WO2018026594A1 - Tunable attribute precision screening assessment platform - Google Patents

Abstract

There is an urgent need for new antibiotics and antimicrobials to treat and cure acute and chronic infections by bacterial pathogens. Current high-throughput practices, do not account for any specifics in bacterial growth or inhibition. Such approaches inefficiently identify many false positives and false negatives of novel compounds. Additionally, this strategy informs little about biochemical changes resulting from bactericidal exposure. We disclose a Precision Throughput approach that can be used to effectively screen new potential antimicrobials under specialized growth conditions relevant to bacterial pathogenesis. This platform can be used to determine new biology and antimicrobial effects of the pathogens Pseudomonas aeruginosa and Group A Streptoccoccus and determine new metrics for characterizing bactericidal activity by a Precision Throughput approach.

Description

TUNABLE ATTRIBUTE PRECISION SCREENING ASSESSMENT

PLATFORM RELATED APPLICATIONS

This application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No. 62/371,464, filed August 5, 2016, which is incorporated herein by reference.

BACKGROUND OF THE INVENTION

New antimicrobials are needed to treat infectious diseases. One clear demonstration of need comes from existing clinical settings as hospital-acquired (nosocomial) infections affect roughly 4% of hospital patients in the U.S. and 8% worldwide. Adding to this nosocomial infection problem is the fact that these infections have an increased likelihood of drug resistance over community-acquired infections. It is further apparent that drug resistance is indeed a global problem. There is also a clear necessity to develop treatments against pathogenic bacteria in a more targeted manner to improve outcomes and limit side effects. Numerous researchers around the globe use a variety of strategies to derive new compounds. Bottlenecks in assessment often arise from existing screening paradigms. In particular, current strategies used to probe for antibiotic activity are grossly inadequate.

Existing approaches are inefficient to screen for compounds that limit bacterial function under relevant pathogenic conditions. Many researchers expend tremendous effort to create and screen compounds to inhibit or kill bacteria; but such "high throughput" screens often fail to identify new compounds that are bactericidal during growth conditions relevant for the bacterium of interest at the point of infection. Conversely, traditional (or "low throughput") bacteriological research that focuses in detail on the growth conditions under which antimicrobials are most effective, are unable to rapidly screen many compounds. With either approach, such work often fails completely to detail the action of an antimicrobial at a single cell level— such relationships are inferred, but not shown, using existing strategies. Consequently, existing methods fall short determining the relevant efficacy profile of any potential drug in vitro.

The identification of a new compound that is inhibitory during growth conditions relevant for the bacterium of interest at the point of infection is often missed during high throughput screening campaigns. In addition, traditional bacteriological assays that focus in detail on the growth conditions under which antimicrobials could be most effective are not designed to screen for many compounds, because the traditional process is low throughput. Accordingly, a solution to these problems would revolutionize how the pharmaceutical industry screens for new anti-infective agents.

SUMMARY A Tunable Attribute Precision Screening (TAPS) antimicrobial assessment platform for evaluating antimicrobial activity of chemical libraries under conditions that are actually relevant to the bacterial infection of interest is disclosed herein. This strategy includes development of a new microfluidic strategy to assess multiple scenarios by creating a vast number of well-defined growth environments, which can be used to research bacterial behavior and antimicrobial action with precise detail. This transformative technology is based on 1) the development of a TAPS platform implemented on a microfluidic device to culture bacteria and screen antimicrobials; and 2) physiologically-relevant media conditions which can provide more informed assessment of antimicrobial susceptibility and resistance on bacterial pathogens using the TAPS platform.

Accordingly, this disclosure provides a method for precision throughput screening of biologically active compounds, the method comprising:

a) mixing a plurality of nutrients in a microfluidic chip, wherein the microfluidic chip comprises:

i) a plurality of microfluidic channels, wherein each microfluidic channel has an inlet for a nutrient that flows hydrodynamically to the microfluidic channel, and wherein the microfluidic channels are communicatively connected with a plurality of dilution channels for mixing of the nutrients; and

ii) an L x M array of cell culture chambers configured to receive the flow of nutrients from one or more microfluidic channels, one or more dilution channels, or a combination thereof, wherein L x M defines the size of the array, and the product of L x M is greater than the sum of inlets;

b) forming an L x M array of varying media conditions relevant to growth of a microorganism by a spatial concentration gradient of cell culture growth media in the L x M array of cell culture chambers, wherein a distinct composition of cell culture growth media is formed by a combination of nutrients flowing into each cell culture chamber;

c) inoculating the microorganism in more than one cell culture chamber, thereby constituting a viable population of microorganisms that can grow at one or more distinct compositions of the varied media conditions in the L x M array of cell culture chambers; d) treating the viable population with a biologically active compound; and e) imaging the viable population by an optical imaging technique or a chemical imaging technique, wherein biological activity of biologically active compounds is assessed by imaging the viable population;

wherein the biological activity of biologically active compounds is precisely screened in an L x M array of varying media conditions relevant to the microorganism.

Thus, this disclosure describes a Tunable Attribute Precision Screening (TAPS) assessment platform that can be used for evaluating the effects of chemical and radiological exposure to cells cultured under conditions that are relevant to the environment of interest. Any culturable cell type, including species of bacteria, archea, yeast, fungi, epithelial cells, and/or mixtures of cells are suitable for assessment by TAPS. Current high-throughput practices to screen potential antibiotics, for example, do not account for any specifics in cell growth or inhibition. Such approaches inefficiently identify many false positives and false negatives of novel compounds. TAPS can be used to screen for antimicrobial activity of chemical libraries under conditions that are actually relevant to the bacterial infection of interest. The TAPS platform utilizes microfluidics to assess multiple scenarios by creating a vast number of well-defined growth environments, which can be used to research cellular behavior and inhibitory action with precise detail. The readout and quantification of inhibitory action via the TAPS platform comes from a combined imaging method. Many possible outcomes besides cell death or rapid growth can signify important effects of chemical and radiation exposure. Assessment of cellular behavior using TAPS is possible by multiple imaging strategies. The key aspects of the TAPS platform are that it allows for precision control of small volumes to grow cells; provides for formation of chemical gradients to test cellular growth and chemical or radiation effects in the same experiment; and this platform is amenable for use in multiple microscopy instruments for both chemical and optical imaging.

BRIEF DESCRIPTION OF THE DRAWINGS

The following drawings form part of the specification and are included to further demonstrate certain embodiments or various aspects of the invention. In some instances, embodiments of the invention can be best understood by referring to the accompanying drawings in combination with the detailed description presented herein. The description and accompanying drawings may highlight a certain specific example, or a certain aspect of the invention. However, one skilled in the art will understand that portions of the example or aspect may be used in combination with other examples or aspects of the invention. Figure 1. Representative examples detailing the influence of changes to carbon and nitrogen source in different media upon biofilm-formation and swarm motility phenotypes of Pseudomonas aeruginosa (P. aeruginosa).

Figure 2. The generation of spatial gradients in microfluidic device resulting in unique culturing environments.

Figure 3A-3F. Representative Lab Micrographs. A) Gfp-tagged P. aeruginosa. B) Co-culture of two strains of P. aeruginosa tagged with Gfp or mCherry . C) Delineation of intracellular protein in Myxococcus xanthus. D) Live-dead image of P. aeruginosa (live = green by Gfp; dead = red by propidium iodide. E-F) Release of nuclear HMGB1 (green) from keratinocytes during GAS infection.

Figure 4A-4C. Raman spectroscopy used to discern planktonic cells from attached biofilms for P. aeruginosa, (a) Raman spectra from planktonic and biofilm cultures, (b) PCA loadings plots from mass spectra reveal clusters of ions found for these different conditions, (c) PCA results for biofilm and planktonic wildtype and a laslrhll quorum sensing mutant.

Figure 5A-5B. Composite CRM images of P. aeruginosa biofilm (a) and planktonic cells (b). Biofilm image reconstructed from glycolipid C-O-C stretching (1000-1050 cm^"1, blue) and protein amide vibrations (1560-1620 cm^"1, red).

