[gmx-users] Beyond the KALP15 in DPPC tutorial and doing analysis in GROMACS

[Justin Lemkul]

On 2/24/15 10:18 PM, Thomas Lipscomb wrote:
> Dear gmx-users,
> Ok Justin here is the information you asked for:
>

My questions were rhetorical.  I honestly don't have time to through all of this 
and tell you how to do a thesis project :)

If you have specific questions about using GROMACS to carry out specific tasks, 
that's the main purpose of this list.

-Justin

>
> See Figure 3 for the antimcirobial bacterial-type membrane disruption models that the antimicrobial peptide maximin 3 might be using.  Note that all models require the interaction of several maximin 3 molecules not just one versus the membrane.  Once I figure out how to get good data with one maximin 3 molecule I might have time to use GROMACS to simulate several maximin 3 molecules to try to find the correct model.  Then future work would be to use GROMACS to figure out how to decrease the toxicity of maximin 3 to mammalian-type membranes enough that maximin 3 would be a viable topical, injectable, and in pill form antibiotic.
> Also I put on dropbox my folder with my stuff that I completed the KALP15 in DPPC tutorial with:https://www.dropbox.com/s/pn2xzsoxs7n7uag/KALP.zip?dl=0
>
>
>
> Antimicrobial Peptides (AMPs)
> AMPs have four general mechanisms for antimicrobial activity not including their antiviral activity.  The first mechanism is thought to be the killing mechanism of the majority of eukaryotic AMP, therefore since Bombina maxima is a eukaryote, maximin 3 is probably using the first mechanism.  The first mechanism is the formation of ion channels or pores across the cytoplasmic membrane of bacteria, which causes membrane perturbation, dissipation of the electrochemical gradient across the cell membrane, and loss of cell content (Parisien, Allain, Zhang, Mandeville, & Lan, 2007) .  The other three mechanisms are used by other antimicrobial peptides.  The second mechanism is inhibition of cell wall biosynthesis (Parisien, Allain, Zhang, Mandeville, & Lan, 2007) .  The third mechanism kills bacteria by the AMP having RNase or DNase activity (Parisien, Allain, Zhang, Mandeville, & Lan, 2007) .  The fourth mechanism is used by phage tail-like bacteriocins to kill other (Bacterioci!
 n, n.d.) b
acteria through specific binding of bacteriocins to the bacterial receptor, which provokes dispolarization and perforation of the cytoplasmic membrane, inducing membrane perturbations (Parisien, Allain, Zhang, Mandeville, & Lan, 2007) . Bombina maxima is a eukaryote so its maximin 3 is likely using the first mechanism.  Of the first mechanism (formation of ion channels or pores), the three most cited models of antimicrobial activity are the barrel-stave, carpet, and toroidal pore models (Chan, Prenner, & Vogel, 2006) .  In the barrel-stave model (Figure 3 part A), the AMP spans the membrane and forms a pore lined with peptides such that the hydrophobic side of the AMP is exposed to the lipid and the hydrophilic portion of the AMP is exposed to the interior of the barrel (Brogden K. A., 2005) , and the pore dissipates proton gradients, etc.  In the carpet model (Figure 3 part B), the AMPs line up parallel to the membrane surface and form a peptide carpet.  This is followed by!
  a deterge
nt-like action induced by the AMPs that causes pore formation by ejecting micelles.  In the toroidal pore model (Figure 3 part C) pores of various lifetimes are created, containing AMPs as well as lipid molecules that are curved inwards towards the pore in a continuous fashion from the surface of the membrane.  After transient pore formation using the heads of the lipids to form the interior edge (Bertelsen, Dorosz, Hansen, Nielsen, & Vosegaard, 2012) , the AMPs end up in both leaflets of the bilayer, which presents a mechanism of shuttling the peptides inside.  Longer-lived toroidal pores may have a lethal effect similar in mechanism to barrel-stave pores.  In the molecular electroporation model (Figure 3 part D) the cationic AMPs associate with the bacterial membrane and generate an electrical potential difference across the membrane. When the potential difference reaches 0.2 V, it is thought that pores will be generated through electroporation.  The sinking raft model (Fi!
 gure 3 par
t E) proposes that binding of the amphipathic AMPs causes a mass imbalance and consequently, an increase in local membrane curvature.  As the AMPs self-associate, they sink into the membrane, creating transient pores which result in the AMPs residing in both leaflets after their resolution.  All these antimicrobial mechanisms derive from hydrophobic, hydrophilic, and charged interactions of the AMPs with the membrane (Paulson, 2013). https://www.dropbox.com/s/33l10m423mqqpu2/AMPactivity.jpg?dl=0Figure 3: Five models of first mechanism (formation of ion channels or pores) AMP activity (Chan, Prenner, & Vogel, 2006) .  Red represents a hydrophilic surface, while blue represents a hydrophobic surface.  A, B, and C all start from the same conformation, with the AMPs associating with the bacterial membrane (top left).  (A) barrel-stave model, (B) carpet model, (C) toroidal pore model, (D) molecular electroporation model, (E) sinking raft model
>
> Previous Work
> The purpose of the research is to continue the work of the previous students on this project.  The first student, James Paulson, found an antimicrobial peptide, in the maximin family, with the broadest toxicity against microbes (Paulson, 2013) , which was maximin 3.  Maximin 1 to 11 were tested computationally and maximin 3 showed the most toxicity to the widest varietyof microbes (Paulson, 2013) .  Maximin 3 is a cationic α-helical protein that has a total of 27 amino acids - containing 190 atoms - in this sequence: GIGGKILSGLKTALKGAAKELASTYLH (Paulson, 2013) .  Maximin 3’s tertiary structure is not in the protein data bank, but was available from Paulson’s thesis (Figure 4).  Paulson’s maximin 3 structure, generated using I-Tasser, suggests that maximin 3 is an α-helix, as expected from previous fluorescence anisotropy work that measured the length of maximin 3 (Paulson, 2013) .  Paulson’s other maximin 3 structure, generated using Rosetta, had better energy mi!
 nimization
 than the I-Tasser structure but showed maximin 3 not as an α-helix but as a “partially unstructured form with some alpha helical propensity”.  The second student on this project, Joseph DeLuca, did lab work showing that maximin 3 bound more tightly to bacterial-type vesicles (25% POPG and 75%POPC) than mammalian-type vesicles (25% cholesterol and 75% POPC) (DeLuca, 2013) .  Vesicles are spherical lipid bilayers.  DeLuca also created a helical wheel model (Figure 5), which shows that maximin 3 is amphipathic by having one side of the alpha helix be hydrophobic and one side be hydrophilic.  Jillian Glatz and Pedro Lebron showed that maximin 3 caused leakage when applied to bacterial-type vesicles (same composition as before?), suggesting that maximin 3 does disrupt membranes (no thesis found in SUNY Purchase library).
>
> https://www.dropbox.com/s/zcx5seu61s4qchs/maximin3ITASSER.jpg?dl=0Figure 4: Maximin 3 most confident structure (I-Tasser) (Paulson, 2013) , an amphipathic alpha helix
>
>
> Sincerely,
> Thomas
>
>
>
>
>

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================