Antimicrobial peptides (AMPs) can rapidly kill bacteria via disrupting the integrity of the cytoplasmic membrane. Although the understanding of the interaction between AMPs and phospholipid membrane is essential for related drug designs, the in situ observation of this process is still challenging, especially at the molecular level. Herein, we develop a new analytical methodology named gradient boosting algorithm-assisted Raman spectroscopy (GB-Raman) for revealing the membrane disruption mechanism of AMPs. The analytical process consists of three steps: first, collecting the Raman spectra of phospholipid membrane in the physiological environment without or with AMPs as two data sets; then, using the well-trained gradient boosting algorithm to automatically extract Raman spectral differences between the two data sets; finally, speculating the membrane disruption mechanisms of AMPs based on the molecular structure information offered by the above spectral differences. When using a well-studied AMP named magainin 2 as a proof of concept, the GB-Raman revealed that the intensity ratio of two vC-C peaks of lipid acyl chains (I1084/I1072) in the phospholipid membrane was increased and their locations were red-shifted after interacting with magainin 2. These spectral changes indicated the disturbance of lipid hydrophobic chain ordering caused by magainin 2, which was matched by previous studies. Another AMP named cathelicidin-BF (BF-30) with an unknown membrane disruption mechanism was also explored. The extracted Raman spectral differences originated from the decrease in the intensity of vC-O-C peak, δC-H peak, and vC-C peaks of lipid acyl chains after interacting with BF-30, implying that BF-30 may disrupt phospholipid membrane like detergents. The above speculation was further verified by other technologies such as isothermal titration calorimetry (ITC). This study opens a new avenue to research on AMPs and provides deep insight into their membrane disruption behaviors, leading to great potential in drug development against bacterial resistance.
Li et al. (Tue,) studied this question.