The voltage-gated sodium channel current (I${}_{\text{Na}}$) is widely found in the atrium and is one of the most important depolarization cation channels in the cardiomyocyte membrane [26]. It is the major determinant of the upstroke of AP. Proteins of Na${}_{\text{V}}$1.5 (encoded by SCN5A gene) are responsible for I${}_{\text{Na}}$ [27]. In Akita type 1 diabetic mice, I${}_{\text{Na}}$ was measured both in isolated RA and LA myocytes [21]. Current-voltage (I-V) curves of atrial myocytes demonstrated that I${}_{\text{Na}}$ was reduced. The decreased I${}_{\text{Na}}$ density of diabetic mice occurred in association with a decline of maximum conductance (Gmax) and a mode straight shift of the voltage dependence of activation. Voltage dependence of inactivation was not altered in atrial myocytes of Akita mice. The expression of SCN5A mRNA and Na${}_{\text{V}}$1.5 proteins were reduced in the atrium of Akita mice compared with normal controls. The alterations of I${}_{\text{Na}}$ resulted in prolonged P-wave duration, and reduced atrial conduction velocity in Akita mice. Acute insulin treatment increased I${}_{\text{Na}}$ due to enhanced insulin signaling through activation of phosphatidylinositol 3,4,5-triphosphate (PIP3).

In atrial myocytes of db/db mice [13], I${}_{\text{Na}}$ amplitude, I${}_{\text{Na}}$ steady-state activation, or fast and slow time constants of I${}_{\text{Na}}$ activation were similar to control mice. The steady-state inactivation curve was shifted to the right, which suggested a larger window current. SCN5A mRNA and Na${}_{\text{V}}$1.5 protein levels was similar in db/db atrium compared with control. In metabolic type 2 DM with STZ-injection followed by diet induced insulin resistance, the density of I${}_{\text{Na}}$ were similar in the control and T2DM rat myocytes [24].

Sodium channels are activated during the depolarization phase and then rapidly deactivated. However, some channels reopen as a late or persistent sodium current (I${}_{\text{Na-L}}$) that participate in repolarization [28]. The basal I${}_{\text{Na-L}}$ is mainly generated from the Na${}_{\text{V}}$1.5 isoform and is regulated by calmodulin-dependent kinase II [29]. Several studies show that the increase of I${}_{\text{Na-L}}$ can markely prolong the duration of AP in cardiomyocytes and is important in the development of AF [30, 31, 32]. I${}_{\text{Na-L}}$ was increased in isolated atrial myocytes of DM mice compared to controls [33]. In diabetic mice, the application of the I${}_{\text{Na-L}}$ inhibitor (GS967) inhibited I${}_{\text{Na-L}}$, shortened APD, and reduced the incidence of AF by high-frequency electrical stimuli. A recent study in knock-in mice fed a high fat diet, which ablates phosphorylation of the Na${}_{\text{V}}$1.5 channel and prevents augmentation of I${}_{\text{Na-L}}$, increased AF inducibility [31]. In conclusion, the increased susceptibility to AF in diabetic mice was associated with increased I${}_{\text{Na-L}}$ and the subsequent prolongation of AP.

The voltage gated calcium channel is another important cation influx channel. The L-type calcium current (I${}_{\text{CaL}}$) contributes to a depolarizing current which is actived during the repolarization phase of AP. It is responsible for the maintenance of the platform stage. Proteins of Ca${}_{\text{V}}$1.1~1.4 and Ca${}_{\text{V}}$3.1~3.3 are responsible for I${}_{\text{CaL}}$ and T-type Ca${}^{2+}$ currents, respectively.

In atrial myocytes of STZ-induced diabetes, the maximum current density of I${}_{\text{CaL}}$ was significantly higher compared with control. The steady-state I${}_{\text{CaL}}$ activation curve was shifted to the left and the activation slope factor was decreased, while the inactivation curve was shifted to the right and the inactivation slope factor was higher in the diabetic group [6]. These results suggested the I${}_{\text{CaL}}$ was easily activated and was difficult to be inactivated in DM. Ca${}_{\text{V}}$1.2 protein expression was also increased in the diabetic atrium. Selective inhibition of protein kinase C (PKC)-$\beta$ using ruboxistaurin (RBX) can reduce nuclear factor kappa-B (NF-$\kappa$B)/transforming growth factor-$\beta$ (TGF-$\beta$)/Ca${}_{\text{V}}$1.2 expression and I${}_{\text{CaL}}$ activation, and inhibit abnormal atrial remodeling in diabetic rats. In ZDF rats, the protein expression level of Ca${}_{\text{V}}$1.2 in the atrium and current density of I${}_{\text{CaL}}$ were significantly lower in the atrial myocytes, while the kinetics of I${}_{\text{CaL}}$ were similar to the control group [18]. In the atrium of metabolic type2 diabetic rats, Ca${}_{\text{V}}$1.2 mRNA and protein expression were significantly decreased, whereas the level of Ca${}_{\text{V}}$3.1 was upregulated [14]. I${}_{\text{CaL}}$ was reduced and the T-type Ca${}^{2+}$ current was increased in diabetic atrial myocytes. Long term rosuvastatin treatment alleviated these pathological changes in diabetic rats. The results of studies involving I${}_{\text{CaL}}$ have not been consistent, and may be related to the use of different animal models and the duration of diabetes.