Figure 6A-6B. First and second generation microfluidic chip designs for (a) 2 x 1 matrix design that provides four unique culturing conditions, (b) 2 x 2 matrix design that provides 25 unique culturing conditions. P. aeruginosa and GAS can be grown and imaged in the cell culture chambers using optical and chemical imaging.

Figure 7. Carotenoid response of soil isolate Pantoea sp. YR343 to hydrogen peroxide (H2O2) stress. The carotenoid signature region identified at ~1550nm (see inset) is absent in the crtB mutant. When exposed to increased H2O2, the normalized intensity of the YR343 strain decreases markedly with a dose of lmM H2O2.

Figure 8A-8C. Raman analysis of 24h P. aeruginosa pellicle biofilms of FRD1 and PAOIC with different carbon sources, (i) Raman spectral loadings, (ii) principal component heat maps, and (iii) SERS images integrated over 2800-3000 cm^"1 for representative biofilms of P. aeruginosa (A-B) FRD1 (C) and PAOIC. Glucose (A and C) and glutamate (B) were the carbon sources.

Figure 9A-9D. Representative examples detailing the influences of nutrient composition upon antibiotic-mediated growth inhibition for (A-B) P. aeruginosa and (C-D) GAS in batch culture. P. aeruginosa challenged with (A) tobramycin or (B) carbenicillin when growing on either LB or FAB-minimal-glucose media. GAS susceptibility to kanamycin (C-D) varies greatly depending upon the nutrient environment.

DETAILED DESCRIPTION

Current strategies used to probe for antibiotic activity are significantly limited.

Existing approaches are inefficient to screen for compounds that limit bacterial function under relevant pathogenic conditions. Many researchers expend tremendous effort to create and screen compounds to inhibit or kill bacteria; but such "high throughput" screens often fail to identify new compounds that are inhibitory during growth conditions relevant for the bacterium of interest at the point of infection. Conversely, traditional (or "low throughput") bacteriological research that focuses in detail on the growth conditions under which antimicrobials are most effective, are unable to screen many compounds. With either approach, such work often fails completely to detail action of an antimicrobial on a single cell level— such relationships are inferred, but not shown, using current strategies. Herein is described a "precision throughput" antimicrobial assessment platform for screening of antimicrobial activity of chemical libraries under conditions that are actually relevant to the bacterial infection of interest. This disclosure includes development of a new microfluidic strategy to assess multiple scenarios and a combined imaging methodology that can yield detailed metrics to assess bacterial inhibition. The description herein contributes to a revolution regarding how the pharmaceutical industry can screen compounds.

Definitions

The following definitions are included to provide a clear and consistent understanding of the specification and claims. As used herein, the recited terms have the following meanings. All other terms and phrases used in this specification have their ordinary meanings as one of skill in the art would understand. Such ordinary meanings may be obtained by reference to technical dictionaries, such as Hawley 's Condensed Chemical Dictionary 14^th Edition, by R.J. Lewis, John Wiley & Sons, New York, N.Y., 2001.

References in the specification to "one embodiment", "an embodiment", etc., indicate that the embodiment described may include a particular aspect, feature, structure, moiety, or characteristic, but not every embodiment necessarily includes that aspect, feature, structure, moiety, or characteristic. Moreover, such phrases may, but do not necessarily, refer to the same embodiment referred to in other portions of the specification. Further, when a particular aspect, feature, structure, moiety, or characteristic is described in connection with an embodiment, it is within the knowledge of one skilled in the art to affect or connect such aspect, feature, structure, moiety, or characteristic with other embodiments, whether or not explicitly described.

The singular forms "a," "an," and "the" include plural reference unless the context clearly dictates otherwise. Thus, for example, a reference to "a compound" includes a plurality of such compounds, so that a compound X includes a plurality of compounds X. It is further noted that the claims may be drafted to exclude any optional element. As such, this statement is intended to serve as antecedent basis for the use of exclusive terminology, such as "solely," "only," and the like, in connection with any element described herein, and/or the recitation of claim elements or use of "negative" limitations.

The term "and/or" means any one of the items, any combination of the items, or all of the items with which this term is associated. The phrases "one or more" and "at least one" are readily understood by one of skill in the art, particularly when read in context of its usage. For example, the phrase can mean one, two, three, four, five, six, ten, 100, or any upper limit approximately 10, 100, or 1000 times higher than a recited lower limit.

As will be understood by the skilled artisan, all numbers, including those expressing quantities of ingredients, properties such as molecular weight, reaction conditions, and so forth, are approximations and are understood as being optionally modified in all instances by the term "about. " These values can vary depending upon the desired properties sought to be obtained by those skilled in the art utilizing the teachings of the descriptions herein. It is also understood that such values inherently contain variability necessarily resulting from the standard deviations found in their respective testing measurements. When values are expressed as approximations, by use of the antecedent "about," it will be understood that the particular value without the modifier "about" also forms a further aspect.

The terms "about" and "approximately" are used interchangeably. Both terms can refer to a variation of ± 5%, ± 10%, ± 20%, or ± 25% of the value specified. For example, "about 50" percent can in some embodiments carry a variation from 45 to 55 percent, or as otherwise defined by a particular claim. For integer ranges, the term "about" can include one or two integers greater than and/or less than a recited integer at each end of the range. Unless indicated otherwise herein, the terms "about" and "approximately" are intended to include values, e.g., weight percentages, proximate to the recited range that are equivalent in terms of the functionality of the individual ingredient, composition, or embodiment. The terms "about" and "approximately" can also modify the end-points of a recited range as discussed above in this paragraph. As will be understood by one skilled in the art, for any and all purposes, particularly in terms of providing a written description, all ranges recited herein also encompass any and all possible sub-ranges and combinations of sub-ranges thereof, as well as the individual values making up the range, particularly integer values. It is therefore understood that each unit between two particular units are also disclosed. For example, if 10 to 15 is disclosed, then 11, 12, 13, and 14 are also disclosed, individually, and as part of a range. A recited range (e.g., weight percentages or carbon groups) includes each specific value, integer, decimal, or identity within the range. Any listed range can be easily recognized as sufficiently describing and enabling the same range being broken down into at least equal halves, thirds, quarters, fifths, or tenths. As a non-limiting example, each range discussed herein can be readily broken down into a lower third, middle third and upper third, etc. As will also be understood by one skilled in the art, all language such as "up to", "at least", "greater than", "less than", "more than", "or more", and the like, include the number recited and such terms refer to ranges that can be subsequently broken down into sub-ranges as discussed above. In the same manner, all ratios recited herein also include all sub-ratios falling within the broader ratio. Accordingly, specific values recited for radicals, substituents, and ranges, are for illustration only; they do not exclude other defined values or other values within defined ranges for radicals and substituents. It will be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint.

One skilled in the art will also readily recognize that where members are grouped together in a common manner, such as in a Markush group, the invention encompasses not only the entire group listed as a whole, but each member of the group individually and all possible subgroups of the main group. Additionally, for all purposes, the invention encompasses not only the main group, but also the main group absent one or more of the group members. The invention therefore envisages the explicit exclusion of any one or more of members of a recited group. Accordingly, provisos may apply to any of the disclosed categories or embodiments whereby any one or more of the recited elements, species, or embodiments, may be excluded from such categories or embodiments, for example, for use in an explicit negative limitation.

The term "contacting" refers to the act of touching, making contact, or of bringing to immediate or close proximity, including at the cellular or molecular level, for example, to bring about a physiological reaction, a chemical reaction, or a physical change, e.g., in a solution, in a reaction mixture, in vitro, or in vivo. An "effective amount" refers to an amount effective to treat a disease, disorder, and/or condition, or to bring about a recited effect. For example, an effective amount can be an amount effective to reduce the progression or severity of the condition or symptoms being treated. Determination of a therapeutically effective amount is well within the capacity of persons skilled in the art. The term "effective amount" is intended to include an amount of a compound described herein, or an amount of a combination of compounds described herein, e.g., that is effective to treat or prevent a disease or disorder, or to treat the symptoms of the disease or disorder, in a host. Thus, an "effective amount" generally means an amount that provides the desired effect.

