Fig 1.
LL-37 and derived peptides employed in this study.
Panel A. Shows peptide regions corresponding to the parent LL-37 as indicated with pairs of arrows and residue numbers. Note that GI-20 corresponds to residues 13–32 with the positions of I13 and G14 are swapped (9). In addition, the C-terminus of GI-20, as well as FK-13 and KR-12, is amidated. These LL-37 fragments are named in the same manner as LL-37 by taking the first two amino acids in single-letter code followed by peptide length. Panel B. Biophysical properties of the peptides obtained from or calculated using the Antimicrobial Peptide Database (http://aps.unmc.edu/AP). Panel C shows three-dimensional structures of intact LL-37 and its derived fragments. Hydrophobic surfaces are represented with filled space model in white. It is evident that the hydrophobic surfaces of both LL-37 and LL-23 are discontinuous. A mutation of Ser9 to Val9 made LL-23V9 more active against both bacteria [13] and viruses (this study). However, GI-20, corresponding to the central helix of LL-37, had the greater activity against IAV than LL-23V9. These structures are determined by NMR spectroscopy in the presence of membrane-mimetic micelles [28, 29]. The structures of LL-37 and KR-12 are reported in refs. [14], LL-23 and LL-23V9 in ref. [13], GI-20 in ref. [29] and FK-13 in ref [30].
Fig 2.
Effects of LL-37 and derived peptides on replication of Phil82 and PR-8 viral strains.
For panels A-C, aliquots of the Phil82 H3N2 IAV strain were pre-incubated with control buffer (PBS) or the indicated concentrations of LL-37, FK-13, KR-12, LL-23, LL-23V9, or GI-20 peptides and then these samples were used to infect epithelial cell monolayers and tested for infectious foci 24 hrs later using anti-nucleoprotein antibodies and fluorescence detection (see Methods). Panel A shows activity of all of the peptides in MDCK cells. Panels B and C show activity of LL-37, LL-23, LL-23V9, and GI-20 in human bronchial/tracheal (HBTE) and small airway (SAE) epithelial cells, respectively. Panels D-F show results of similar experiments done using the PR-8 virus. LL-23 did not cause inhibition of PR-8 in these experiments. LL-23V9 caused modest inhibition of the virus. Inhibition by LL-23V9 again was significantly greater than LL-23 and LL-37 and GI-20 caused significantly greater inhibition than either LL-23 or LL-23V9. The average number of infected cells per well were 141, 100 and 78 for MDCK, HTBE, and SAE cells, respectively. Results are mean±SEM of 4 or more experiments and expressed as mean±SEM % of control infectious foci. Please refer to S1 Data for raw data for this and other figures. * indicates p<0.05 vs control buffer alone (unpaired t test). & indicates p<0.05 for LL-37 or GI-20 compared with other peptides and control (ANOVA analysis). # indicates where LL-23V9 caused significantly greater inhibition than LL-23 (ANOVA analysis).
Table 1.
Effects of LL-37 derived peptides on cell viability as assessed by LDH assay.
Fig 3.
Effects of LL-37 on plaque assays and neuraminidase inhibition with Phil82 IAV.
Panel A compares the ability of LL-37 and GI-20 to inhibit plaque formation by Phil82 IAV. The virus was either pre-incubated with the peptides prior to the assay, or the peptides were added to the monolayers after allowing 45 min for virus to infect the cells (“Delayed Addition”). Panel B shows neuraminidase inhibition assay results using LL-37, sLL-37, or LL-37-related peptides. Phil82 virus was pre-incubated with the peptides and then NA activity was assayed as described in Materials and Methods. Oseltamivir was used as a positive control and it caused marked reduction of NA activity. Results are mean±SEM of 4 or more experiments. * indicates p<0.05 vs control buffer alone (unpaired t test).
Fig 4.
Effects of LL-37 or sLL-37 on replication of seasonal or pandemic H1N1.
The effects of LL-37 or sLL-37 on the indicated H1N1 strains were tested using the infectious focus assay as in Fig 2. Panel A shows results with the Cal09 pandemic H1N1 strain Cal09 and the seasonal H1N1 strain NY01. Panel B compares effects if LL-37 on two additional recombinant H1N1 strains that had the HA only (Mex 1:7) or HA and NA (Mex 2:6) of the Mex09 pandemic strain combined with the other gene segments from NY01. Panels C and D show results of similar experiments using HBTE and SAE cells, respectively. Results are mean±SEM of 4 or more experiments. * indicates a significant decrease in viral foci (p<0.05) compared to control. # indicates a significant increase in viral foci (p<0.05) compared to control.
Fig 5.
Effects of LL-37 and related peptides on neuraminidase activity of Cal09.
The figures shows effects of LL-37, sLL-37 or related peptides on NA activity of Cal09. NA activity was measured as in Fig 3. Results are mean±SEM of 4 or more experiments.* indicates a significant decrease in viral foci or NA activity (p<0.05) compared to control.
Fig 6.
Effects of LL-37 on viral replication as assessed by quantitative RT-PCR.
Results shown are percent of control viral RNA copy numbers for samples treated with LL-37 or sLL-37. Panel A shows amounts of cell-associated or supernatant virus after 45 min of incubation with MDCK cells. There were no significant differences in viral RNA quantities between control and LL-37 treated samples in panel A except for cell associated NY01 (* indicates p<0.05 vs control). Panel B shows results in which RNA was isolated from cells and supernatants after 24 hours of infection. LL-37 at concentrations ≥0.4μM significantly decreased viral RNA in supernatants and cells for NY01 strain but not for the Cal09 strain in panel B. All results are mean±SEM of 4 or more separate experiments.
Fig 7.
Effects of CRAMP and HNP-1 on replication of seasonal or pandemic H1N1 strains.
The effect of pre-incubating Cal09, NY01, Mex 1:7, or Mex 2:6 H1N1 strains on infectivity was assayed as in Fig 5 using the infectious focus assay. Results are mean±SEM of 4 or more experiments. * indicates p<0.05 vs control buffer. ** indicates p<0.01 vs control buffer.
Fig 8.
Effects of LL-23, LL-23V9 or GI-20 on replication of Cal09 or NY01 in various cell types.
All of the peptides caused significant inhibition of NY01 in MDCK cells and HBTE cells. For SAE cells LL-23V9 and GI-20 caused significant inhibition for NY01 but LL-23 did not. GI-20 caused significant inhibition of Cal09 in all three cell types. LL-23 and LL-23V9 caused modest but statistically significant inhibition of Cal09 in MDCK cells but not in HBTE or SAE cells. LL-23 (11μM) caused significant increase in replication of Cal09 in SAE cells. Results are mean±SEM of 4 or more experiments. * indicates a significant decrease in viral foci (p<0.05) compared to control. ** indicates a significant increase in viral foci (p<0.01) compared to control.
Table 2.
Comparison of approximate 50% viral neutralizing concentrations of antimicrobial peptides in MDCK cells.