Scientific Publications - Work Done by Microbiology Reader Bioscreen C

Current Pharmaceutical Design, 2002, 8, pp. 99-110

Antimicrobial Compounds  of Low Molecular Mass  are Constitutively Present  in Insects:  Characterisation of  ß-Alanyl-Tyrosine

K. Meylears, A. Cerstiaens, E. Vierstraete, G. Baggerman, C.W. Michiels and A. De Loof, L. Schoofs
 

ABSTRACT

Abstract: The number of bacterial and fungal strains that have developed resistance against the classical antibiotics continues to grow. The intensified search for new antibiotic lead compounds has resulted in the discovery of numerous endogenous peptides with antimicrobial properties in plants, bacteria and animals. Their possible applications as anti-infective agents are often limited by their size, in reference to production costs and susceptibility to proteases. In this article, we report recent isolations of antimicrobial compounds from insects, with molecular masses less than 1 kDa. Experimental approaches are discussed and the first data on the antimicrobial properties of ß-alanyl-tyrosine (252 Da), one of such low molecular mass compounds isolated from the fleshfly Neobellieria bullata, are presented. We also offer evidence for the constitutive presence of antimicrobial compounds in insects of different orders, in addition to the previously identified inducible antimicrobial peptides.

INTRODUCTION

Miracles Don’t Last Forever. The Need for New Antimicrobial Agents

Reports of infectious micro-organisms having acquired resistance against antibiotics for which they were previously susceptible, have become more and more frequent. The indiscriminate and excessive use of antimicrobial agents in human and animal disease treatment and as growth promoters in animal feed, has created a continuous selective pressure, promoting the development of resistant strains. Moreover, the accumulation of spontaneous, resistantrendering mutations over time within a bacterial population and the horizontal spreading of these mutations across major taxonomic barriers, dangerously accelerates the development of strains resistant to a wide range of conventional antibiotics. Confounding were the reports, made for the first time in a number of Japanese hospitals in the late 1950s, revealing the accumulation of several bacterial genes coding for antibiotic resistance on so-called R-plasmids. These plasmids seemed to be freely exchangeable within a bacterial population, thus rapidly creating a subpopulation of multiple resistant bacteria. Where barely 30 years ago we thought that we had finally won the battle against infectious diseases, today bacteria are steadily forcing us into the defensive again. The World Health Organization (WHO, Geneva) estimates today that 1500 people die each hour from an infectious disease, half of these being children under five years of age [www.phrma.org].

An element that is partly responsible for the current lack of effective antibiotics, is the fact that pharmaceutical companies reduced or even stopped the search for new antibiotic agents during the 1960s, when the traditional screening of soil and marine mico-organisms provided no new results. Instead, efforts were directed to the chemical modification of existing antibiotics, to generate second and third generation compounds with enhanced properties. The most recent true novel antibiotic class, the fluoroquinones, were introduced already 30 years ago. Pharmaceutical companies have meanwhile reached the limit of possible chemical modification of existing antibiotics, while they have at each new manipulation eventually been bested by a bacterial countermeasure. As such, there were no new antibiotics available against the new or re-emerging infectious diseases caused by resistant strains.

However, intensified research has meanwhile resulted in two promising new candidates for novel antibiotic classes: The oligosaccharide agents [1] and cationic peptide antimicrobials [2]. Especially this last group of bioactive peptide compounds, derived mainly from animals, has been intensively studied during the last decade [2,3,4], but many questions concerning cytotoxicity, stability and production costs remain.

Insects, a Valuable Source of Antimicrobial Compounds

The search for lead compounds from natural sources has resulted in the isolation of numerous peptides and polypeptides with antimicrobial properties from plants [5], bacteria [6], and mostly animals. Almost 50% of the reported antimicrobial substances were identified in invertebrates and predominantly in insects. The first antibacterial peptides to be isolated from an insect in 1981 were two cecropins [7]. Insects have since then proven to be an important source of antimicrobial peptides. [reviews 8,9]. Being a large group of different species (80% of the fauna presently known to man), with representatives in almost all ecological niches, insects are the most comprehensive group among the animals. The evolutionary success of insects is largely due to their potent innate immune capabilities. Except for the discrimination between antibacterial and antifungal immune responses reported in Drosophila [10], insects do not display any immune specificity or immunological memory, but their response to invading cells is immediate and powerful. The first, cellular line of defence are the circulating haemocytes, the insect blood cells, that will isolate the invading cells and initiate fagocytosis [review 11]. In the mean time, a powerful systemic response is activated, resulting in the synthesis and release of several antimicrobial peptides into the haemolymph. These peptides will then work synergistically to destroy the invading cells, by destructing the bacterial cytoplasmic membrane or inhibiting bacterial housekeeping enzymes. Although similarities are found throughout orders or genera of insects, each species also seems to have developed particular compounds in response to its specific environmental conditions.