The terms "inhibit", "inhibiting", and "inhibition" refer to the slowing, halting, or reversing the growth or progression of a disease, infection, condition, or group of cells. The inhibition can be greater than about 20%, 40%, 60%, 80%, 90%, 95%, or 99%, for example, compared to the growth or progression that occurs in the absence of the treatment or contacting.

The term "substantially" as used herein, is a broad term and is used in its ordinary sense, including, without limitation, being largely but not necessarily wholly that which is specified. For example, the term could refer to a numerical value that may not be 100% the full numerical value. The full numerical value may be less by aboutl%, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, about 10%, about 15%, or about 20%.

The term "biologically active" refers to the beneficial or adverse effects of a compound on living matter, such as, but not limited to bacteria. Among the various properties of chemical compounds, biological activity plays a crucial role since it suggests uses of the compounds in the medical applications. Activity is generally dosage-dependent. General classes of biologically active compounds include, but are not limited to anti -infectives, antimicrobials, antibacterials, antibiotics, bactericides, anti-virals, fungicides which are chemical agents that are adverse to the growth and viability of a population or colony of microorganisms such as a bacterium, virus, or fungus.

The term "medium" (or "media" pi.) used herein refers to the composition of chemical components used in a cell culture to promote or stimulate the growth or

pathogenesis of a microorganism or microorganisms to varying degrees. In some media, the compositions may not be suitable for growth of a viable population or colony of pathogens or microorganisms. Other media may have compositions in which, for example, certain microorganisms thrive. Varied concentrations and ratios of specific chemical constituents can, for example, encompass actual or simulated conditions for pathogenesis of bacteria in animals in which the biological activity of an antibiotic or potential new antibiotic can be measured. Thus, the precision throughput screening of compounds for biological activity that is adverse to a population of inoculated or seeded microorganisms in cell culture media under varied nutrient conditions can be accomplished in this manner.

The term "nutrient" used herein refers to the individual chemical components used together to create a medium such as water, glucose, saccharides, fatty acids, amino acids, vitamins, minerals such as iron and/or magnesium, saliva or other body fluids, and the like.

The term "microfluidic chip" refers to a device which can perform one or more laboratory functions on an integrated circuit of only a few millimeters or a few centimeters to achieve automation and high-throughput screening. A microfluidic chip can handle extremely small fluid volumes down to less than pico liters. The microfluidic chip described herein comprises microfluidic channels and dilution channels as shown in Figures 2 and 6. The microfluidic channel (or micro channel) has inlets for various media and nutrients which flow directly, or indirectly through dilution channels (where nutrients are diluted by mixing through other microfluidic and dilution channels in a network of channels in the microfluidic device) to cell culture chambers. In this manner, various gradients of cell culture media can be efficiently created to form a spatial concentration gradient of distinct nutrient

compositions in an array of cell culture chambers.

The terms "hyperspectral imaging" refers to spectral imaging which collects and processes information from across the electromagnetic spectrum. The goal of hyperspectral imaging is to obtain the spectrum for each pixel in the image of a scene, with the purpose of identifying materials, or detecting processes. Certain objects leave unique 'fingerprints' in the electromagnetic spectrum. Known as spectral signatures, these 'fingerprints' enable identification of the materials that make up a scanned object. For example, the Raman biomolecular signature of an imaged cell colony of microorganisms can be bimolecularly analyzed by various data extraction methods.

Embodiments of the Invention

This disclosure provides various embodiments of a method for precision throughput screening of biologically active compounds, the method comprising:

a) mixing a plurality of nutrients in a microfluidic chip, wherein the microfluidic chip comprises: i) a plurality of microfluidic channels, wherein each microfluidic channel has an inlet for a nutrient that flows hydrodynamically to the microfluidic channel, and wherein the microfluidic channels are communicatively connected with a plurality of dilution channels for mixing of the nutrients; and

ii) an L x M array of cell culture chambers configured to receive the flow of nutrients from one or more microfluidic channels, one or more dilution channels, or a combination thereof, wherein L x M defines the size of the array, and the product of L x M is greater than the sum of inlets;

b) forming an L x M array of varying media conditions relevant to growth of a microorganism by a spatial concentration gradient of cell culture growth media in the L x M array of cell culture chambers, wherein a distinct composition of cell culture growth media is formed by a combination of nutrients flowing into each cell culture chamber;

c) inoculating the microorganism in more than one cell culture chamber, thereby constituting a viable population of microorganisms that can grow at one or more distinct compositions of the varied media conditions in the L x M array of cell culture chambers; d) treating the viable population with a biologically active compound; and e) imaging the viable population by an optical imaging technique or a chemical imaging technique, wherein biological activity of biologically active compounds is assessed by imaging the viable population;

wherein the biological activity of biologically active compounds is precisely screened in an L x M array of varying media conditions relevant to the microorganism.

In additional embodiments of the L x M array of cell culture chambers the value of L is greater than 1, greater than 5, greater than 10, greater than 20, greater than 50, greater than 100, or any whole number between 1 to 1000. In various other embodiments of the L x M array of cell culture chambers the value of M is greater than 1, greater than 5, greater than 10, greater than 20, greater than 50, greater than 100, or any whole number between 1 to 1000. Examples of an L x M array of cell culture chambers include but are not limited to 2 x 2, 3 x 3, 4 x 4, 5 x 5, 6 x 6, 7 x 7, 8 x 8, 9 x 9, 10 x 10, 20 x 20, 100 x 100, 2 x 3, 3 x 4, 4 x 5, 5 x 6, 2 x 4, 4 x 8, 3 x 7, etc.

In various embodiments, a microorganism enters the L x M array of cell culture chambers through a plurality of secondary micro-channels to avoid contamination of the microfluidic channels and the dilutional channels by the microorganism. In other embodiments, the cell culture chambers have an outlet to avoid contamination of the microfluidic channels and dilutional channels by a microorganism from the cell culture chamber.

In some embodiments, the microorganism enters the L x M array of cell culture chambers simultaneously in all the cell culture chambers. In some other embodiments, a different microorganism enters each one or more rows of the L x M array of cell culture chambers, or a different microorganism enters each one or more columns of the L x M array of cell culture chambers.

In some embodiments, the biologically active compound enters the L x M array of cell culture chambers simultaneously in all the cell culture chambers. In some other embodiments, a different biologically active compound enters each one or more rows of the L x M array of cell culture chambers, or a different biologically active compound enters each one or more columns of the L x M array of cell culture chambers.

In some embodiments, a distinct composition of cell culture growth media is formed by a combination of nutrients flowing into each one or more rows of the L x M array of cell culture chambers, or a distinct composition of cell culture growth media is formed by a combination of nutrients flowing into each one or more columns of the L x M array of cell culture chambers.

In other various embodiments, the microfluidic chip is configured for linearly proportional gradients (for example 0, 0.25, 0.5, 0.75, and 1.0), or the microfluidic chip is configured for geometrically proportional gradients for example 0, 1/16, 1/8, ¼, ½, and 1).

Some embodiments include a technique for imaging the viable population of microorganisms is non-invasive to the viable population.

In additional embodiments, the biological activity is assessed by a chemical imaging technique, wherein the chemical imaging techniques comprises infrared spectroscopy, Raman spectroscopy, or hyperspectral imaging. In some embodiments, the chemical imaging technique comprises confocal Raman microscopy (CRM), wherein biological activity is assessed by a Raman biomolecular signature of the viable population. In other embodiments, the Raman biomolecular signature of the colony distinguishes between living microorganisms and dead microorganisms in the viable population.

In yet other embodiments, the Raman biomolecular signature of the viable population is analyzed by principal component analysis (PCA), the Raman biomolecular signature of the viable population is analyzed by hierarchal cluster analysis (HCA), or a combination thereof.

