DESIGN, SYNTHESIS AND ANTIMICROBIAL ACTIVITY OF NOVEL ANTIMICROBIAL PEPTIDES

In the past decade, a large number of naturally occuring thiazole and oxazolecontaining peptides have been isolated from microbial and marine origins. These peptides exhibit important antimicrobial, antiviral, and anticancer activities. Solid phase synthesis of peptides provides the methods to prepare and investigate novel compounds for possible use as antibiotics. Three combinatorial peptide libraries, library 1, an enantiomeric peptide library, library 2, with guanidinium, thiazole and benzothiophene side chains and library 3, designed from a template natural product, microcin Bl 7, were tested against Vibrio anguillarum (wild type), Escherichia coli (ZK 4) and Saccharomyces cerevisiae in zone inhibition and cell culture growth inhibition assays. Library 2, with guanidinium, thiazole and benzothiophene side chains exhibited significant antimicrobial activity towards V anguillarum and E. coli, with greater activity than did the enantiomeric peptide library, library 1, or the library based on microcin Bl 7, library 3. a-Helical antimicrobial peptides act by general permeabilization of bacterial membranes. Pleurocidin, a 25 amino acid peptide was extracted from the skin secretions of winter flounder. Pleurocidin is histidine rich, forms a well-defined amphipathic a-helix and exhibits no toxicity towards eukaryotic cells. Two novel C-terminally truncated pleurocidin amide peptides were synthesized by Solid Phase Peptide Synthesis (SPPS) and their identities were confirmed by Matrix Assisted Laser Desorption Ionization Mass Spectrometry (MALDl-TOF MS). The antimicrobial activities of the two C-terminally truncated peptides were


CHAPTER 2 ANTIMICROBIAL ACTIVITIES OF PLEUROCIDIN AMIDE AND RELATED PEPTIDES TRUNCATED AT THE C-TERMINUS
A library of tetrapeptides with guanidinium, thiazole, and benzothiophene side chains had higher potency (lower MIC values) against the candidate microorganisms than did an enantiomeric library of thiazolylalanine building blocks or a library of designed analogs of a natural antimicrobial peptide, (microcin Bl 7).
The increased positive charge and the presence ofbenzothiophene side chains, which have the ability to intercalate between DNA base pairs may have contributed to the enhanced antimicrobial activity.
Based on these results, it is recommended that future combinatorial peptide libraries should be designed to increase the number of positive charges and the number of building blocks with benzothiophene side chains.

INTRODUCTION
In the past decade, a large number of thiazole and oxazole-containing peptides have been isolated from microbial and marine origins. These peptides exhibit important antiviral, anticancer, antibacterial and antifungal activities< 1 >. GE Proteolysis of the first 26 amino acids Given the structural importance of thiazole and oxazole heterocycles, synthetic combinatorial libraries with unnatural amino acid building blocks containing thiazole and oxazole heterocycles in the backbone of the peptide or on the side chains can be designed and their antimicrobial activities can be investigated.
Library 1 is an enantiomeric tetrapeptide library made of two building blocks, L-3-( 4 .. thiazolyl) alanine and D-3-( 4-thiazolyl) alanine< 21 >.     Library i is a tetrapeptide library of 16 compounds. Library 2 design included a greater number of amino acid building blocks with heterocyclic side chains, such as imidazole (H) and benzothiophene (S). Arginine (R) was introduced to increase the overall positive charge on the tetrapeptidescis).
Increasing the positive charge on the tetrapeptides may increase their transport across negatively charged bacterial membranes as well as increasing their binding affinities to DNA backbones.
The sequences of the designed peptides comprising combinatorial library 2 are shown in Figure 1.5.     Given that the thiazole, oxazole and bisheterocycle moieties may be essential to the antibacterial activities of microcin B 1 7, these moieties were systematically connected in a combinatorial library (library 3).
Compounds L 3 1-9 have a central glycine residue that separates the two heterocyclic moieties, which confers flexibility on the molecules.
Compounds L 3 10-13 have two heterocycles directly attached to each other giving it a certain degree of rigidity which is even higher in compounds L 3 14 and L 3 15 with three heterocycles directly attached to each others.
Compound L 3 16 is a microcin B 17 fragment.
Three combinatorial libraries with thiazole and oxazole-containing amino acids that had been designed and synthesized were tested for cytotoxicity against the fish pathogen, Vibrio anguillarum, the enterobacterium Escherichia coli and the yeast Saccharomyces cerevisiae in cell culture growth inhibition assays and in zone inhibition assays.
Zone inhibition assay is a classical method to screen antimicrobial agents.
In zone inhibition assays, only the lethal activity of the antimicrobial compounds can be assessed after incubating the antimicrobial compounds with growing cultures of micro-organisms on agar plates for a period of 12-24 hours.
In cell culture growth inhibition assays on the other hand, both lethal and inhibit<!)ry effects of antimicrobial compounds can be determined.
It has been reported that synthetic analogs of grarnicidin S antibacterial activities were highly dependent on the type of the assay used with growth inhibition assays showing greater activity against gram-negative bacteria than agar-based assays< 30 >. Factors such as variation in the diffusion rates of the compounds in the peptide libraries under screening may lead to different reported MIC values than cell culture growth inhibition assays.
In this chapter, both zone inhibition assays and growth inhibition assays were used to screen the peptide libraries compounds but growth inhibition assay data was used to determine the MIC values.
The selection of candidate micro-organisms followed that V. anguillarum causes vibriosis, a fish disease that causes severe economic losses. E. coli is an opportunistic enteric pathogen which causes enteric diseases in immunely compromised humans, E. coli (ZK4) is a genetically engineered species that is microcin-sensitive which makes it a good selection to screen compounds in library 3 and compare their activities against microcin B 17. S. cerevisiae is a representative example of an eukaryotic micro-organism.
Species selectivity of the antimicrobial activities of the combinatorial libraries against prokaryotic pathogens (V. anguil/arum and E. coli) and eukaryotic pathogens (S. cerevisiae) were also investigated.