The Advantages of Antimicrobial Compounds with Low Molecular Mass

Although several interesting antimicrobial compounds have meanwhile reached the stage of clinical trials [12,13, www.phrma.org], possible applications are often limited to topical treatments. Being mostly peptidergic in nature, these natural antimicrobial compounds are susceptible to proteolytic breakdown by proteases, encountered during internal digestion in the gastro-intestinal tract or transport in the bloodstream, and consequently have too limited half-life upon systemic application. Moreover, most antimicrobial peptides have a high content of Lys and Arg, which are the main target of trypsine-like proteases. As such, it is important to identify smaller, perhaps even non-peptidergic substances with antimicrobial activity, as these can be expected to be more stable. Being small, they are also more likely to be suitable for production by chemical synthesis, thus bypassing the need for more expensive and labour intensive recombinant expression systems.

Such small (less then 1kDa), possibly non-peptidergic compounds with antimicrobial activity do indeed exist in insects. The first report was made by Leem and co-workers in 1996. From immunised adult fleshfly Sarcophaga peregrina, they isolated the inducible antibacterial compound N-β-alanyl-5-S-glutathionyl-3,4-dihydroxyphenylalanine (5-S-GAD, Fig. 1a) with a molecular mass of 573 Da [14]. Two years later in 1998, the same team reported the isolation of the antibacterial and antifungal compound p-hydroxycinnamaldehyde (148 Da, Fig. 1b) from larvae of the saw fly Acantholyda parki [15]. The latter compound was also present when the larvae where not previously immune challenged. In the same year, two small substances were isolated from larvae of the grey fleshfly Neobellieria bullata: β-alanyl-tyrosine (252 Da, Fig. 1c) and 3-hydroxykynurenine (224 Da), which exhibited paralytic activity when injected into adult insects [16]. One of these, β-alanyl-tyrosine, also displays antimicrobial properties, that will be presented in this article.

EXPERIMENTAL APPROACHES: THE ISOLATION

Extraction Procedures

When attempting to isolate antimicrobial peptides, researchers often start the purification from the haemolymph in which these peptides are released, or from the fat body or haemocytes where the synthesis takes place. When looking for a new class of small, perhaps non-peptidergic antimicrobial substances, we chose not to limit our search in this way, bearing in mind that these compounds may very likely also have a role in processes other than immunity (discussed in a later paragraph), and that their presence and production as such may not be restricted to the haemolymph or fat body. When nothing is known about the physiological location of the compounds, or when the insects are smallsized, whole body extracts should be prepared. The insects are washed with water containing disinfectant, then rinsed with deionised water and excess water is removed by filter paper. Next, the insects are thoroughly homogenised in acidified methanol (methanol/water/acetic acid; 90/9/1; v/v/v). It is by using an acidified extraction solvent that we ensure that the larger proteins and polypeptides are denaturated and precipitate. An extraction procedure using boiling water/1M acetic acid resulted in a decrease of antibacterial activity and was discarded (results not shown). The homogenate is centrifuged at 10,000g for 30 min at 4°C. Methanol is removed from the supernatant by evaporation and the residue is dissolved in 0.1% trifluoroacetic acid (TFA) in deionised water. Lipids will disturb the interactions with the HPLC column and can block the passage of the sample. They are removed from the sample by sequential extraction with ethylacetate and nhexane, each time removing the upper lipophilic layer that is consequently formed. Traces of these organic solvents are afterwards removed by evaporation and the remaining hydrophilic fraction is further concentrated under vacuum. The residue is again dissolved in 0.1% TFA in water, filtrated (0.45 µm) and loaded on a solid phase extraction column (Mega Bond Elute C18 cartridges, Varian). The sample is eluted in three or four crude fractions with 0.1% TFA in rising concentrations of acetonitrile (ACN). The fractions are dried under vacuum and screened for antimicrobial activity. The active fractions will be submitted to further purification using High Pressure Liquid Chromatography (HPLC).

Isolating Small Molecules by HPLC

Most known antimicrobial peptides have been purified using mainly traditional reversed phase chromatography. Some antimicrobial substances, however, are very hydrophilic and will consequently not be retained on reversed phase columns. Preliminary screening of extracts prepared as described above, also yielded high antimicrobial activity in fractions eluted with 0.1% TFA in 5% ACN. Table 1 outlines data on the antibacterial activity of the solid phase fractions of a whole body extract of Galleria mellonella larvae. The term “equivalent” is routinely used in literature on isolations, to indicate the quantity of sample theoretically derived from an extract of one animal.

The compounds that elute with 0.1% TFA in 5% ACN bind only weakly to the hydrophobic phase of the column and cannot be purified by reversed phase chromatography. Logically, one would turn to the opposite separation principle, namely normal phase chromatography. But it was observed that these hydrophilic substances do often not readily dissolve in the hydrophobic solvent used for loading the sample onto a normal phase extraction column. In such a situation, ion-exchange and gel filtration chromatography can be used. Reversed phase chromatography can still be helpful as a “subtractive isolation” procedure, in which the non-binding hydrophilic active compounds are separated from the binding hydrophobic compounds. Based on differences in molecular mass or charge, a further purification may then be obtained by respectively gel filtration or ion exchange chromatography.