In some embodiments, the viable population is extracted from the culture chamber for analysis by one or more separate analytical methods. In various other embodiments, the microorganism is bacteria. In additional embodiments, the biologically active compound is an anti-infective, an antimicrobial, an antibacterial, an antibiotic, or a bactericide.

In other additional embodiments, the nutrients in the media include succinate, an amino acid, a carbohydrate, an ammonium ion, a nitrate ion, an essential metal ion, a host derived medium, molecular oxygen, or a combination thereof. In yet other additional embodiments the nutrients include metal ions, for example, iron, copper, zinc, manganese, chromium, magnesium, calcium, cobalt, nickel, cadmium, potassium, sodium, lead, tin, or selenium. Embodiments of a host derived medium include, for example, a mixture of saliva and mucus coughed up from the respiratory tract, typically as a result of infection or other disease and often examined microscopically to aid medical diagnosis. Other examples of a host derived medium include, but are not limited to secretions from animal, such as tears, saliva, sweat, and pus.

In some embodiments, the microfluidic chip is permeable to a gas, or is not permeable to gas. In other embodiments, the microfluidic chip is constructed from gas permeable polydimethylsiloxane (PDMS).

In various embodiments, a biologically active compound hydrodynamically flows to the L x M array of cell culture chambers, and the biologically active compound enters the microfluidic chip through an inlet on a microfluidic channel. In other additional

embodiments, the hydrodynamic flow of nutrients is at positive pressure, and the

hydrodynamic flow of a biologically active compound is at positive pressure.

In yet other embodiments, the microfluidic chip has a hydrodynamic flow rate ranging from about 1 nL/hr to about 1000 μΙΤτΐΓ, 1 nL/hr to about 100 μί/τΐΓ, 1 nL/hr to about 10 μί/τΐΓ, 1 nL/hr to about 1 μί/τΐΓ, 1 nL/hr to about 500 nL/hr, 1 nL/hr to about 100 nL/hr, or 1 nL/hr to about 50 nL/hr,. In some other embodiments, more than one biologically active compound is screened on the microfluidic chip.

In various embodiments, the product of L x M is greater than 1 , greater than 10, greater than 20, greater than 30, greater than 40, greater than 50, greater than 100, greater than 200, greater than 500, or greater than 1000. In other embodiments, the sum of inlets is less than or equal to 10, less than or equal to 20, less than or equal to 50, less than or equal to 100, or less than or equal to 1000.

In additional embodiments, the mechanism of action of a biologically active compound is elucidated from the spatial concentration gradient of cell culture growth media in the L x M array of cell culture chambers. Results and Discussion

In general, research to investigate bactericidal or inhibitory effects of antimicrobials follows one of two general strategies: 1) screen many compounds against bacteria grown under monotonous conditions or 2) screen select compounds to determine a specific mechanism of action in detail. We submit that neither of these traditional strategies is efficient to fulfill a societal need to identify new compounds that can treat bacterial infections. Bacterial pathogens are the leading causes of infant mortality, hospital-acquired infection mortality, and overall mortality from communicable diseases in the United States. Studies need to address the urgency for the discovery of new antibiotics given increased resistance to existing compounds. We disclose a strategy that can screen many compounds while accounting for the relevant growth characteristics of bacteria during infection.

Screening for a bactericidal effect irrespective of growth environment may be precise but is grossly inaccurate because the goal should be to treat pathogens where they are pathogenic.

Certainly, research to discern antimicrobial mechanism(s) of action on specific bacteria has led to many important discoveries. However, the nature of earlier work is slow and tedious, often only describing mechanisms of novel compounds years after they might first be considered novel. Such a low throughput strategy can account for the importance of specific growth conditions to pathogenesis, but is not efficient to screen larger libraries of new potential antimicrobials. Thus, there is a clear need for a better method and strategies to discover antimicrobials that overcome the shortcomings of current methods.

Development of a systematic strategy to examine and quantify the susceptibility of pathogenic bacteria to new antimicrobial compounds under relevant growth conditions using microfluidics and correlated imaging is necessary. A microfluidic platform that controls nutrient gradients to examine bacterial behavior in response to potential antimicrobial compounds and chemical imaging strategies that can provide better and faster metrics for cell stress and death would advance screening technologies. Our strategy utilizes a "precision throughput" approach that can screen many compounds, but not at the expense of appreciating the importance of bacterial growth phenotypes that mediate pathogenesis. Thus, our approach improves efficiency of drug discovery by minimizing false positives and false negatives resulting from indiscriminately growing bacteria in a general manner and assessments that rely on antiquated metrics of bactericidal activity.

It is well-established that nutrient availability affects pathogen colonization, growth and virulence, but little is known about variation of bactericidal action as a function of nutrient conditions. This disclosure can use a microfluidics platform to create a vast number of well-defined growth environments, which can be used to research bacterial behavior and bactericidal action in great detail. While the literature provides certain examples of microfluidic devices, the true potential of microfluidic technology has not been realized. We disclose a platform that can be used to fill critical technical gaps by: A) screening for antimicrobials under more relevant growth conditions for pathogenic bacteria; B) identifying physiological conditions of bacteria growth that influence antimicrobial susceptibility and resistance; and C) defining imaging and chemical imaging profiles of antimicrobial susceptibility and pathogen behavior.

Our "precision throughput" approach, can be used to screen the Notre Dame Chemical Compound Collection maintained by the Notre Dame Warren Family Research Center for Drug Discovery and Development against the bacterial pathogens Pseudomonas aeruginosa and Group A Streptococcus pyogenes. Because no pathogenic bacterium has ever been studied in the manner we propose, our research provides not only new methodology but also enables discovery of how pathogenic bacteria survive under differing relevant conditions and how these pathogens might be disrupted with specific novel antimicrobials.

Approach

New antimicrobials are needed to address current and future healthcare needs. In addition to increased awareness and reports of antibiotic resistance, there is a clear need to treat existing pathogenic bacteria in a more targeted manner to improve outcomes and limit side effects. Initiatives such as the NIH NCATS New Therapeutic Uses provide for the opportunity to

test existing compounds for bactericidal effects.

Additionally, many researchers use a variety of strategies to derive new compounds. We assert there is a bottleneck in assessment. This disclosure provides advances to how potential antimicrobials are screened and how antimicrobial activity is assessed. Existing strategies and metrics are antiquated and inefficient. We disclose a strategy that can screen many compounds while accounting for the relevant growth characteristics and stress responses exhibited by bacteria during infection is paramount to success. Efforts from bench to clinic can be improved by employing a precision strategy to treat pathogens where they are pathogenic. The disclosure herein details a multi-faceted approach that can be implemented to transform and revolutionize how the pharmaceutical industry screens antimicrobial compounds. Current methods in antimicrobial screening

Numerous approaches are used currently to screen new compounds for antimicrobial activity. One common approach uses a microtiter-based assay to test some measure of bactericidal activity of a compound library in a 96-well plate format. The simplest outputs measure optical density where low values indicate growth inhibition or bactericidal activity. While it is generally understood that this approach can identify few potential candidates amongst many comparatively inert compounds, the limitations of this common strategy can be severe. Well-reasoned and meticulous screens can still fail to yield promising candidate compounds. As detailed by Payne et al. (Nat Rev Drug Discov., 2007, 6, 29), when using a high throughput approach with the intent of identifying compounds that could target specific proteins or actions of pathogenic bacteria, most investigation yielded no viable leads and several initial hits for activity were subsequently discarded when investigated in greater detail.

Alternative approaches used to increase sensitivity and specificity of these assays utilize specific reporter- based methods in live-culture assays; for example, luciferase bioluminescence of lactate dehydrogenase promoter-lux system in a modified Streptococcus mutans background decreases drastically if the bacteria are not respiring. Even more specific genomics-based in vivo and in silico assays can be used to measure the properties of the antimicrobial candidates with these biomolecules nearly independent of bacterial growth. However, these more specific and more sensitive approaches have yielded very few new candidate compounds.