There are several types of potassium channels in cardiomyocytes. It has been reported that the main repolarizing potassium currents (I${}_{\text{K}}$) are transient outward potassium currents (I${}_{\text{to}}$), rapid-delayed rectifier potassium currents (I${}_{\text{Kr}}$), slow-delayed rectifier potassium currents (I${}_{\text{Ks}}$) and steady-state potassium currents (I${}_{\text{ss}}$) in the human heart ventricle, while they are fast transient-outward potassium currents (I${}_{\text{to, f}}$), ultra-rapid delayed rectifier potassium current (I${}_{\text{Kur}}$) and I${}_{\text{Ks}}$ currents in the atrium. I${}_{\text{to}}$ and I${}_{\text{Kur}}$ participate in the phase 1 repolarization process of myocardial AP, and I${}_{\text{Kr}}$, I${}_{\text{Ks}}$ participate in the phase 2 and phase 3 repolarization process of AP. Proteins of K${}_{\text{V}}$4.2 (encoded by KCND2) and K${}_{\text{V}}$4.3 (encoded by KCND3) are responsible for I${}_{\text{to}}$, and K${}_{\text{V}}$1.5 proteins (encoded by KCNA5) are responsible for I${}_{\text{Kur}}$ [34].

Bohne et al. [13] found atrial I${}_{\text{K}}$, mainly including I${}_{\text{to}}$ and the I${}_{\text{Kur}}$, were decreased in atrial myocytes of db/db mice. The decrease of I${}_{\text{to}}$ occurred in association with reductions in the expression of KCND2 mRNA and K${}_{\text{V}}$4.2 proteins (mRNAs for KCND3 were reduced and K${}_{\text{V}}$4.3 proteins were similar). The reduction in I${}_{\text{Kur}}$ was not related to mRNA or protein expression (no differences in KCNA5 mRNA or K${}_{\text{V}}$1.5 protein levels). There were no differences in calcium-activated potassium currents in atrial myocytes of db/db mice. Atrial current density of I${}_{\text{to}}$ and I${}_{\text{Kur}}$ in ZDF diabetic rats was less than that in controls and the expression levels of the protein K${}_{\text{V}}$4.3 and K${}_{\text{V}}$1.5 were significantly downregulated [18]. No significant differences were found in the kinetics of I${}_{\text{to}}$. Polina et al. [21] also found I${}_{\text{K}}$ carried by K${}_{\text{V}}$1.5 channels were reduced in type 1 diabetic Akita mice. They measured I${}_{\text{K}}$ in atrial myocytes with and without a prepulse to inactivate I${}_{\text{to}}$. Peak total I${}_{\text{K}}$ was reduced in diabetic atrial myocytes while I${}_{\text{to}}$ (the difference currents between measurements with and without the inactivating prepulse) were similar between wildtype and diabetic mice. The I${}_{\text{Kur}}$, as measured by 4-aminopyridine sensitive I${}_{\text{K}}$, was reduced, and western blot showed no differences in K${}_{\text{V}}$4.2 and K${}_{\text{V}}$4.3 protein levels of the atrium from wild-type and diabetic mice; however K${}_{\text{V}}$1.5 protein was reduced with no difference in mRNA expression. Inward rectifier K${}^{+}$ currents (I${}_{\text{k1}}$) mainly affected resting membrane potential. No significant difference in I${}_{\text{k1}}$ densities were found between control and diabetic atrial myocytes [13, 24].

There are numerous studies showing that small conductance calcium-activated potassium channels (SK channels) play important roles in diabetic AF. The SK channels have three isoforms including SK1 (K${}_{\text{Ca}}$2.1, encoded by KCNN1), SK2 (K${}_{\text{Ca}}$2.2, encoded by KCNN2) and SK3 (K${}_{\text{Ca}}$2.3, encoded by KCNN3). The SK currents were significantly reduced and the AP duration was prolonged in atrial myocytes of GK rats [25]. Compared with control rats, the expression of SK2 channel was decreased and the expression of the SK3 channel was increased in atrial myocytes of GK rats. Metformin reversed SK channel alterations in the diabetic atrium. Liu et al. [35] also reported that SK2 protein levels was decreased and SK3 protein elevels were increased in the atrium of T2DM rats. Metformin treatment prevents the SK channel alterations by inhibiting the PKC/extracellular signal regulated kinase pathway. Long term treatment of metformin also upregulated the SK2 channel and downregulated the SK3 channel by inhibiting the nicotinamide adeninedinucleotide phosphate oxidase 4/p38 mitogen-activated protein kinase (MAPK) signaling pathway [36].

Howarth et al. [37] evaluated gene expression in the sinoatrial node of GK rats and found hyperpolarization-activated cyclic nucleotide-gated channels (HCN) were downregulated. The reduction of HCN isoforms were also reduced in the sinoatrial node of diabetic rats induced with STZ injection, indicating HCN might be an important contributor to the dysfunction of sinoatrial node in DM [38]. mRNA and protein expressions of hyperpolarization-activated cyclic nucleotide-gated channel 2 (HCN2) were reduced exclusively in the ventricles of STZ rats [39]. However, HCN2 expression in the atrium of STZ rats and H9c2 cells treated with high glucose were unchanged.

Higher protein expression levels of Na${}^{+}$-Ca${}^{2+}$ exchanger current (NCX) were observed in the STZ-induced diabetic group [6]. Yang et al. [4] observed the electrophysiological abnormalities of diabetic rats were accompanied by more severe oxidative stress and higher protein expression of NCX in the atrium. The protein level of NCX in the atrial tissue of diabetic rats was upregulated without alterations in mRNA. Allopurinol (a xanthine oxidase inhibitor) intervention can downregulate its protein level, which indicates that NCX activation plays a key role in diabetic electrical remodeling of the atrium, and antioxidant treatment improves electrical remodeling by inhibiting NCX expression.