A1 water

Phenotype MicroArray

IDENTIFYING ANTIMICROBIALS
The accumulated color is measured over time in an AND THEIR MECHANISM OF ACTION USING
OmniLog that can simultaneously track up to 4,800 PHENOTYPE MICROARRAYS
independent assays. A kinetic response curve, which parallels microbial growth, is generated so that a growth NTRODUCTION
The advent of multiple drug-resistant microbes has prompted renewed interests in finding novel pharmacophores to treat infectious disease. Biochemical and whole cell screening are currently being used to [Oleandomycin] /MIC
discover novel anti-microbials from natural products and chemical libraries. After an inhibitor is found, two 1. Is the natural product a novel antimicrobial? Incubation Time / Hr
2. What is the mechanism by which the novel inhibitor OmniLog Response Curves

S. aureus ATCC29213 incubated with oleanomycin and monitored in an
Determining the identity of an active natural product can be OmniLog at 36oC for 48 hours. Minimal inhibitory concentration (MIC), a lengthy process. Repeated cycles of time consuming in this example, is defined as the absence of growth in a microtiter well chemical fractionation and bioassays are needed to purify a chemical moiety for physical identification methods. Such time investments are wasted if the natural product is already The ability to measure microbial growth kinetics at a variety of inhibitor concentrations allows researchers to quantify biological effects to an unparalleled degree. For novel antimicrobials, a large effort is also needed to Biolog PM technology takes advantage of OmniLog determine their mechanism of action (MOA). Discovering growth data to identify known chemicals and to infer an inhibitor by biochemical screening does not guarantee that it will inhibit its presumed target in vivo. Genetic and biochemical approaches are labor intensive and may fail to PRINCIPLES OF CHEMICAL IDENTIFICATION
discover a MOA. Biolog’s Phenotype MicroArray (PM) The ability to identify chemicals and infer their MOA technology offers a unique way to identify natural product rests on Biolog’s ability to generate high quality hits and to infer a MOA of an inhibitor. isobolograms. Isobolograms are graphs that display the interactions that two inhibitors have on microbial growth. PHENOTYPE MICROARRAY TECHNOLOGY
Typically, isobolograms are generated by mixing two PM technology is a cellular analysis system that combines chemical inhibitors in different proportions at fractional proprietary assays, high-throughput instrumentation minimal inhibitory concentrations (MICs) and then (OmniLog), and software. The assays are pre-filled and determining if those combinations prevent microbial dried in 96-well microplates that can monitor chemical growth. A line is then drawn to separate those sensitivities. Cell response in each assay well is determined combinations of chemical concentrations that allow by the amount of color development produced by reduction growth from those that completely inhibit growth. Depending on the chemicals, one of three interactions can be observed based on the shape of the line drawn in the isobologram. Indifference (or additivity) is recognized when the inhibition by one chemical can be added to the inhibition caused by the other. Synergy is recognized when one chemical increases the inhibitory effects of the other. Antagonism, its converse, occurs when one chemical lessens the inhibition caused by the other. These effects need not be symmetrical. In addition, the degree of synergy and antagonism can be quantified (Figure 3). Cellular Response in a PM Plate
Phenotype MicroArray
The magnitude of synergy, or antagonism, between any two inhibitors can be quantified. Using an array of Antagonism
chemicals, a scoring matrix based on isobologram data is generated for each inhibitor. These matrices are then Indifference
used in clustering algorithms to group the inhibitors based [Chem B]/MIC
on their pattern of synergy or antagonism. A chemical (or natural product) would be identified based on matching its isobologram matrix score to one of the entries in a database. Novel inhibitors that cluster within a group of inhibitors are inferred to have the same MOA. [Chem A]/MIC
S. aureus (or another model cell type) is mixed with known antibiotics at fractional MICs (0, 0.25, 0.5, 0.75 Classical Isobolograms and Chemical Interactions
and 1). The treated cells are added to one chemical sensitivity PM plate that contains an array of inhibitors Lines indicate minimum concentrations at which the combination of also at fractional MICs. The growth of the bacteria at chemical inhibitors A (Chem A) and B (Chem B) have completely inhibited each combination of inhibitor concentration is recorded in bacterial growth. Chemical inhibitors are present as a fraction of their minimal inhibitory concentration (MIC). an OmniLog over 48 Hr. Gradient isobolograms are then constructed using the growth data (Figure 5 is an example Although useful, classic isobolograms ignore the effects inhibitors have on the rates of microbial growth. Biolog’s PM technology captures these growth rate changes over a range of different chemical combinations. This enables the construction of detailed isobolograms that map gradients of microbial growth as a function of chemical concentrations Indifference