The new antimicrobial teixobactin was discovered using yet another approach that enabled novel compound discovery. Nutrients and growth factors were diffused through a microfluidic device inserted into a soil environment to promote growth of uncultured bacteria in their natural environment. By not relying upon a microbial isolation strategy to grow antimicrobial-producing organisms in pure culture, this approach side-stepped convention to obtain new compounds for screening.

We note that these previous strategies, however, do not focus on finding

antimicrobials that show activity on bacteria under conditions important to pathogenesis.

Research to detail the specific mechanism and efficacy of antimicrobial action upon a pathogenic bacterium often utilizes a protocol that investigates one (or a very few) antimicrobial molecule at a time. Such studies are critical to determine mechanisms of action for antimicrobials and are equally important as an initial step in screening overall suitability of novel compounds for drug development. Much of that research specifically considers growth of target bacteria and conditions pertaining to pathogenesis via media or animal model examination, however, the detailed nature of that work precludes scrutiny of many compounds. Pathogenic bacteria basics

This disclosure includes the implementation and development of a strategy to probe for bactericidal activity using two pathogenic bacteria as model systems: Pseudomonas aeruginosa and Streptococcus pyogenes. Both P. aeruginosa and S. pyogenes are responsible for numerous acute and chronic infections of the lung, skin, eye, nasopharynx, intestine, and bone.

Pseudomonas aeruginosa is an opportunistic pathogen responsible for both acute and persistent infections. While this ubiquitous environmental bacterium rarely infects healthy adults, it is a common pathogen among susceptible populations, such as individuals with cystic fibrosis (CF), burn victims, ventilator patients, and those who have had intestinal reconstruction. P. aeruginosa is among the most common nosocomial pathogens for intensive care unit patients.

In many of these clinical situations, including the lungs of individuals with CF, P. aeruginosa is the dominant organism and exists predominantly in biofilms, in which its pathogenicity and resistance to antibiotics are significantly upregulated. For P. aeruginosa and many bacterial pathogens, the most persistent of these infections, and the hardest to treat with antibiotics, are those that form bacterial biofilms. Biofilms are surface-associated, socially organized communities of cells.

Streptococcus pyogenes (Group A Streptococcus; GAS) is a clinically-relevant Gram- positive bacterial pathogen that causes an array of diseases, ranging from simple pharyngitis and impetigo to life-threatening toxic shock syndrome (TSS) and necrotizing fasciitis.

Approximately 700 million cases of GAS infection occur annually worldwide, with approximately 18 million of these cases considered severe. In addition, -700,000 cases of sepsis occur in North America each year, with a mortality rate of -30-50% in patients afflicted by its more severe forms. A resurgence of GAS infections in developed nations and in developing countries where severe GAS-related diseases are often endemic, has placed this bacterial pathogen as a serious disease concern associated with morbidity and mortality. The persistence of GAS infections may be linked to the emergence of new serotypes and pathovars of the bacteria with altered abilities to interact and thrive within diverse host environments. Nutrient variation impacts bacterial behavior, pathogenesis, and antibiotic activity

It is widely understood that bacterial growth and pathogenesis are greatly influenced by the availability of nutrients. Yet, this knowledge is often conveniently applied selectively. Given that is well-established how nutrient availability affects pathogen growth and nutrient availability affects pathogen virulence, it is unfortunate that more attention not been given to bactericidal action in consideration of nutrient conditions. Common practice tolerates use of poorly- chemically defined media for bacterial growth— these nutrient rich mixtures allow for rapid and high- yield growth but rarely represent the growth conditions of any specific host environment. The following discussion includes just some examples of nutrients that can influence bacterial growth.

Many controlled in vitro studies have shown that basic phenotypes of P. aeruginosa are influenced by the constituents available in the growth environment. Figure 1 shows two types of representative examples where simple substitutions between carbon compounds (succinate, glutamate, or glucose) and/or nitrogen (NH4⁺-ammonium vs. NC ^'-nitrate) affect the biofilm formation or swarm motility phenotype of P. aeruginosa growing in pure culture, controlled, laboratory assay experiments. Production of P. aeruginosa rhamnolipid, an agent with surfactant and cytotoxic properties is established to increase when cultures are growing with excess glucose and nitrate as sources of carbon and nitrogen, respectively.

Aspects of carbohydrate utilization upon the behavior of GAS have also been established. Switching between organic acids and simple sugars affects growth and gene expression, and specifically, production of streptolysin S. In addition, transcriptome analysis of GAS has linked carbohydrate utilization and other metabolic pathways to distinct stages of bacterial community growth and virulence gene production.

The influence of nutrient composition and availability also extends directly to antibiotic activity against bacteria. For example, for the aminoglycoside gentamicin, it was noted that increases in magnesium in the growth medium correlated with decreased inhibition of P. aeruginosa. Oxygen availability has also been linked directly with antibiotic tolerance where P. aeruginosa shows increased resistance to tobramycin, ciprofloxacin, carbenicillin, ceftazidime, chloramphenicol, or tetracycline when oxygen is limited.

Amino acids

The importance of amino acids to virulence of P. aeruginosa and other bacteria comes from several lines of evidence. Amino acid auxotrophs have been recovered from infections for several pathogens: Staphlyococcus aureus, Listeria monocytogenes, Burkholderia cepacia, Streptococcus suis, and Salmonella typhi. When investigating the biology of upper airway infections for CF patients, P. aeruginosa amino acid auxotrophs are routinely isolated from lung sputum samples. Amino acids are a major source of carbon (and nitrogen) to support growth of P. aeruginosa in CF sputum. Proline, alanine, arginine, lactate, glutamate, and aspartate (i.e., five amino acids and one hydroxyl acid) have been identified as the preferred carbon sources in CF sputum. Discerning how antibiotics affect bacteria when growing on select amino acids can greatly improve our understanding of pathogenesis.

Certainly numerous studies have linked amino acid responses by bacteria to alternate behavior and disease. A few examples include: Changes to the gut microbiome population dependent upon carbohydrate and amino acid availability were recently linked to the malnutrition disease Kwashiorkor. Proline-dependent responses have been linked to

Campylobacter jejuni and Clostridium difficile pathogenesis. Lastly, the urinary pathogen Proteus mirabilis was recently shown to require glutamine for swarming.

For P. aeruginosa, the amino acid-dependent responses are less characterized than for E. coli and some other bacteria but several responses are known. P. aeruginosa shows increased resistance to Polymixin B and Colistin when growing on glutamate. More direct links with virulence are also established as amino acid- dependent phenotypes have been observed in addition to the isolation of auxotrophic mutants from CF sputum. Production of virulence protein Exotoxin A is enhanced by glutamate, pyocyanin production is increased by aromatic amino acids, and swarming is improved by small additions of amino acids.

Less is known about the importance of amino acids to GAS infection, but it is established that amino acids can play a role. There is some evidence to suggest that arginine availability is required for GAS to effectively evade nitric oxide stress (iNOS) displayed by host innate immune mechanisms. In addition, catabolism of the alpha-amino acid citrulline has been demonstrated to protect GAS in low-pH environments during infection.

Iron

Iron (Fe) is an essential and often growth-limiting element for pathogenic bacteria. In addition to growth, it is well established that available iron influences production of many pathogenic virulence factors.