Equipment
Analytical HPLC experiments were performed on a Waters HPLC system which consisted of a gradient controller, two 501 pumps and a 486 tunable UV I VIS absorbance detector set at 220 run and a Vydac 218TP C 18 10 micron (250 x 4.6 mm I. D.) protein and peptide column.
Scanning of the growth inhibition plates was done on a Dynex MRX, (Dynex Technologies Inc, VA).
E. coli (ZK4): E. coli (ZK4) was streaked from a frozen sample (-70°C) on a Luria Bertani (LB 10) medium and incubated at 3 7°C overnight. A string of individually growing colonies was collected with a sterile loop and put into 5 ml of 1 x LB 10 in a sterile centrifuge tube.
The solution was shaken at 37°C for 3 hours. The optical density (OD) was measured at 650 run and the dilution factor to give a bacterial concentration of 2 x 10 3 colony-forming unit (cfu) I ml was calculated using the equation that 1 OD at 650 run is equal to 2.5 x 10 8 cfu I ml.
The bacterial suspension was inoculated into 0.1 x LB 10, 1 % agarose medium and poured into a petri dish. Peptide solution (5 µl) was added to the wells in the agar and incubated for 3 hours at 37°C. Full-strength LB 10, 1 % agarose medium was added on top of the previous layer and the plate was incubated at 37°C overnight and the diameters of the observed zones of inhibition were measured.

V anguillarum:
V anguil/arum was streaked from a frozen sample (-70°C) on a Luria Bertani (LB 20) medium and incubated at 28°C overnight. A string of individually growing colonies was collected with a sterile loop and put into 5 ml of 1 x LB 20 in a sterile centrifuge tube.
The solution was shaken at 28 °C for 3 hours. The optical density (OD) was measured at 650 nm and the dilution factor to give a bacterial concentration of2 x 10 3 colony-forming unit (cfu) I ml was calculated using the equation that 1 OD at 650 nm is equal to 2.5 x 10 8 cfu I ml. The bacterial suspension was inoculated into 0.1 x LB 10, 1 % agarose medium and poured into a petri dish. Peptide solution (5 µl) was added to the wells in the agar and Full strength YPD broth (100 µl) was then added to each well and the absorbance was measured at 570 nm at different time intervals. flow rate 1 ml I min; UV detection at 220 nm.

HPLC Analysis
The absorbance of the peptide peak at 220 nm represents the amount of peptide in a known injected volume of peptide solution and was used to calculate peptide concentrations.

Media Preparation:
Yeast extract, tryptone peptone (pancreatic digest of casein), YPD broth and YPD Bacto agar were from Difeo, sodium chloride (enzyme grade) was from Fisher. Distilled water was obtained from Milli-Q Plus PF system (Millipore).
Add distilled water to 1 liter and autoclave at 120-124°C for 15 min.
Add distilled water to 1 liter and autoclave at 120-124°C for 15 min.

YPD broth
Formula per liter Bacto yeast extract 10 g.
50 grams of YPD broth was dissolved in 1 liter distilled water and sterilized at 120-124°C for 15 min.

YPD Bacto agar
Formula per liter Bacto yeast extract 10 g.

HPLC Analysis of Peptide Libraries
Studies have shown a direct correlation between peptide hydrophobicities and the% CH 3 CN required to elute them from a reverse phase co1umn< 31 >.
The more hydrophobic peptides eluted at higher CH 3 CN percentage.
Peptides in library 2 are arranged in order of decreasing retention times in Table 1.4.
The number of L-3-(3-benzothienyl) alanine (S) residues was correlated with the hydrophobicity of the compounds in library 2.
The presence of 2 histidine and 1 arginine residues makes compound # 8 and compound# 16 the least hydrophobic compounds in the library.
Peptides in libraries 1 and 3 had overall lower retention times than did compounds in library 2.
The lower % CH 3 CN correlated with the less hydrophobic nature of the building blocks used in libraries 1 and 3 compared to the more hydrophobic L-3-(3-benzothienyl) alanine (S) residues used in the design of library 2.