The salt containing solvents, usually applied to elute the sample from an ion exchange column, may interfere with the antimicrobial screening assay and should be removed before testing. Therefore, this step should be used only when the sample is almost pure and can be detected by its chromatographic absorbance pattern. Also, salts can disturb the ionisation of the sample, needed to determine its mass by mass spectrometry, although Matrix Associated Laser Desorption/Ionisation (MALDI) and nanoflow Electrospray Ionisation (ESI) mass spectrometry can tolerate low amount of salts [17]. For the concentration and isolation of small, hydrophilic compounds the following method is recommended: start with semi-preparative gel filtration; then, subtractive reversed phase isolation can be used; ion exchange chromatography may be the next step; desalt the selected fraction(s) by analytical gel filtration. It is worthwhile to mention the existence of new chromatography columns with higher selectivity for hydrophilic compounds such as the Hypercarb analytical column (Hypersil, Cheshire, U.K.), in which the solid phase is characterised by carbon sheets causing a polar retention effect. Depending on the nature of the sample, such a column may render the use of ion-exchange chromatography superfluous.

An Appropriate Bioassay

Following each chromatographic separation, the fractions are screened for antimicrobial activity using a liquid growth inhibition assay. This test is easy, sensitive and is compatible with large numbers of fractions. The eluted fractions are concentrated under vacuum to remove all traces of TFA and ACN, after which the residue is dissolved in microbial growth medium. The samples are transferred to the flat-bottom wells of a microtiter plate and an inoculum of the chosen test organism is added. The inocula are taken from a microbial suspension obtained by incubating the micro-organisms overnight (bacteria) or longer (fungi), so as to obtain a dense culture of viable cells, and then diluting a sample into fresh medium to a concentration ranging from 10_ to 10⁵ CFU/ml (colony forming units, representing viable bacteria). The volume of the inoculum may vary between 10 to 100 µl. The bioassay incubation time can range from several hours to overnight, depending on the growth rate of the organism. Growth inhibition is evaluated either by direct visual observation, under an inverted microscope, or by absorbance reading at approximately 600nm. In the latter case, microscopic observation remains advisable, as a lower absorbance rate may be due to growth reduction, growth inhibition, or even aggregation of the growing micro-organisms. This test can be conducted as an end-point determination of activity, if growth is evaluated only after a predetermined incubation time. To obtain a more dynamic analysis, and to ensure the detection of possible secondary antimicrobial effects as described in the next paragraph, the samples can be incubated in a microplate absorbance reader with temperature control and mixing possibilities. At regular time intervals, the absorbance is measured and used to draw a growth curve.

One should be cautious when interpreting the growth inhibiting activity of the screened fractions. Different types of antimicrobial activity can be distinguished: (1) a complete lack of growth. Either the micro-organisms were killed by the compound, indicating a bactericidal effect, or the onset of growth was inhibited. In the latter case, the microorganisms will stay dormant but can resume growth when the interfering compound is removed from the medium, thus indicating a bacteriostatic effect. By simple observation of lack of growth in the microtiter plate, the observer cannot distinguish between the two possible causes. (2) If the active compound is present in a sub-optimal concentration, or if the fraction contains a less active compound, growth reduction will be observed, which may become indiscernible after a prolonged incubation time, as the less sensitive cells continue exponential growth and the cells will ultimately grow to the same density as controls. (3) In some cases, the micro-organisms were observed to start exponential growth as normal, but stop multiplying after a certain period of time, rendering a culture at a significantly lower cell concentration than the control cultures i.e. Fig. (3). A possible explanation for this effect is that the active compound is interfering with a bacterial pathway, needed for the synthesis of a growth factor, such as a vitamin or amino acid. When this growth factor becomes exhausted in the medium, further growth is inhibited. (4) In the literature it has been reported that some active compounds need several hours to kill the bacterial cells, i.e. drosocin needs 6-12 hours to display its bactericidal effect in vitro [18]. During the isolation the researcher will usually give priority to the fractions with direct and complete growth inhibiting effect, but it may be very rewarding to isolate the compounds that cause the secondary effects as well.

Usually, fast growing bacteria such as Escherichia coli are used as test organism for the bioassay. As some antibacterial peptides are active only against Gram positive or Gram negative bacteria, it is advisable to test a representative bacterial strain for each class. One may also chose to include a mould or yeast as test organism to screen the extracts for antifungal activity. Available time and amount of sample will determine the choice of test organisms.

EXPERIMENTAL APPROACHES:

CHARACTERISATION OF THE ANTIMICROBIAL ACTIVITY OF Β−AY

Once an active compound has been identified, the nature and extent of its antimicrobial activity are determined. Using the example of β-alanyl-tyrosine (β−AY), various procedures of characterisation are presented next.

Measurement of Growth Inhibiting Activity

The compound was screened for activity against a variety of micro-organisms, including bacteria and fungi, using the liquid growth inhibition assay as described above. Using serial dilutions, the minimal inhibitory concentration or MIC was determined. The MIC is defined as the lowest final concentration of the compound, at which no growth is observed during a predetermined period of incubation. The assay is performed in a microtiter plate, allowing the simultaneous testing of a large number of samples against several test organisms. Rising concentrations of β-AY were dissolved in 40 µl of suitable microbial growth medium and added to a 10 µl inoculum of chosen test organisms, which is taken from a culture prepared by 100-fold dilution of a stationary phase culture. After a predetermined incubation time, growth was evaluated microscopically.