Iron metabolism has been studied in great detail for P. aeruginosa, particularly via genetic and transcriptional methods. Iron uptake is known to be regulated at many levels and in response to several variables - for example, cellular iron sufficiency, the availability of a particular iron source, the pH and C tension. Pivotal to iron uptake by P. aeruginosa are siderophores: small molecular weight Fe ⁺ chelators that bacteria secrete to both mobilize Fe from wherever it is bound and transport it (via specific cell surface receptors) through the outer membrane. Siderophores are abundantly overproduced by bacteria under iron stress, accrue in proportion to the size of the bacterial colony, and can themselves act as protein- associated signaling molecules, inducing for example the production of virulence factors as well as their own cell surface receptors. P. aeruginosa makes two siderophores: pyoverdine, a high-Fe ⁺-affinity fluorescent molecule, and lower-affinity pyochelin. Siderophores and iron have been shown to have broad roles for biofilm formation in both clinical and laboratory studies, particularly in the production of highly structured "mushroom stalk and cap" biofilms. The absence of siderophores, effected by knocking out both siderophore-producing systems, results in flat, unstructured biofilms. This phenotype can be reversed by supplying sufficient iron in another actively transported, exogenous chelator. Further work using cocultures of siderophore-producers and siderophore-auxotrophs showed that the two assumed distinct locations in the mature biofilm: producers in the stalks, and auxotrophs in the caps. This suggests the development of specialized metabolisms at different parts of the mature biofilm. These components of the P. aeruginosa iron metabolic pathway serve as variables that we can manipulate in straightforward ways in order to understand their contribution towards host colonization, and possibly, towards antibiotic susceptibility.

GAS infections are also strongly linked with iron, as the hallmark characteristic of Group A Strep is β- hemolysis of red blood cells, which leads to an increase in available iron. Especially relevant in severe invasive diseases mediated by GAS, hemolysis of red blood cells and acquisition of iron sources are designed to meet the needs of rapidly growing GAS in blood systems for dissemination and sepsis outcomes. Recent evidence has highlighted the increased role of iron acquisition in highly virulent and invasive patient isolates of GAS. The streptococcal hemoprotein receptor (Shr) is a surface-localized GAS protein that binds heme- containing proteins and extracellular matrix components. The critical role of Shr in the pathogenesis of the highly virulent M1T1 strain of GAS has been determined by

demonstrating that the Shr mutant exhibited a growth defect in iron-restricted medium, and also attenuated for virulence in in vivo models of skin and systemic infection. Therefore, Shr- mediated iron uptake has been shown to be critical to GAS growth in human blood systems, and is required for full virulence of serotype M1T1 GAS in models of invasive disease.

Recent transcriptome profiling has also demonstrated that highly invasive GAS isolates increase iron uptake and processing systems compared to their non-invasive GAS

counterparts. These data suggest that manipulation of iron sources can significantly influence GAS bacterial behavior, and allow us to precisely test antibiotic susceptibility and other parameters in an iron-rich environment such as that would occur in the context of an invasive, disseminated GAS infection. Comparative measures of antibiotic efficacy in environments that simulate non-invasive vs. invasive GAS disease conditions have not been studied.

Glucose

Utilization of glucose and similar hexose sugars via glycolysis (Embden-Meyerhof-

Parnas Pathway) or Entner-Doudoroff, or conversely, an alternative carbon source funneled through an alternative pathway can have a profound influence on bacterial behavior. For example, strains of Legionella pneumophila that are deficient for glucose-6-phosphate dehydrogenase (required for glucose utilization) were severely outcompeted by other L. pneumophila strains. Investigation of multiple Salmonella typhi strains has shown that many are unable to utilize glucose while other strains exhibit robust growth.

The influences of glucose are marginally understood for P. aeruginosa. It is known that media with glucose promotes production of the surfactant and virulence factor rhamnolipi. Additionally, it is known that biofilm and motility phenotypes are often quite different in glucose- minimal medium compared to growth on organic acids. Although the distinct role for glucose in GAS pathogenesis has yet to be determined, it has been shown that maltodextrin utilization by GAS plays a critical role in the ability of the bacterium to successfully colonize the oropharynx. Given that pathogens like P. aeruginosa, GAS, Staphylococcus aureus, and others can infect multiple host cell types with distinct nutrient microenvironments, it would be useful to better understand the links between overall nutrient composition and pathogenesis.

Host-derived composite media

Several studies have been conducted detailing the growth of bacteria and pathogenic bacteria on various media that better represents the host environment. Among the most detailed approaches is the creation of synthetic CF sputum (SCFM) based upon a chemical analysis of several CF sputum samples. Other approaches have been more modest but also highly informative. Growth of bacteria in environmental conditions related to the skin and mouth have been directly assessed using dilutions of human sweat and human saliva as growth media. This host-medium has even been utilized specifically with a microfluidic approach to gain insight into bacterial behavior, but specific factors or nutrients in saliva that influence oral disease are less clear.

The ability to examine growth of pathogenic bacteria under nutrient conditions that specifically pertain to host colonization and the onset of infection is needed. The description herein addresses that need by developing and modifying bacterial growth media for in vitro experiments that are amenable to examining the importance of specific gradients and nutrient concentrations to bacterial phenotype and antimicrobial activity. The described methods can be applied to specifically examine amino acids, iron, glucose, and host-derived composite media as nutrient factors that influence antimicrobial activity upon P. aeruginosa and GAS.

Microfluidic Basics

Microfluidic devices have a number of characteristics that make them particularly well suited for cell culture systems, including the ability to deliver solutions in a controlled and reproducible manner, the ability to form spatial concentration gradients, use small volumes to prevent unnatural dilution and control the concentrations of both liquid and gaseous reagents, and perform massive parallelism to test many conditions at once. An example of a common microfluidic mixing device is shown in Figure 2. First, the small critical dimensions of microfluidic devices stabilize pressure driven flow, providing low Reynold's numbers for the conditions used in these systems. Furthermore, the flow patterns can be characterized in these systems using particle tracking, tracking of dye fronts, and particle imaging velocimetry resulting in flows that are both well characterized and reproducible. Second, with low Reynold's numbers and laminar flow conditions provides a great opportunity to form spatial gradients, enabling rapid testing of reagents over a broad range of concentrations. Third, the ability to use small solution volumes allows for better mimicking of in vivo conditions by keeping the solution volumes at natural levels and preventing dilution of extracellular signaling molecules. Fourth, control of both solution phase and gaseous species can be easily controlled. In some internal environments, such as the liver, oxygen concentrations are about 9%, which is greatly reduced below ambient concentrations. Fifth, the small size of the fluidic chambers and fluid delivery structures enables massive parallelism, making it possible to increase throughput for detailed assays well beyond what is achievable with current systems.

The ability to examine specific growth conditions to efficiently probe for bactericidal and inhibitory actions of new antibiotics is needed. This need can be addressed by testing bacterial growth and antibiotic susceptibility under controlled variation to environmental conditions using a microfluidic format.

Imaging and Chemical Imaging Basics

Optical Imaging

In laboratory research, the primary methods to rapidly assess bacterial behavior involve microscopy. Implementing an optical light or fluorescence microscopy with fluorescent proteins, dyes, and probes often provides for non-destructive analysis such that the same sample can be imaged over time to research time-dependent phenomena, system dynamics, and fate. Certain research laboratories are experienced in utilizing light and fluorescence microscopy methods to examine bacterial behavior at both single-cell and community scales. Figure 3 shows representative examples of microscopy images obtained by the Shrout and Lee laboratories to probe for green-fluorescent bacterial cells or mixtures of cells expressing different fluorescent proteins; imaging of localized fluorescent protein markers within bacteria; co-imaging of fluorescent cells that have been stained with exogenous dyes to demonstrate the heterogeneous nature of individual cells within a population; and release of nuclear HMGB1 from keratinocytes during GAS infection.

Chemical Imaging

It is not currently feasible to use standard detailed nucleic acid characterization techniques that can be used to detail gene and translational expression patterns in situ.

Relatedly, if a microscopy technique involves diffusion and uptake of a chemical dye, the dynamic behavior of bacteria in an experiment may not be captured in phase. Thus, faster answers are much desired. Vibrational spectroscopy is one altemative with great promise for rapid and accurate identification of bacteria. When coupled to multivariate statistical tools, the functional group specificity of vibrational spectra combined with the inherent variation in morphology and protein expression can be used to identify changes in bacterial behavior, even in cases involving multiple species. The following discussion concerns just some examples of useful chemical imaging information that can be applied to map biological activity.