Zone Inhibition Assay
Compounds in library 2 were tested against E. coli ZK4 and V. angui/larum in a zone inhibition assay. Antimicrobial activity was tested using serial dilutions of the peptides, and the diameters of the zones of inhibition were measured after overnight incubation. As expected, the diameters of the zones of inhibition decreased with decreasing peptide concentrations. Tables 1.5.1, 1.5.2, 1.6. l and 1.6.2 list the library 2 peptide concentrations and the measured zones of inhibition of E. coli ZK4 and V. anguillarum respectively.
Sodium hydroxide and microcin B 17 were included in the zone inhibition assay as positive controls.
The MIC* is considered to be the minimum amount of peptide that gives a zone of inhibition.
For compounds in library 2, we also employed growth assays that were more sensitive than zone assays (section 1.3.3). The peptide concentrations tested in the zone inhibition assays were above the MIC values of peptides in library 2 determined using the growth assay technique.
The growth inhibition assays monitor the growth curves of bacteria after a 3 hour incubation of the bacteria with the peptides. This approach aims to elucidate the mechanism by which these peptides exert their inhibitory effects on the bacteria.
In a zone inhibition assay, the diameter of the clearing zone of inhibition corresponds to the bactericidal effects of the compounds tested.
In a growth inhibition assay, the inhibition of bacterial growth is monitored over a long period of time and the difference between bactericidal and bacteristatic effects can be observed.
Library 1 contained enantiomeric and diastereomeric compounds that did not show any variation in their antimicrobial activities. They all had identical MIC* values of 4.5 µmoles, higher than MIC values of library 2 compounds.
The concentration ranges that were tested in the zone inhibition assay were above the MIC values of the compounds in library 2. The activities of the library 2 compounds were comparable against E. coli and V. anguillarum.
Cell culture growth inhibition assays were used to report MIC values of library 2 tetrapeptide amides against E. coli, V. anguillarum and S. cerevisiae (section

Cell Culture Growth Inhibition Assay
The growth inhibition assay is a kinetic assay. Logarithmically growing cultures of E. coli, V anguillarum and S. cerevisiae were diluted into nutrientpoor medium and pre-incubated with the peptides for 3 hours in 96-well plates, cultures were then diluted with nutrient-rich media and the bacterial growth was monitored by measuring the optical density at 570 nm.
Controls in the assay included a freely growing bacterial suspension (positive control), and uninoculated medium (blank). To control for variability in the assay, the experiments were repeated on different days, and in each assay every peptide concentration was run in triplicate.
Time points between 6 and 8 hours (9 and 11 hours post-exposure) were selected to measure the MI Cs of the peptides.
For each peptide concentration, the average of the optical density in three wells was used to calculate the average optical density in the wells at each time point. A plot of the peptide concentration versus the % growth relative to the positive control was used to determine the MIC values. The growth inhibition MIC values of peptides in library 2 against gram negative bacteria, E. coli and V anguillarum were comparable to each other but the activity against a fungus, S. cerevisiae showed a different pattern.
Peptides, L1 1 and L 2 9 both with three L-3-(3-benzothienyl)alanine (S) residues, had a ten fold lower MIC values than peptides L 2 14, L 2 15 and L 2 16 with one (S) residue (Tables 1. 7  The activity of peptides in library 2 against E. coli did not change by changing _ the position of arginine residue between the N and the C terminus for compounds L 2 1, (RSSS-NH 2 ), and L2 9, (SSSR-NH2).
The interchange between L-3-(4-thiazolyl)alanine, (L), and D-3-(4thiazolyl)alanine, (D), at position two did not affect the activity of peptides in library 2 since peptide pairs L 2 2 and L 2 3, L 2 14 and L2 15 had the same MIC values against E. coli while peptide pairs L 2 10 and L 2 11, L 2 6 and L 2 7 had less than two fold differences in their MIC values against E. coli.
This pattern is also observed in the activity of peptides in library 2 against V.
anguillarum. This pattern is lost in the activity of peptides in library 2 against

S. cerevisiae.
A second set of experiments were aimed at further characterization of the interactions of compound in library 2 with V. anguillarum. For this, we chose to test the most active compounds in library 2 L2 1, L2 2, L2 5 and L2 9.
The MIC values were compared at 9 hours and 19 hours after incubating the The observed decrease in bacterial growth at 9 hours may be due to a bacteriostatic effect, where the peptides inhibit the growth of bacteria but the bacteria may recover.
At 19 hours, the inhibition of growth of bacteria is more likely to be due to a bactericidal effect. The MIC values at 19 hours were two fold higher than those at 9 hours.
Peptides in library 1 did not show variations in their antimicrobial activities against E. coli ZK4 and V. angu.illarum. The MIC values were consistently around 6.3 µM.
Although members in library 1 were shown to bind DNA with different binding affinities(32>, this had no measurable effect on their activities against E.
Library 3 peptides did not cause inhibition of the growth of either E. coli or V.
anguillarum in the µmolar concentration range tested. Although the compounds in library 3 were designed analogs of a template natural antimicrobial peptide, microcin B 17, and incorporated certain elements of activity that are unique to the natural compound, this lack of potency compared with its naturally-occuring micrtemplate may be due to the larger size of microcin B 17 and additional intervening amino acids that may play a role in providing an optimum spacing between the heterocyclic-. containing amino acid moieties.