β-AY displayed antibacterial activity against both Gram negative and Gram positive bacteria. Antifungal activity is demonstrated by complete growth inhibition of the yeast Saccharomyces cerevisiae and a 75% growth reduction of the yeast Candida albicans at 40 mM, but no activity was detected for the yeast Candida utilis or the mould Beauveria bassiana up to a concentration of 40 mM. The MIC values are in the millimolar range, while most MIC values found for the inducible antimicrobial peptides are in the micromolar range. The low molecular mass compounds isolated by Leem et al. showed an IC50 (50% inhibitory concentration) ranging from 0.02 to 0.08 mM [14,15]. It should be noted however, that these high concentrations are physiologically relevant in the insect as β-AY is present in very high amounts in the larvae and prepupae (discussed in later paragraph).

To obtain a more dynamic analysis of the growth inhibiting action, the test bacteria were incubated over a longer period of time in a Bioscreen C Microbiology reader (Labsystems Oy, Finland). Absorbance was measured with a 'wide bandpass' filter at regular intervals. The antibiotic chloramphenicol (30 µg/ml) was used as a positive control for the bacteria (except for S. aureus, M. luteus and L. innocua which where found not to be susceptible). The amino acids β-alanine and tyrosine were used as a negative control. The different growth curves as registered by the Bioscreen C Microbiology reader revealed mostly an increased lag-phase and/or a reduced exponential growth rate. In Fig. (2), the growth curve of Shigella sonnei LMG 10473 during the first 10 hours of incubation is presented. β-AY completely inhibited growth of S. sonnei for 21 hours (14 mM), 26 hours (16 mM) and longer depending on the concentrations applied. Most bacterial strains eventually started growing and reached cell densities equal to that of the control, depending on the concentrations of β-AY applied. A different pattern of growth inhibition was observed for Pseudomenas aeruginosa DPMB-B5 and Streptococcus faecalis DPMB-B4. In Fig. (3) the growth curve of the latter is presented. In this case, rising concentrations of β-AY did not delay the onset of exponential growth, but reduced the cell density reached in stationary phase.

Bacteriostatic and Bactericidal Effects of β−AY on B. Thuringiensis

In order to determine whether the effect of β-AY was bacteriostatic or bactericidal, a culture of Bacillus thuringiensis DPMB-B1, was incubated overnight at 37°C in Grace’s Insect Medium/Basal Medium Eagle (1/1) (IM).

A 200 µl sample of bacterial culture in stationary phase (approximately 5x10⁷ CFU/ml) was centrifuged (2000 g / 5 min.) to collect the bacteria. The supernatant was decanted and the cells were resuspended in 200 µl potassium phosphate buffer (100 mM; pH 7), with (test samples) or without (control samples) a given concentration of β-AY. The samples were then incubated at 37°C. At specific time intervals, a sample of the bacterial suspension was taken, diluted and plated on tryptic soy agar to count the number of surviving cells as colony forming units (CFU) after incubation at 37°C for 24 hours.

When B. thuringiensis was transferred from IM growth medium to phosphate buffer (Fig. 4), cells continued to divide slowly during the first hours but then stopped growing and survived at almost constant cell numbers. The phosphate buffer does not contain the nutritients required for the exponential growth of B. thuringiensis. The presence of 24 mM β-AY seems to eliminate the initial period of cell division, but causes no bactericidal effect up to 24 hours incubation. At 32 mM β-AY, the number of viable cells remains stable during 5 hours of incubation, but then decrease from 4.4x10⁷ to 10⁶ CFU/ml in the next 19 hours. This is an indication of a lethal effect of β-AY at concentrations equal to 4 times the MIC value.

The test was repeated with a concentration of 32 mM β- AY, replacing the phosphate buffer with IM growth medium (Fig. 5). The number of viable cells dropped below the detection limit of 20 CFU/ml within the first hour.

The test was repeated, in IM growth medium, but using an initial cell density of approximately 5x10⁴ CFU/ml instead of 10⁷ CFU/ml. In these conditions, cells continued to grow and reached again a cell density of 10⁷ CFU/ml after 4.5 hours of incubation. The bacteriostatic effect of β-AY at a concentration equalling the MIC (8 mM) was confirmed as growth was inhibited up to 24 hours (Fig. 6). At a concentration of 24 mM β-AY, viable cells became undetectable within 30 minutes, providing evidence for a strong and rapid lethal effect at this concentration.

From these experiments we can conclude that lower concentrations of β-AY have a bacteriostatic effect on B. thuringiensis while high concentrations have strong lethal effect. The strong bactericidal effect was observed only for bacteria in microbial growth medium, and not for bacteria in phosphate buffer.

Also, the MIC value that quantifies the bacteriostatic effect, is dependent of the growth medium used. When B. thuringiensis DPMB-B1 was incubated in IM growth medium, the MIC value of β-AY was determined to be 8 mM. However, in TSB growth medium, the MIC was 16 mM. Similar observations were made for the test strains of S. aureus, S. sonnei and S. Enteritidis which were also found to have MIC values in TSB that were approximately twice as high as in NB. Moreover, the MIC value of B. cereus was determined to be 8 mM in IM and NB but 24 mM in TSB.