Vibrational Raman scattering provides non-invasive, label-free information on sample functional groups that is better suited to biological samples than IR imaging. Many cellular constituents are Raman-active, producing vibrational bands at characteristic "fingerprint" frequencies to identify cellular components. In addition, Raman signals are unaffected by water, which scatters weakly, adding to the suitability of Raman spectroscopy to studies of bacteria and their biofilms, Raman spectroscopy has been utilized for characterization and identification of bacterial species, characterization of biofilms, structural analysis of cellular components, and probing medically relevant bacterial species.

Raman scattering can also be performed in a confocal microscopy format which affords sub-μιη three- dimensional spatial resolution, and in this format, it has been utilized for single cell mapping and analysis. Furthermore, its chief drawback, sensitivity, has been addressed through advances in spectral instrumentation, so that it is now routine to acquire an entire spectrum from a sub-μιη spot in a few milliseconds.

The ability to detail the impact of antibiotics on single cells over time is needed. This need can be addressed by obtaining chemical functional group information from confocal Raman microscopy (CRM) in our experiments. The in-situ characterization of functional microbial communities presents a number of challenges that can be effectively addressed using the information available in Raman experiments. For example, CRM imaging has been used to study biofilms of S. epidermidis and ammonia oxidizing bacteria directly, without labeling. Fourier transform Raman scattering has been applied to the identification of physiologically relevant compounds from Antarctic cryptoendolithic microbial communities, and under favorable conditions even closely related species can be distinguished, through the application of chemometric partem recognition techniques, such as principal component, linear discriminant and hierarchical cluster analyses. In a suite of techniques inspired by fluorescence in situ hybridization (FISH), microbial samples are grown in ¹ C/¹⁵N-enriched media, which incorporates, for example, into phenylalanine to produce spectra that are easily distinguishable from proteins displaying native amino acids. Interestingly, the same perturbations can also be read out by SIMS, rendering the MS-Raman pair uniquely well- suited to study single, labeled microbial cells in complex microbial communities.

The Bohn research group in collaboration with the Shrout research group have demonstrated the applicability of CRM imaging by characterizing the transition from planktonic cells to biofilms in P. aeruginosa grown under static conditions. Raman spectra from planktonic cells, see Figure 4(a), are composed of signals from molecular vibrations of individual cell components - lipids, proteins, nucleic acids and carbohydrates, while biofilms tend to be dominated by excreted components, both polymeric EPS as well as small signaling molecules and glycolipids. In our CRM studies, Raman spectra of P. aeruginosa wild-type and QS- mutant planktonic cells were acquired between 100 cm^"1 and 3600 cm^"1 and subsequently compared to biofilms grown from the same original cells. Spectra were analyzed from 600 - 1800 cm^"1; Raman bands in the range 2750 - 3050 cm^"1 are dominated by C-H stretching vibrations, and can carry important information about cell membrane fluidity.

Even a cursory visual inspection of the spectra in Figure 4(a) reveals the differences between planktonic cell and biofilm states for P. aeruginosa. The Raman spectrum from a planktonic cell reveals a series of DNA/RNA base vibrations that are well defined, the strongest of these being the characteristic thymine out-of- plane C-0 bend at -747 cm^"1, along with the guanine and adenine ring breathing vibrations at 1585 cm^"1, an adenine ring vibration at 1310 cm^"1 among others. Upon biofilm formation, dramatic changes occur in the spectrum. Biofilms can be many μηι in thickness, well beyond the confocal depth of CRM, making it possible to move the confocal plane in the biofilm away from the substrate, thus greatly reducing the magnitude of the substrate background. Thus, it is not surprising to lose scattering from the substrate (Si/SiC ), but the DNA/RNA-related bands at 747 (thymine), 1126 (cytosine), 1310 (adenine), and 1447 cm^"1 (cytosine) as well as the strong band at 1585 cm^"1 are diminished or disappear altogether, while a new band at 1601 cm^"1 grows in. These observations are consistent with the growth of a relatively thick biofilm which dilutes the contribution of nucleotide-derived peaks, which must come from the cells themselves, in the spectrum. The band at 1030 cm^"1 is of particular note and is assigned to a glycolipid, in this spectrum the C-0 stretching vibration originating from the rhamnose sugar of the secreted rhamnolipid. This peak is not observed in the spectrum from planktonic cells. P. aeruginosa and other Pseudomonas species are known to excrete rhamnolipids, in concert with biofilm formation, so the observation of this band is a marker of the planktonic cell-to-biofilm transition. More surprising is the appearance of the sharp peak at 999 cm^"1, which is not observed in the planktonic cell spectrum. This band, assigned to phenylalanine, is characteristic of protein, as is a weaker band observed at 617 cm^"1. Phenylalanine and tyrosine are the two amino acids that contribute predominantly to vibrational spectra from most cells, with tyrosine (not observed) being assigned to peaks at 849 cm^"1 and 1617 cm^"1. Thus, in addition to the expected rhamnolipid signal, the observation of these bands suggests the upregulation of protein expression.

For comparison, representative spectra for a QS-mutant (AlasArhll), a strain deficient for rhamnolipid production (QS is required for rhamnolipid production) show dramatic differences between planktonic cell and biofilm spectra (not shown). In contrast to the wild- type, DNA/RNA bands at 747, 1126, 1310, 1447, and 1585 cm^"1 in the planktonic cell spectrum are also observed in the 72h QS-mutant "biofilm" spectrum, clearly pointing to hindered biofilm development in the QS mutant, as does the concordance of peak positions between the planktonic and "biofilm" spectra.

Recent Raman imaging experiments show striking differences between planktonic cells and biofilms. The image in Figure 5(a), assembled from vibrational bands largely assigned to C-H stretches (2800-3050 cm^"1), shows several P. aeruginosa cells in close proximity. Aside from the obvious difference in size scale, the biofilm, Figure 5(b), exhibits a much more heterogeneous internal structure. It is particularly noteworthy that, consistent with the spectral differences in Figure 4, the image of the biofilm is dominated by bands at 1000- 1050 cm^"1 (rhamnolipid, blue) as well as protein-derived vibrational bands (red).

Examination of media

The constituents of published media for various infection models can be modified for testing in our microfluidic format. We can create controlled micro-environments of select nutrients such that stepwise concentrations within a larger gradient are examined in the same controlled in vitro assay experiment.

Minimal media with specifically controlled and variant nutrients.

Numerous minimal media have been derived and investigated for probe for growth of P. aeruginosa, GAS, and many other pathogens. We can utilize published medium

compositions to further develop and test our methodology to integrate the microfluidic platform to grow cultures of P. aeruginosa and GAS and image these bacteria over time using optical and chemical imaging. The most complex "host-inspired" minimal medium (i.e., fully-defined chemical salts) that currently exists is synthetic cystic fibrosis sputum medium (SCFM) developed by the Whiteley Laboratory at UT-Austin. This can be used to explore antimicrobial action in a simulated lung infection environment.

Complex host-inspired media.

The SCFM medium described above is very unique in that research has already been completed to characterize CF sputum chemistry. As we are interested in probing for bacterial pathogen growth relevant to many infections the disclosed method is not limited to published minimal medium compositions to conduct research. Other research to specifically investigate bacterial growth and colonization within the context of specific host environments has utilized a more direct approach of sterilizing, filtering, fractioning, and/or diluting actual mammalian fluids as described above. In addition, a similar approach in tandem with experiments conducted using minimal medium to probe for growth of P. aeruginosa and GAS in less-defined host-inspired medium compositions can be utilized.

Development of a Microfluidic Platform

Microfluidic device design

The microfluidic systems can be made using standard methods using PDMS substrates. These devices can be fabricated with 3D channel structures and thin PDMS layers, and additionally these devices can be fabricated from glass. The devices can use pressure driven (hydrodynamic) flow to insure reliable delivery of solutions and reproducible flow rates. The devices can be modeled using COMSOL Multiphysics to ensure the more complex systems are maintaining equivalent flows and nutrient delivery. The channel structure, solution delivery system, and laminar (non-turbulent) flow ensures consistent delivery of solutions across the cell culture chamber. An excellent exploratory size for the cell culture chambers has the dimensions 500 μηι X 500 μηι X 40 μηι providing a total volume of 10 nL. To reduce the dilution of signaling compounds, the culture media flow rate can be configured to 10 nL/hour as measured by particle tracking. This chamber design is large enough to avoid most clogging problems, while still small enough to provide efficient imaging and efficient use of spacing for further multiplexing. A liquid handling manifold can be created, wherein flows controlled with a program written in LabView that automatically record the solution delivery parameters and correlate with other data streams. The manifold would connect directly with the microfluidic chip. The microfluidic chips can be disposable, single-use devices. Because PDMS is highly permeable to oxygen, the oxygen concentration can be controlled by placing the device in a chamber with a controlled gas composition, making it possible to supply oxygen without liquid flow.

