Objective: Cardiac cine magnetic resonance imaging (MRI) is one of the important means to assess cardiac functions and vascular abnormalities. However, due to cardiac beat, blood flow, or the patient’s involuntary movement during the long acquisition, the reconstructed images are prone to motion artifacts that affect the clinical diagnosis. Therefore, accelerated cardiac cine MRI acquisition to achieve high-quality images is necessary for clinical practice. Approach: A novel end-to-end deep learning network is developed to improve cardiac cine MRI reconstruction under free breathing conditions. First, a U-Net is adopted to obtain the initial reconstructed images in k-space. Further to remove the motion artifacts, the Motion-Guided Deformable Alignment (MGDA) method with second-order bidirectional propagation is introduced to align the adjacent cine MRI frames by maximizing spatial-temporal information to alleviate motion artifacts. Finally, the Multi-Resolution Fusion (MRF) module is designed to correct the blur and artifacts generated from alignment operation and obtain the last high-quality reconstructed cardiac images. Main results: At an 8× acceleration rate, the numerical measurements on the ACDC dataset are SSIM of 78.40±4.57%, PSNR of 30.46±1.22dB, and NMSE of 0.0468±0.0075. On the ACMRI dataset, the results are SSIM of 87.65±4.20%, PSNR of 30.04±1.18dB, and NMSE of 0.0473±0.0072. Significance: The proposed method exhibits high-quality results with richer details and fewer artifacts for cardiac cine MRI reconstruction on different accelerations under free breathing conditions.