These findings suggest that the antibacterial activity of β- AY depends on the medium composition and/or on the physiological state of the cells, since the latter can be influenced by the medium composition.

Activity of Structural Analogues

Structural analogues of the compound β-AY were screened for antimicrobial activity using the bioassay as described above. The concentrations tested were equal to and the double of the MIC value as determined for β-AY on the respective micro-organisms in TSB or MB growth medium. The results are displayed in table 3.

The negative results with β-AS, β-AV and β-AF indicate that the hydroxylated phenol ring on the second amino acid, is a requirement for activity. The modified amino acid β- alanine also seems to play an important role in retaining activity since Fmoc-β-AY did not inhibit growth and AY was much less active.

Chromatography was used to determine the presence of β-AY after incubation at higher temperatures. Comparison of HPLC profiles indicated that no degradation had occurred during 1 week of incubation at 36°C in an aqueous solution, or 3 days incubation at 36°C in Ham’s F12 cell culture medium. After incubation, the solutions of β-AY were dried and resuspended in bacterial growth medium to a concentration equal to the MIC value of B. thuringiensis DPMB-B1. The endpoint growth inhibition assay as described above, indicated that antibacterial activity had remained intact. Moreover, an aqueous solution of β-AY, autoclaved at 120°C for 20 minutes, then dried and resuspended in growth medium to MIC concentration, still inhibited growth of B. thuringiensis DPMB-B1. These data indicate a heat-stable antimicrobial activity of β-AY.

The chemical structure and low molecular mass of β-AY, make this modified dipeptide an unlikely target for proteases. However, in the feeding larvae and prepupae of Sarcophaga, the concentration of β-AY rises to very high levels in the haemolymph, which are quickly reduced at the onset of pupariation due to the action of specific hydrolases from the fat body [19]. The identity of these hydrolases and their presence in vertebrates has not yet been described.

Cytotoxicity

Although injection of β-AY in adult insects causes paralysis [16], no cytotoxic effect could be observed in vitro on insect neuronal cells. When applied to rat C6 glioma cells, rat primary neurons or human neuroblastoma cell lines IMR32 and SHSY5Y, no cytotoxic effect of β-AY was detected up to 10 mM concentrations [16]. Intraspinal injection in rats gave no significant effect at concentrations as high as 198 mM [Klement E. et al., unpublished results]. However, the effects of β-AY application in vertebrates must be further examined. Besides for clinical applications, β-AY may be used as antimicrobial additive to cell culture growth media, or as preservative in foods, pharmaceuticals, cosmetics and other products susceptible to microbial deterioration.

MODE OF ACTION

When reviewing the thus far elucidated mode of action of antibacterial peptides isolated from insects [reviews 20,9], roughly two major groups can be considered:

(1)

Most of the inducible antibacterial peptides act on the bacterial membrane, either by physical disruption or disturbing membrane synthesis: I.e Defensins [21] and cecropins [22] kill bacteria by perforating the cytoplasmic membrane. Bacteria are killed immediately, i.e. one minute of contact with 0.5 µM defensin will effectively kill M. luteus in the growing and resting phase [23,24]. On the other hand, attacins only partially integrate in the outer membrane, disturbing the incorporation of outer membrane proteins, which results in the inhibition of their synthesis at the transcriptional level, and an increase of the membrane permeability [25].

(2)

Another group of antimicrobial peptides act by specific inhibitory binding of bacterial housekeeping proteins. I.e. drosocin, pyrrhocoricin and apidaecin act only on Gram negative bacteria. Their activity is found to be stereospecific [18, 26,27] and bacterial cell death in vitro ensues only after several hours. Recent reports show that antimicrobial activity is correlated to the specific inhibition of the bacterial heat shock protein DnaK [28,29].

The antibacterial activity of the low molecular mass compound 5-S-GAD can be inhibited by the presence of catalase [14], indicating that its activity is dependent on the production of H2O2. Later it was suggested that two protons are generated from the hydroxyl groups of the dopa substructure, probably via the formation of a quinone and linked to an oxygen molecule by a bacterial terminal oxidase [30]. The induction of apoptosis on a macrophage cell line by 5-S-GAD is also thought to be due to the formation of H2O2 [30]. The observed selective antitumor activity of 5-SGAD [31] is, however, not inhibited by catalase, and may be based upon the observed inhibition of the autophosphorylation of protein tyrosine kinase [30,32]. The composing substructures β-alanyl-dopa and gluthatione did not exhibit this inhibiting effect on tyrosine kinase [33].

The exact mode of action of p-hydroxycinnamaldehyde and β-AY has not been elucidated as yet. However, the cytotoxicity of hydroxylated phenolic compounds and their oxidation products has been long recognised and is thought to be due to the production of free radicals [34,35] and to their inhibitory interactions with proteins in general [36,37,38]. The inhibitory effect of such compounds on micro-organisms is, among other things, dependent on the rate of hydroxylation [39,40]. The main effectors of activity are quinones, which are aromatic rings with two ketone substitutions. Hydroxylated amino acids can be converted into quinones by phenoloxidases such as tyrosinase (Fig. 7).