CONCLUSIONS AND FUTURE DIRECTIONS
Library 2 was the most active library among the three libraries tested.
The features that make library 2 more effective may be the presence of arginine residues and I or the benzothiophene moiety.
It has been shown that poly-arginine peptides can help transport larger peptides across electrically charged membranes< 33 >. Bacterial membranes have a negative membrane potential as opposed to the electrically neutral eukaryotic membranes due to the nature and distribution of fatty acids in the lipid bilayer.
The potency of library 2 peptides may be due to improved transport across the bacterial membranes as opposed to the less charged peptides in the other libraries.
Benzothienylalanine residues have unnatural benzothiophene groups in their side chains, which are aromatic planar and have the potential to intercalate between DNA base pairs. Intercalation can cause cytotoxicity due to topoisomerase inhibition. Arginine residues may also enhance localization of the peptide on the DNA by electrostatic attraction to the negatively charged phosphate backbone of DNA.
In order to further investigate the mechanism of action of library 2 peptides, the peptides were assayed for topoisomerase inhibition (Appendix A).
Further directions include designing additional peptide libraries incorporating . more arginine residues and I or incorporating novel amino acid residues with side chains that can intercalate between DNA base pairs.

CHAPTER 2 ANTIMICROBIAL ACTIVITIES OF PLEUROCIDIN AMIDE AND RELATED PEPTIDES TRUNCATED AT THE C-TERMINUS ABSTRACT
Antimicrobial peptides (AMPs) are found in insects, fish and mammals where they play an integral role in the host defense mechanism against pathogenic micro-organisms. Many AMPs act via general permeabilization of the microbial membranes. So-called amphipathic a-helical peptides are an abundant and widespread class of AMPs. However, such short peptides are generally unstructured in solution, yet due to their cationic nature they are electrostatically attracted to negatively charged lipids and form a-helices in the microbial membrane. Two extensively-studied AMPs were isolated from insects. Cecropin A was isolated from the giant silk moth (Hylophora cecropia) while mellitin was isolated from bee venom (Apis mellifera) and unfortunately had a strong hemolytic toxicity.
In order to maximize the antimicrobial activity while minimizing undesirable properties, such as hemolytic activity, several strategies have been used to develop novel AMPs. Amino acid residues can be deleted or substituted at either the N-terminus or the C-terminus. Hybrid formation of AMPs from different sources has been used to design cecropin -mellitin hybrids having increased antimicrobial activity and lower hemolytic toxicity.
Pleurocidin is a 25-arnino acid peptide from the skin secretions of the winter flounder (Pleuronectes americanus). Pleurocidin could potentially form a single well-defined amphipathic a-helix and exhibits no hemolytic activity.
Previous SAR studies using pleurocidin have indicated that changing the Cterminal carboxylic acid group to a C-terminal amide increases the potency (decreased the MIC value by 4 fold) against Vibrio angu.illarum.
Given the small size of pleurocidin compared to other naturally-occuring AMPs, its lack of hemolytic activity, and its potential ability to form an amphipathic a-helix, we chose to further investigate the SAR of pleurocidin amide .
. In this chapter, we describe the synthesis of two novel C-terminally truncated peptide amides derived from pleurocidin, pleurocidin amide, and CP-29, a cecropin-melittin hybrid. The antimicrobial activities of these peptides were explored against Vibrio angu,i//arum (wild type), a marine bacterium that causes vibriosis in fish, and Escherichia coli (ZK4), a microcin sensitive strain of E.coli.

Problem
Antibiotics are therapeutic agents, interfering with functions that are essential to bacterial growth or survival. Antibiotic resistance has  polymer support and subsequently cleaved off the polymer at the conclusion of the synthesis. Much research has been done since then to improve the type of polymer support, linkers, protecting groups, and the overall synthetic strategy.
To avoid base-catalyzed racemization of amino acid residues during synthesis of the peptide chains, the peptide chains are synthesized from the C-terminus . to the N-terminus using N-protected amino acids. Coupling reagents such as  ser, lys 6 with ile, ile 8 with lys, gly9 with leu, ile 10 with thr, gly 11 with thr, and leu 14 with lys. CP-29 yielded a higher amphipathic a-helix content than CEME and gave a higher activity against gram-negative bacteria< 52 )_

Peptide design
Pleurocidin is an extremely promising AMP first isolated in 1997. Pleurocidin is shorter (25 amino acid residues) than other natural AMPs (e.g.: cecropin A has 37 amino acid residues), forms a welldefined amphipathic a-helix and exhibits antimicrobial activity with no hemolytic activity. In this work, CP-29 and pleurocidin amide were both synthesized using solid-phase methods, and served as positive controls in the antimicrobial assays.

CP-29< 52 )
H-KWKSFIKKLTTAVKKVLTTGLPALIS-NH 2 In this chapter, we describe the further investigation into the SAR of pleurocidin through design and synthesis of two new C-terminally-truncated peptide amides. Pleurocidin was chosen as a good lead compound for the design of novel AMPs due to its potent MIC against Vibrio anguillarum (0.76 uM), absence of hemolytic toxicity, short length, and ability to form an amphipathic a-helix. The goal was to achieve improved antimicrobial activity while simplifying the synthesis by shortening the length of the peptide.
Previous SAR studies on pleurocidin showed that addition of a lysine residue · to the N-terminus of pleurocidin did not enhance its antimicrobial activity against Vibrio anguillarum, but that formation of a C-terminal amide yielded an 8-fold increase in the activity of pleurocidin against Vibrio anguillarum < 53 >_ SAR studies of AMPs that act through formation of pores through the bacterial membranes has shown that a minimum number of 15 amino acid residues is needed for the peptide to span the lipid bilayer of the bacterial membrane< 51 >.
The first designed peptide A, an 18 amino acid-long peptide amide starting from the N-terminus of pleurocidin should therefore be long enough to span the membrane. A comparison ofEdmundson< 54 >helical wheels and hydrophobicity plots (Kyte-Doolittle< 55 > ) of pleurocidin indicated that peptide A could potentially form an amphipathic a-helix, and the C-terminal amide should improve the stability of the helix.

