at Karolinska Institutet for professional biobank service.
Funding
This research was supported by grants from the Swedish Research Council (2018-02653), the Swedish foundation for Strategic Research (KF10-0039), the Swedish Brain foundation, and the Swedish Federal Government under the LUA/ALF agreement (ALF 20170019, ALFGBG-716801), all to M.L.. E.S. was funded by ALF Grants Västra Götalandsregionen. The funding source was not involved in analysis, interpretation of data or in writing the manuscript.
eLetters
No eLetters have been published for this article.