The bactericidal effect of β-AY at concentrations higher than the MIC value, was rapid for bacterial cells that are suspended in microbial growth medium but not for bacterial cells in phosphate buffer. The most likely explanation is that only actively metabolizing cells are sensitive to the bactericidal effect of β-AY. It is too early to speculate on the mode of action of β-AY, though the lack of a bactericidal effect in phosphate buffer and the small molecular size of the compound make it unlikely that it would have a membrane perforating activity. Rather, the observation that growing cells are more sensitive, suggests that β-AY may interact with a vital metabolic process.

Other low molecular mass molecules with antimicrobial activity from insects remain to be identified. Their small molecular size (less than 1 kDa) however, will attribute to their chances of being able to cross the microbial membrane and exhibit their antimicrobial activity on intracellular processes or structures.

PAST SUCCESSES AND FUTURE PROSPECTS

Successful Isolations from Diptera

So far, all insect low molecular mass compounds with antimicrobial properties have been isolated from flies. The antibacterial 5-S-GAD was isolated from adults of the fleshfly Sarcophaga peregrina [1 4 ] . phydroxycinnamaldehyde was isolated from larvae of the saw fly Acantholyda parki S. [15] and β-alanyl-tyrosine has been isolated from the larvae of the grey fleshfly Neobellieria bullata [16].

Recently, we have isolated and partially identified another low molecular mass compound (438 Da) with antibacterial activity from larvae of the grey fleshfly. Similar to β -AY, this compound is also present prior to immunisation of the insects. During the isolation procedure, the original compound of 438 Da was degraded, but a dominant fragment of 374 Da retained a significant degree of activity. The molecular masses were determined by use of an Electrospray Ionisation-Quadrupole-orthogonal acceleration- Time of Flight (Esi-Q-Tof) mass spectrometer (Micromass, UK). After accurate mass determination, special software (Micromass, UK) is used to calculate possible elementary atom compositions of the compound, using a predetermined set of atomic masses so as to match the determined mass within an accuracy range of 10 parts per million (ppm). A riboflavine-like structure was suggested (Fig. 8).

Further fragmentation by ESI-Q-Tof in MS_ mode of the 374 Da compound and commercial available riboflavin (376 Da) resulted in similar fragmentation spectra, as presented in Fig. (9). The dominant 242 Da fragment that is clearly in common, was in turn subjected to elemental mass composition and a known structure was calculated: dimethylalloxazine, which is the basic substructure of the flavines (Fig. 10). Several fragments obtained from the 374 Da compound (spectrum b) have masses that are exactly 2 Da less than those obtained from synthetic riboflavine (spectrum a). As the mass of the dimethylalloxazine fragment is identical, this mass difference may be due to an oxidation in the saccharide side chain (Fig. 10).

Riboflavin itself (376 Da) was also simultaneously purified from Neobellieria as a less active fraction. An early report already mentioned an antibacterial activity of riboflavin analogues [41]. Unfortunately, not enough material was retained to perform the Nuclear Magnetic Resonance (NMR) analysis needed to elucidate the complete structure of the original and most active 438 Da compound. Efforts will be made to complete the analysis in the near future.

Prospects for Novel Leads

(a) Constitutive Antibacterial Activity

Are these small molecules with antimicrobial activity only present only in the order of Diptera? To answer this question, several crude methanolic extracts from larvae of insects of various orders were prepared as described above. These extracts were loaded on solid phase C18 columns and eluted in 4 crude fractions. These crude fractions were screened for antibacterial activity (Table 4) using the insect pathogen Bacillus thuringiensis LMG 7138 in the liquid inhibition assay as described previously.

Highest antibacterial activity was found in the fractions eluting with 0.1% TFA in 0% or 30% ACN, with the exception of the Spodoptera extract, which displayed high antibacterial activity in all fractions. It is important to note, that the animals were neither previously injured nor artifically infected in order to induce an immune response prior to extraction. Previously known antibacterial peptides from insects have been isolated only after bacterial induction of the immune response. The data presented in Table 4 however, indicate the presence of a constitutive reservoir of antibacterial substances in unchallenged insect larvae. As the larvae of insects, and in particular the larvae of Diptera, often grow and feed mostly on decomposing material, they are intensely exposed to an environment rich in microbes. A constitutive reservoir of antimicrobial compounds may exist in parallel to the induced antimicrobial peptides that are released in the haemolymph in response to septic injury. Synthesis of these induced peptides occurs mainly in the fat body (equivalent of the mammalian liver) and in certain haemocytes, although a more local secondary response in barrier tissues such as tracheal or gut epithelium [42,43] can also be induced in addition to the systemic response. It has been reported however that a constitutive expression and storage of antibacterial peptides, occurs in granular cells of other invertebrates such as the shrimp [44], crabs [45,46] and the mussel [47] as part of the innate immunity. In 1998, the constitutive expression of an antibacterial peptide was observed in insects, more particularly in the midgut of the coconut beetle (Coleoptera) [48] and earlier this year, Lamberty et al. reported the constitutive presence of an antifungal peptide in certain haemocytes and the salivary glands of the fungus-growing termite Pseudocanthotermes spiniger (Isoptera) [49]. It therefore seems that insects have two ways of fighting infection: (1) The infection-induced transcription of genes coding for synthesis of antimicrobial peptides, mainly in the fat body, and their release in the haemolymph. (2) The constitutive production and storage of antimicrobial compounds, particularly in granular cells. These compounds may be destined for release from storage into the haemolymph through exocytosis or cell degranulation, or destined for fusion with microbescontaining phagosomes. Whether an infective induction is an absolute requirement for eventual release of these constitutively synthesised compounds into the haemolymph, or whether some may be continuously secreted, remains to be elucidated.