Peptide 1:
H-GWGSFFKKAAHVGKHVGK-NH 2 Upon comparing the sequence of pleurocidin to other AMPs with highly homologous sequences (e.g.: dermaceptin and ceratotoxins), the residues 19-21 appear to be highly conserved, so a longer peptide B was designed to include them. The second designed peptide B, is a 21 amino acid-long peptide amide starting from the N-terminus of pleurocidin:

Peptide 2:
H-GWGSFFKKAAHVGKHVGKAAL-NH 2 To evaluate the effectiveness of the design, the antimicrobial activities of all . four synthetic peptides were evaluated using Vibrio angu.illarum (wild type) and Escherichia coli (ZK4).

Chemicals:
Dimethylformamide ( using PTC amino acid derivatives normalized to alanine and phenylalanine.

Synthesis
MBHA resin, (1 g., sub 1.1 mmoles/g., 1 % DVB) was swollen overnight in 20 ml DMF in an inverting type shaker. DMF was drained the next day, and the resin was washed with DCM (2 x 20 ml). The synthetic-program used in the synthesis of peptide 1, peptide 2, pleurocidin amide and CP-29 is shown in     (2 x 20 ml), and i-propanol (2 x 10 ml). The peptide-resin was then transferred to the HF reaction vessels (PF A Teflon) and dried overnight under vacuum.

HF Cleavage:
The HF treatment was done in an HF apparatus, and simultaneously accomplished the cleavage of the peptide from the resin, the removal of the 0benzyl protecting group on serine, and the removal of Cl-Z protecting group on lysine. p-cresol (0.5 ml), p-thiocresol (0.5 ml) scavengers were added to the dry resin and 9 ml of anhydrous HF was condensed in the vessel at -50°C in a bath of i-propanol and C0 2 cs)· The vessel was then warmed to 4°C and the mixture was stirred for 1 hour on ice. The HF was subsequently evaporated away with the aid of nitrogen (the HF was trapped with KOH) and then with water aspirator.

Peptide Purification:
Following HF treatment, peptide-resins were typically washed with 5 ml ether to remove most of the scavengers and protecting groups, and then extracted first with 10 ml of 10 % acetic acid (fraction A) and subsequently with 10 ml of glacial acetic acid (fraction B). The combined peptide extracts A and B were lyophilized and then re-dissolved in a minimum volume of 10 % acetic acid for gel filtration chromatography. The concentrated peptide extract was loaded on a Sephadex® G-25 column (31 cm x 3 cm). The crude peptide was eluted with 10 % acetic acid, flow rate 1.5 ml/min., UV-detection at 254 nm.
Fractions (3 ml each) were collected and the initial fractions representing excluded material were pooled and lyophilized (Peak A in Figure 2.4). The crude peptide was then dissolved in deionized water and was analyzed by

Peptide Synthesis
Four solid-phase syntheses of peptide amides ranging from 18-26 amino acids were completed without premature termination in 2-3 days each. Judging by increases in the weight of the resins, the yields of protected peptides ranged from 44.4% for pleurocidin amide, 46.6% for peptide 2 to 48% for peptides 1 and CP-29. Following HF deprotection, cleavage from the resin, and gel filtration chromatography to remove scavengers and protecting groups, the crude product mixtures were characterized using analytical gradient HPLC. It is clear that the same synthetic protocol can lead to variable success in obtaining the desired peptide.
The soiid-phase method yielded more than one peptide species in some crude product mixtures and this was indicated by the multiple peaks observed in the analytical gradient HPLC chromatograms that were obtained following gel filtration. Peptide 2 (Figure 2.7) gave the cleanest chromatogram for a crude product, and therefore it seems that this synthesis went most smoothly.
Peptide 1 (Figure 2.5) also showed one predominant product, although the overall chromatogram was not as clean. Analytical HPLC chromatograms of crude product mixtures from the syntheses of pleurocidin amide (Figure 2.9) and CP-29 (Figure 2.11) contained a significant number of additional peaks, some products having similar retention times to the largest peak, and therefore presenting possible purification problems.
Peptide 1 was extracted from the resin following HF cleavage using 10% acetic acid, applied to a gel filtration column and the first peak that eluted from the column (Sephadex® G-25 column, exclusion ~5,000 Da) was collected and lyophilized to give a yellow solid. In case the peptide was not sufficiently soluble in 10% acetic acid, the same resin was then re-extracted with glacial acetic acid, the extract applied to the same gel filtration column and the first peak was collected and lyophilized to give a white solid. Both excluded peaks were then combined and injected onto the HPLC for analysis of the crude product (Figure 2.5). By experimenting with isocratic conditions on the analytical column, the separation shown in figure 2.6. l was achieved using 25 % acetonitrile. The crude peptide in peak II was then isolated using a preparative HPLC column eluted using 25 % acetonitrile. Individual fractions were re-injected on the analytical HPLC column before pooling those fractions containing peak IL The purity of the final product was calculated to be 84% based on HPLC, because some of peak I was still present.
Peptide 2 was processed in the same manner to achieve >99% purity as shown in figures 2.8. l and 2.8.2. Peptide 2 eluted at 37 % CH 3 CN in the gradient run from 5-85 % CH3CN and was then purified using 27 % CH 3 CN.
CP-29 was purified to 90% from the crude product of the CP-29 synthesis, peak IV eluted at 67 % CH 3 CN in the gradient run, and was purified at 62 % CH3CN (figures 2.11 and 2.12).