Biolog OmniLog PM Isobolograms
[Chem B]/MIC 0.5
S. aureus strain was ATCC29213. Protein inhibitors streptomycin, tetracycline and erythromycin displayed antagonism, indifference and synergy, respectively, in combination with tetracycline. Lines in graphs Growth Gradient
represent inflection times extracted from OmniLog growth data (in Hr): 6-10 (blue), 10-14 (maroon), 14-18 (yellow), 18-22 (light green), 22-26 (purple), 26-30 (salmon). 30-34 (dark blue) and 34-38 (light purple). MIC is defined as a maximal growth rate occurring at 24 Hr. Using Biolog isobologram data, a synergy or antagonism [Chem A]/MIC
magnitude is calculated. These values are used in generating a matrix that describes the chemical Biolog Isobolograms use Microbial Growth Rates to
interaction between the chemicals in the PM plate and the Describe the Interactions between Two Antimicrobials
added inhibitors being tested. The matrix is imported into a standard clustering program to generate the following Gradient of high (red) and low (white) levels of microbial growth observed at different concentration combinations of chemical inhibitors A (Chem A) and B (Chem B). Chemical inhibitors are present as a fraction of their minimal inhibitory concentration (MIC), defined as a maximal growth rate observed at 24 Hr. Normal maximal growth rates occur at 7 Hr. Phenotype MicroArray
their MOA. Fluoroquinones like norfloxacin inhibit topoisomerase IV and have a different cellular MOA than nalidixic acid which inhibits DNA gyrase. The four groups of protein synthesis inhibitors observed are not unexpected. These inhibitors have diverse MOAs, including mis-incorporation of amino acids into proteins, chain initiation and chain termination. Oxolinic Acid
Ofloxacin
Reproducibility of this analysis is demonstrated by the Norfloxacin
Nalidixic acid
clustering of tetracycline, each tested on a different day. Phleomycin
All three of the tetracycline entries grouped together Tetracycline_2
Tetracycline_3
along side its structural analog doxycycline. Tetracycline_1
Doxycycline
Chlormaphenicol
Erythromycin
Oleandomycin
Puromycin
Bacteria are exquisitely sensitive to their chemical Cefotaxime
Cefazolin
environment. Thus, the interaction of the bacteria’s Cefamandole Naftate
biochemical machinery and inhibitory chemicals will Streptomycin
Amikacin
dictate its growth characteristics. Biolog uses OmniLog data to capture growth information, which can be exploited to generate a high-resolution isobologram. The Biolog Isobologram Data used to Group Known synergy and antagonism magnitudes taken from such
Antibacterials against S. aureus
isobolograms can be used to cluster chemicals with known mechanisms of action (Figure 6). Such chemical Synergy (green), antagonism (red) and indifference (dark red, black or information can be used to assign MOA by inference and dark green) observed between inhibitors in the PM plate (top row) was used to differentiate known inhibitors (right side). Synergy and antagonism values were derived from Biolog isobolograms and clustered using the synergy/antagonism pattern (or fingerprint) is already in a manhattan distance complete linkage (maximal distance) algorithm. S. aureus strain ATCC29213 was used. Discernable groups of inhibitors are color-coded. The number following tetracycline refers to the day it was Even greater sensitivity can be obtained with Biolog gradient isobologram data. Expansion of the chemical The chemicals in the PM plate (Figure 6; top row) were library will enable more resolving power between and used to differentiate known inhibitors (Figure 6; chemicals within inhibitor classes. Additional resolving power may in the column). Based solely on the matrix provided by the be obtained considering the detailed shapes of the discriminatory chemicals in the PM plate, Biolog isobologram curves. Using different subsets of chemicals isobologram data were able to cluster the inhibitors by in clustering will further increase resolution within MOA. Indeed, chemicals with identical or nearly identical structures group together as demonstrated by the seven Phenotype MicroArray technology can be used with a 1. All the topoisomerase IV inhibitors (ofloxicin and variety of bacteria and fungal species, including chemical sensitive mutant strains (tolC or acrAB) that can reduce 2. DNA gyrase (nalidixic acid) and DNA nicking the amount of material needed for testing. 3. Tetracyclines (tetracycline and doxycycline). A database is being generated using Staphylococcus 4. Macrolides (oleandomycin and erythromycin) that aureus to report on antibacterial activity. A variety of other bacterial species and yeast could also be employed 6. All of the cephalosporins (cefamandole naftate, 7. Aminoglycosides (streptomycin and amikacin). In S. aureus, the separation of norfloxacin, oflaxacin and oxolinic from nalidixic acid may be due to differences in

Source: http://www.biolog.com/pdf/PM_Application_DrugID_05.pdf