The mode of fighting infection by granular storage of antimicrobial peptides and release upon contact, is part of the innate immunity of crustaceans [50] and mussels [47], and has been observed for mammalian granulocytes [51,52]. On the other hand, in H. cecropia and G. mellonella it was demonstrated that only the granular haemocyte type internalise bacterial LPS [53], making the presence of storage granules filled with antimicrobial compounds very acceptable. It is obvious that there may be a close link to the cellular immune response, during which different types of haemocytes will concentrate around invading microorganisms [12].

(b) Preliminary Screening of an Extract of Galleria

Galleria mellonella larvae were collected at the wandering stage, when they move away from the food to start pupariation. An acidic methanolic extract was prepared as described above and eluted from solid phase Mega Bond cartridges in four crude fractions. The 30% ACN solid phase fraction was then loaded on a preparative HPLC C18 column. The obtained chromatogram is presented in Fig. (11). The fractions were screened for antimicrobial activity as described above using two bacterial strains and a yeast.

The presence of several active fractions was observed. A compound with activity against both the Gram positive and the Gram negative bacterial test strain as well as the yeast eluted at approximately 7% ACN. Growth inhibiting activity was observed against B. thuringiensis and S. cerevisiae in fractions eluting at 12% and 16 to 18 % ACN. The fraction eluting at 20% ACN displayed only antifungal activity while the fraction eluting at 25% ACN was only active against the Gram negative E. coli. These data indicate the presence of several different antimicrobial compounds present in Galleria larvae, prior to any manual induction of the immune response.

A first attempt to further purify certain fractions was troubled by a persistent loss of activity. Different storage conditions were evaluated and although specific conditions proved to be slightly more suited for specific fractions, no significant differences were observed between storage in acetonitrile or dried, at 4°C or –20°C. We observed several times that activity was lost after further separation on analytical columns. Interestingly, the original fraction still inhibited microbial growth. This observation may be due to conformational changes during the chromatographic separation that render the compound inactive. Stereospecificity of the action of antimicrobial compounds has been reported [18,26,27]. Another possible explanation could be the dependence of activity on the presence of a cofactor that becomes detached, or a synergic action between two different compounds that are separately not or much less active. The observed loss of activity may be a factor to explain the absence of previous reports of isolations of low molecular mass substances with antimicrobial properties from Galleria.

(c)The Constitutive Presence of Antimicrobial Compounds During Development in Tenebrio Molitor

Extracts were made at different developmental stages of the yellow mealworm Tenebrio molitor (Coleoptera), and screened for antimicrobial activity, using a liquid growth inhibition assay.

Table 5 shows that the highest antimicrobial activity is usually found in the 0% and 30% ACN fractions. This could also be deducted from the data in table 4. The highest overall antibacterial and antifungal activity are found in larvae and pupae. When pupating, the animals are vulnerable, as they cannot physically move away from an infectious environment. At this point in development, however, metamorphosing processes take place and many metabolites are formed, and some of these may be effectors of the observed antimicrobial activity (see next paragraph).

It is apparent from our data that antibacterial compounds in Tenebrio molitor are most sensitively detected with B. thuringiensis, which is a known insect pathogen. It follows from this observation that this bacterial species represents a good bioassay test organism to monitor the purification of active compounds from Tenebrio molitor.

NOT JUST FOR DEFENSE

The physiological functions of these low molecular mass compounds in insects, may be diverse and not restricted to immunity. Some may be secondary metabolites from pathways directed towards non-immune functions.

This is observed in plants where several secondary metabolites such as tannins, terpenoids, alkaloids and flavenoids are known to exhibit antimicrobial activity in vitro [54]. These metabolites are however also responsible for various features such as plant pigmentation (quinones and tannins), odour (terpenoids) and flavour. The phenolic compound p-hydroxy-cinnamaldehyde in plants is a precursor of the cell wall component lignine [13], but has also been demonstrated to have antibacterial and antifungal properties in the saw fly [1 5 ] . Recently, hydroxycinnamaldehydes have been reported to be fungicidal, but not bactericidal [55].

In insects, an injury or invasion of various non-self components induces a proteolytic cascade in the haemolymph that will converse the phenoloxidase precursor, present in the haemolymph or released by haemocytes upon induction, into the active enzyme [reviewed in 56]. Phenoloxidase catalyses the synthesis of melanin pigment from phenolic substances. Hence, the deposition of dark pigment in injured cuticle and to the formation of brown nodules [57,58], or melanotic encapsulations [59,60] around invading micro-organisms or parasites. The metabolites (quinones) of this pathway are known to exhibit cytotoxic and antimicrobial activity [see “mode of action”,35].