MALDI-MS
Matrix-assisted laser desorption ionization (MALDI) time of flight (TOF) mass spectrometry is a relatively new mass spectrometry technique. It has rapidly evolved as a valuable tool for the detection and characterization of biopolymers such as peptides, proteins, oligonucleotides and oligosaccharides.
The use of laser beams to induce ionization without matrix assistance was more or less restricted to small biomolecules (below 1000 Da) bearing aromatic residues. The basic rational to employ matrices for MALDI-TOF MS was to assist in the ion formation of large biomolecules. The matrix candidate should be water soluble, not too volatile and chemically not aggressive. a-Cyano-4-hydroxycinnamic acid has been used successfully as a matrix for MALDI. One of the most important conditions for the matrix to work well is an inti'mate mixing of the analyte and the matrix. Optimal molar ratios range from 1: 1000 to 1: 10,000 (analyte to matrix) so that the analyte molecules are  corresponding to the fragment pleurocidin (13-25) amide.
The actual mass of peptide A was less than the expected mass for peptide 1, indicating that the major product of the synthesis was likely a deletion peptide.
The difference between the found mass (1793.15) and the expected mass ( 1940.07) could indicate the loss of a phenylalanine residue giving a deletion peptide with a calculated mass of (1793.00) for the M+H ion which is in close agreement with the measured m/z (1793.15). No other amino acid deletion could give the observed mass. Since the only two phenylalanine residues are fortunately located adjacent to one another in the sequence of peptide 1, the synthetic product could be unambiguously identified as a 1 7 amino-acid peptide having the sequence: Peptide A (Peptide 1-des-5 or 6 Phe) H-GWGSFKKAAHVGKHVGK-NH2 Peptide B, peak III also had less than the expected mass of CP-29. An alanine residue was probably deleted in the synthesis of CP-29 although the calculated mass still differs from the measured one by two mass units (   Ave. (1) (Ala+Phe) (calc) Table . 2.9 Amino acid analysis (AAA) of acid hydrolysate of purified peptide 2

Monitoring integrity of synthetic peptides
HPLC analysis was used to determine the purity and concentration of each peptide solution used in the antimicrobial assays. The extinction coefficient (E 22 o) of peptide 2 was calculated from the AAA data using Beer's law A=E*b*c. A solution of peptide 2 (100 µl) was acid hydrolyzed using 12 N HCL 25 µl was subsequently derivatized using PITC method. The residue was brought up in 40 µl distilled water and 6 µl was injected in the HPLC system for quantification. The amount of peptide 2 (8450 picomoles) was normalized. to phenylalanine and alanine. HPLC analysis of the intact peptide solution of the same concentration showed an absorbance of 0.0065 AU I µl in the peptide peak indicating that peptide 2 has E 220 0.3466 L *µmole-1 * cm-1 • Since absorbance (A) at 220 nm is primarily due to the amide bonds in the backbone of the peptide, the number of amide bonds is proportional to the intensity of absorbance at 220 run. Therefore the E 22 o for any similar peptide can be determined by adjusting for the number of amide bonds. The data from AAA of peptide 2 was used to calculate an appropriate E 220 and quantify peptide A, pleurocidin amide and peptide B using absorbance at 220 run.  Table 2.10 Calculated extinction coefficients for synthetic peptides using the results obtained from AAA of an acid hydrolysate of peptide 2 and analytical HPLC results on the same solution.

Antimicrobial activities of synthetic peptides:
The synthetic peptide solutions were diluted and tested for their antibacterial activities against Vibrio anguillarum (wild type) and Escherichia coli (ZK4) in cell culture growth inhibition assays. The bactericidal activities of the synthetic peptides were explored by co-incubating peptide solutions, in serial concentrations, with the bacteria for a period of 48 hours. 37°C at an initial cell count (10 5 ) cfu I ml for 23 hours. Bacterial cells were co-incubated with the peptides in 96-well plates at the indicated temperatures and times so that the effect of the peptide on the bacterial growth could be assessed. Each peptide was assayed in duplicate at eleven different concentrations. In addition, tachyplesin, an antimicrobial peptide that acts via a different mechanism than pleurocidin was used as a control peptide.
Bacterial growth was monitored by absorbance at 650 run.

Conclusions
Studies have shown that when gradients of increasing CH 3 CN are used to elute peptides adsorbed on a C 18 column, the more hydrophobic peptides elute at a higher percentage of CH 3 C~3 1 >. Hydrophilic contaminants from the synthesis and HF cleavage should elute first, followed by the peptide products in order of increasing hydrophobicity. The largest peak, indicated by an arrow on each chromatogram, was selected for further characterization. In the case of pleurocidin, when the hydrophobicity of the peptide was evaluated using the Hopp-Woods algorithm, the C-terminal region was markedly more hydrophobic. Thus it was expected that peptide 1 would elute at a lower % CH 3 CN than peptide 2, and this was the case.
Assuming that in each synthetic step, the reaction went to completion, the strongest intensity peak should correspond to the full length peptide, while other peaks likely represent failure sequences (deletion or terminated sequences), that are shorter than the desired product. Since each coupling reaction wa_ s monitored during the synthesis using the ninhydrin test, and the next ainino acid was not added until the ninhydrin test was negative, this assumption seemed justified. Since the synthesis proceeded from the Cterminus of the peptide, any failure sequences would contain the same pattern of amino acids but would be shorter, and thus should elute before the desired full-length peptide. Incompletely deprotected peptides would be expected to be more hydrophobic, and thus elute later than the desired product. In the case of both peptide A and peptide 2, the largest peak was also the last significant peak to elute from the column, and was expected to be the full-length peptide.
As confirmed by the MALDI-TOF results, this strategy worked well in the case of peptide 2, where peak IV in figure 2.8.1 had the correct molecular weight and amino acid analysis. In the case of peptide A, which turned out to be a deletion peptide, missing a phenylalanine residue at position 5, the desired product, peptide 1, would be expected to elute later than the failure sequence which'was purified and characterized. The mass spectrometry results showed that during the solid-phase synthesis of pleurocidin amide and it's analogs, there was a difficult coupling of one of the adjacent phenylalanines, which merits further study. The problem with the synthesis of CP-29 was more obscure, and c<.1.iuld be due to some problem with the peptide synthesizer or the HF cleavage process.
In our experiments, peptide B was not active against either V. angu.illarum or E. coli. The literature reported value of the MIC for pleurocidin amide against V. angu.illarum is 0.74 µMC 53 >. Our value was 6.25 uM and we determined the MBC to be 50 uM.

CHAPTER 3: Synthesis of Novel Thiazole-Containing Amino Acid
Building Blocks For Combinatorial Peptide Libraries.

ABSTRACT
Many thiazole and oxazole-containing natural products exert biological activities e.g.: anticancer, antiviral, antibacterial and antifungal activities.
Thiazole and oxazole rings may be important pharmacophores in these compounds. One traditional approach to thiazole synthesis is through Hantzsch synthesis, a one hundred year old method. The Hantzsch approach requires prior preparation of thioamides, which involves several steps. The thioamides easily revert to nitriles upon heating. Despite the difficulty of preparing the starting materials, the Hantzsch cyclocondensation itself proceeds smoothly in high yields.

INTRODUCTION
Many thiazole and oxazole-containing natural products exert important biological activities e.g.: anticancer, antiviral, antibacterial and antifungal activities. Thiazole and oxazole rings may be important pharmacophores in these compounds(!>. In this chapter, we describe the synthesis of a novel thiazole-containing amino acid building block: Methyl 2-S-(1 '-Boc-2'-phenylethyl) thiazole-4- One approach to the synthesis of thiazole rings is through Hantzsch synthesis ( eq 3 .1 ), a one hundred year old method that, despite its antiquity, is still + eq (3.1) The mechanism of Hantzsch synthesis involves initial nucleophilic attack by sulfur followed by cyclocondensation< 56 >. The Hantzsch approach requires prior preparation of thioamides, which involves several steps. The thioamides easily revert to nitriles upon heating. Despite the difficulty of preparing the starting materials, the hantzsch cyclocondensation itself proceeds smoothly in high yields. Another approach to thiazole rings involves the condensation of C-terminal aldehydes (2) prepared from the corresponding amino acids (5, R=amino acid side chain.) with L-cysteine methyl ester to form a thiazolidine (3)< 57 >_ Thiazolidines can be dehydrogenated by active manganese dioxide in the presence of pyridine in benzene to yield the Boc-thiazole amino acids (4).
L-amino acids are not expected to undergo racemization when they are converted to aldehydes and then condensed with L-cysteine methyl ester< 57 >.
The cyclocondensation is not stereospecific, and is expected to give a mixture of diastereomeric thiazolidines. Chemical shifts (8) are given in ppm from tetramethylsilane (TMS).

RESULTS AND DISCUSSION
The synthetic strategy reported in this chapter is a general method that can be used to prepare heterocyclic amino acids. Replacing Boe-phenylalanine with other Boe-amino acids, it can be used to synthesize additional thiazolecontaining amino acid building blocks for combinatorial library synthesis. The preparation of aldehydes via N-methoxy-N-methylamides is increasing in popularity due to the ease of preparation of amides and the selective reduction to form amino acid aldehyes< 60 ). In the formation of the Weinreb amides, DCC is a classic coupling reagent used to form amides without racemization and the reaction finished within 10 min., the precipitation of dicyclohexylurea pushed the reaction forward.For the reduction of N-methoxy-N-methylamides to yield the amino acid aldehydes, the mechanism is thought to proceed through initial formation of a stable complex that is then hydrolyzed to give the aldehyde< 59 ).
In this section, we describe our methods to assay inhibition of DNA topoisomerases I and II by sixteen compounds from a tetrapeptide library. In a topoisomerase inhibition assay, the active enzyme will relax supercoiled DNA, inhibition prevents relaxation.
Increasing the inhibitor concentration should eventually completely inhibit DNA relaxation. The products are analyzed using agarose gel electrophoresis. Relaxed DNA has a bigger radius than its supercoiled counterpart, thus would migrate slower than supercoiled DNA in an agarose gel. The amount of relaxed DNA is quantitated by scanning the gel for the intensity of fluorescence of the relaxed and supercoiled DNA. Add distilled water to 100 ml, pH = 7 .8.
Add distilled water to 100 ml, adjust pH to 7.4.