Also, Natori et al. [30] have reviewed several defence molecules from Sarcophaga, that seem to play a role in development as well as in immunity. Some of these additional functions are based on the same biochemical properties responsible for the antimicrobial activity, while others are surprisingly different, indicating a dual mode of action from the same molecule.

β-alanyl-tyrosine has only been detected in Diptera of the genus Sarcophaga and the tachinid Voria ruralis [61]. Earlier reports [62] considered the high amounts of β-AY in the prepupae of Sarcophaga as a storage form of tyrosine and β-alanine, which are involved in melanisation and sclerotisation of the cuticle and puparium respectively [63,64]. This way the water solubility of tyrosine is increased. Furthermore, the free amino acids are protected from metabolism by other pathways. Another possible physiological role of the β-alanyl-tyrosine present Sarcophaga has been postulated by Chiou et al. [16]. When an extract of wandering larvae of Neobellieria bullata was injected into adults of the same or other insect orders, a paralysing effect was observed. The two responsible compounds were identified as 3-hydroxykynurenine and β- AY. Meanwhile, two more paralytic compounds have been isolated from Tenebrio molitor [review on paralysins ref. 65]. The appearance of these endogenous compounds in insects was highly related to the onset of metamorphosis. Concentrations of β-AY continue to rise during larval development, peak just before pupariation and then quickly drop to undetectable levels in the emerging adult as illustrated in Fig. (12). These observations led to the postulation of a possible role of β-AY in the immobilisation of the larvae, as part of the onset of the pupriation process.

The newly discovered antibacterial effect of β-AY now adds to this range of properties. A single white pupa of Neobellieria bullata contains at least an amount as large as 265 µg β-AY (Fig. 12), as determined by amino acid analysis of the purified fraction and not taking into account the losses during purification, hydrolysis and derivatisation. This high amount corresponds to an approximate concentration of 20 mM in this insect’s body, thus closely approaching or exceeding the MIC value of β-AY against a variety of micro-organisms (Table 2). Apparently, the low specificity of the compound is compensated by its very high concentration present. An in vivo antimicrobial effect of β- AY becomes even more likely if one takes into account the fact that insects contain a reservoir of various antimicrobial compounds which act synergistically. Hence, the MIC value of β-AY may be lower in vivo than in vitro.

Which function of β-AY came first in evolution, and which function(s) were convenient side-effect(s), is intriguing but must remain unanswered.

CONCLUDING REMARKS / HOPES FOR THE FUTURE

The over-prescription and abuse of antibiotics, that has created the selective pressure under which resistant pathogenic micro-organisms have emerged, are being discussed these days. As accumulating evidence and research reports have at last created an important awareness amongst the public and governmental health institutions, a change in practice is being admonished. But in the meantime, the threat of infectious pathogens that are resistant against all the presently available antibiotics is very real.

Many efforts are directed towards the better understanding and possible inhibition of the basic mechanisms of the resistance itself [review 66], as pharmacologists realize that the effective life span of any antibiotic will eventually be limited. Through frequently occurring mutations, pathogenic micro-organisms acquire tools to evade total annihilation. The eventual generation of resistant micro-organisms is unavoidable and the demand for new anti-infective agents is therefore never-ending. A status exists of a continuous race between the development of resistance by the infectious pathogens and the development of new drugs by pharmaceutical research and industry.

Since the discovery of the antimicrobial peptide cecropin in 1981, insects have proved to be a very rich source of substances with antimicrobial properties, which can function as lead compounds for further development of anti-infective agents. Being the largest and most diverse class within the animal kingdom, present in almost all ecological environments, insects represent an extensive source to be further explored. Each insect species is a potential source for one or several new antimicrobial agents, as each species has developed an armory of anti-infective compounds against the specific pathogenic micro-organisms of its particular environment. Using chemical modifications to overcome the limitations of these natural substances, pharmaceutical companies try to develop optimized derivatives for application as new antibiotics. In the mean time, each new discovery represents a hope for a new lead.

Also, similarities between the innate defense systems found in insects and the innate immunity in humans add an extra dimension to the search. Reactions such as phagocytosis and coagulation, and immune mediators such as cecropins or peptidoglycan recognition protein, are some elements of innate immunity that are conserved from insects to mammals [67,68]. Striking structural and functional parallels exist between the NF-κB transcription pathway, regulating the inducible inflammatory responses in mammals, and the activation of the DORSAL gene encoding some of the inducible antimicrobial peptides in Drosophila [69]. By unraveling the insect immune response system, we may find tools to enhance our own.

(Full Text online - PDF)

© 2005 Transgalactic Ltd (manufacturer of Bioscreen C software) | Privacy Statement | P.O. Box 1393, 00101 Helsinki, Finland, phone: +358 9 85172920, fax: +358 9 8749481, e-mail: microbiology@bionewsonline.com
 

Last modified: May 25, 2005