Gel Preparation:
1 % Agarose gels were prepared by heating 1 g. of agarose in 100 ml TBE buffer for 2 min., pouring the solution into gel plates and allowing the gel to cool down at room temperature.
Topo I Assay: The smallest effective concentration of topo I, which fully relaxed the supercoiled plasmid pBR322 was first determined, and then utilized in the topo I inhibition assay. The gel results for this determination are shown in figure 4.2.1, and the plot of enzymatic activity was linear up to 1.5 units (Figure 4.2.2). 20 units of enzyme caused a new product to form which had lower mobility than relaxed pBR 322.
The smallest effective concentration of topo I required to unwind > 80 % of 0.05 µg of plasmid in 30 min. at 37°C was 1 units. A unit is defined by the manufacturer as the amount of enzyme required to convert 1 µg of pGM® -9Zf(-) Vector DNA from supercoiled form to relaxed form in 30 min. at 3 7°C.
Relax~tion Assay: In a clean autoclaved microcentrifuge tubes, 3 µl 10 x reaction buffer, 5 µl pBR (0.01 µg I µl), 2 µl topo I in several concentrations (such that total activity 20, 5.7, 1.54, 0.86, 0.19 Units, dilutions of the enzyme were made in 50 % glycerol, 50 % 10 x reaction buffer) were mixed (10 µl total), vortexed and incubated at 37°C for 30 min. The reaction was stopped by adding 2 µl 30 % bromophenol dye and 10 µl was loaded into the wells of the gel. The gels were run at 75 Volts, 15 milliamps. at room temperature for 2 hours in 1200 ml TBE buffer. The gel was stained with ethidium bromide for 1 hour at room temperature. The bands were visualized by fluorescence on a UV light box, photographed with a digital camera and analyz€d using Kodak software.
Topo I Inhibition by Tetrapeptides: Tetrapeptide solutions were prepared in several dilutions using 90 % 1 x reaction buffer, 10 % glycerol. In a clean micro-centrifuge tube, 1 µl of 10 x reaction buffer, 2 µl of peptide solution and 2 µl of topo I (1 U/µl) were mixed, vortexed and incubated at 37°C for 10 min. 5 µl of pBR 322 DNA, (0.01 µg I µl) was added, the tubes vortexed and incubated at 37°C for 30 min. The reaction was stopped by adding 2 µl of 30 % bromophenol dye to each tube and 10 µl of the mixture was loaded into the wells of the gel. The gels were run at 75 Volts, 15 milliamps. at room temperature for 2 hours in 1200 ml TBE buffer. The gel was stained with ethidium bromide for 1 hour at room temperature. The gel data for the inhibition assay is shown in figure 4.3.
Topo II Assay: Topo IJ was purchased from USB Corp., (OH, USA). The smallest effective concentration to fully relax 0.02 µg supercoiled pBR322 was determined and used in topo II inhibition assays. The smallest concentration of topo II to fully relax supercoiled pBR322 was 1 unit. A unit is defined by the manufacturer as the amount of enzyme that fully relaxes 0.3 µg of negatively supercoiled pBR322 plasmid DNA in 15 minutes at 30°C under standard assay conditions. The gel results are shown in figure 4.4.
The wells in the gel were each loaded with 20 µland the gels were run at 75 Volts, 15 milliamps. for 2 hours in 1200 TBE at room temperature. The gels were stained for 1 hour with ethidium bromide at room temperature. The bands were visualized by fluorescence on a UV light box, photographed with a digital camera and analyzed using Kodak software.
Each well in the gel was loaded with 20 µl of the reaction mixture and the gels were run at 75 Volts, 15 milliamps. for 2 hours in 1200 TBE at room temperature.
The gels were stained for 1 hour with ethidium bromide at room temperature. The bands were visualized by fluorescence on a UV light box, photographed with a digital camera and analyzed using Kodak software.
The gel data for topo II inhibition by tetrapeptides are shown in figure 4.5, figure   4.6, figure 4.7 and figure 4.8.

RES UL TS AND DISCUSSION
The smallest effective concentration of topoisomerase I, which fully relaxed the supercoiled plasmid pBR322 was determined and then utilized in the topo I inhibition assay.    Figure 4.6 shows the inhibition oftopo II by compounds RSSH-NH 2 , RLSH-NH 2 , RDSH-NH 2 , and RHSH-NH 2 . Peptides RLSH-NH 2 and RDSH-NH2 did not inhibit topoisomerase II at the concentrations used in the assay while RSSH-NH2 and RHSH-NH2 inhibited topoisomerase II but not in a concentrationdependant manner. concentrations. This might be due to that the concentrations used were higher than the linear concentration range.