On-chip gradient formation

The microfluidic devices can allow rapid and reproducible testing of numerous experimental conditions to test moderate numbers of compounds over a vast number of experimental conditions. This work can utilize the ability of microfluidic chips to make spatial concentration gradients over an area that correspond to the bacterial growth chamber. Figure 6 illustrates two microfluidic schematics of increasing complexity. Development of more complex designs are possible. Gradients can be formed a single species or multiple species to investigate interactions and synergistic effects. On-chip formation of gradients can provide the ability to test the effects of concentration of single compounds in the media on simple devices, and multiple species in larger systems. While continuous gradients are possible in microfluidic systems, the high degree of cell motility for many bacteria makes a design in which the bacteria cells are loaded into discrete chambers that can be exposed to different concentration of nutrients and antimicrobials.

Multi-component growth environments enabled by increased chip complexity

The chip designs can use a fluidic structure to introduce the media and expose the a gradient in a single variable. Other designed systems can use a more complex microfluidic design to test two or more variables in an N x N matrix format where N is the number of discrete concentrations tested for each of the variables. This design can provide significant gains by allowing the investigation of synergistic effects. Additionally, these more complex N x N matrix designs can screen more than one antimicrobial candidate compound on the same chip.

The Notre Dame Chemical Compound Collection maintained by the Notre Dame Warren Family Research Center for Drug Discovery and Development can be screened by adding nL volumes from the existing the 96- well plate format of the collection using the microfluidic format. Thus, availability of small amounts of compound will not limit a research effort, whatsoever. Each compound can initially be screened separately; however, our platform detailed in Figure 6b can allow for simultaneous introduction of multiple compounds.

Inoculation and growth of P. aeruginosa and GAS in microfluidic chambers

Inoculation of bacteria into the culture chambers represents a critical aspect of the device formation. Therefore, a simpler method was chosen initially which can evolve to more elaborate schemes. Our approach introduces cells in filtered suspensions through a microfluidic channel that is separate from the microchannels used for delivery of the media and bioactive agent. This design can avoid fully exposing the gradient formation system to the cell suspension to reduce contamination and to reduce the potential for clogging. The channel exiting the cell culture chamber can be optimized in size to have low backpressure to prevent contaminating the gradient dilution channels with cells and to prevent contamination of the culture chamber through the waste channel. The flow can be stopped for a period of time to allow cell adhesion, and only small volumes of media must be passed purge the microchannels in which the cells to not belong. If necessary, membranes, microfabricated pillar filters, and microvalves can be added to improve the performance. Loading of the same volume can provide approximately loading of the same numbers of cells with some variance. The variance of the loading can be studied and to determine the effects of the variance on the reproducibility of the cellular assays. Multiple replicates of the same assay can provide reproducible results by averaging out the variance in the number of cells loaded in each chamber. It is recognized GAS cells are known to aggregate more than P. aeruginosa cells, but both can be loadable by suspension with the appropriate pre-filtering, channel dimensions, bacterial inocula.

Inoculated microfluidic chambers can be maintained for 24-48 hours to assess antimicrobial activity. Once bacteria are inoculated, the antimicrobial compound to be tested into the growth chambers can then be introduce. Bacterial behavior and antimicrobial activity can be routinely monitored using optical imaging and some experiments can be performed in parallel for chemical imaging. Some experiments can be conducted entirely on a microscopy stage to detail long term effects, while others can be examined periodically to increase the conditions investigated. Repetition of some assays can be conducted simply for quality control purposes in addition to confirmatory investigation of specific research discoveries that apply statistical significance metrics.

Imaging and Quantification

Optical Imaging

The Shrout and Lee Laboratories have utilized optical microscopy and imaging methods to examine bacterial behavior at both single-cell and community scales. P.

aeruginosa and GAS cultures can be inoculated in microfluidic growth chambers and imaged over time. Given the nature of the microfluidic platform, it is not a precondition that cultured bacteria develop as surface-attached "biofilms". Experiments can utilize GFP or mCherry fluorescent protein-tagged cells (only produced by living cells) as a surrogate for cell survival while cell death can be assessed initially using Live/Dead staining to quantify the ratios of live to dead cells remaining after exposure to these antimicrobials as a metric for activity.

Such assessment of "living" and "dead" bacteria represent a standard approach for antimicrobial action against bacteria. Thus, these metrics are useful starting points for research. It is important to note that bacterial cell death is not a requirement, however, to treat infection. Thus, other metrics of bacterial behavior in response to the chemical library such as cell morphology, motility, and biofilm development can be examined. Further, it is possible to implement additional strategies to assess antimicrobial activity through the iterations applied research.

Chemical Imaging

Raman spectroscopy chemical imaging can be used to discern differences in known biomolecules. In addition, it is possible to probe for and identify new biomolecular signatures that correspond to bacterial stress and antimicrobial action. An example of one possible signature is presented in Figure 7, which details a decrease in the carotenoid synthetic pathway of the bacterial soil isolate Pantoea strain YR343 strain and special mutant with the carotenoid synthetic pathway disrupted (and therefore lacking the strongly colored pigments that give rise to the intense resonantly enhanced Raman bands). Clearly below 0.1 mM (10-4 M) H2O2 the cells are "alive" and above 1 M they are "dead", but the Raman allows for a continually graded measure of cellular disruption with the greatest change observed when the dose increases to 1 mM H2O2. Normalization is done against the integrated CH stretching bands (not shown) which are indicative of total organic content and do not change on H2O2 exposure.

It is also possible to probe behavior of known pathogenic virulence factors that can be identified and quantified using this chemical imaging technique. For example, we can detail differences in siderophore production by P. aeruginosa as this is a known virulence factor for which the importance of iron was detailed above.

This chemical imaging approach is critical to develop new markers to assess pathogen inhibition. Thus, we can assess potential metrics for antimicrobial behavior in a highly iterative fashion to probe, discern, and validate biomolecular signatures that may indicate a desirable antimicrobial response that can prevent pathogen colonization of the host and or expression of virulence factors that actually cause disease. Imaging Data Quantification

The Shrout laboratory has considerable experience in assessing optical and fluorescence data from different mixtures of protein-expressing and dye fluors. Fluorescence emission profiles visualized from light microscopy experiments can be quantified with spatial analysis software (e.g., COMSTAT).

Similarly, the Bohn lab has experience in assessing and characterizing chemical imaging data from Raman spectroscopy and other techniques. Chemometrics - essentially "the entire process whereby data, i.e. numbers in a table, are transformed into information used for decision making," - is a powerful and essential part of chemical imaging, since each image pixel is composed of an entire Raman spectrum. For research, it is important to define how different chemometric tools can be used to elucidate subtle relationships hidden in the complex spectral datasets. Unsupervised learning techniques, such as principal component analysis (PCA) and hierarchal cluster analysis (HCA), have been shown to work well for initial analysis of chemical images, as they can be implemented to identify data clustering without a priori conditions, a valuable feature in Raman spectroscopy. PCA is particularly powerful in identifying the subtle variations in spectral response that accompany spatial and temporal differences in chemical messenger secretion. We have observed striking

spatiotemporal variations of Pseudomonas quinolones (in the quinolone stretching vibration in the region 1340-1390 cm^"1) in Pseudomonas biofilms by mapping the higher order principal components, variations which are masked in the more common approach of mapping Raman spectral intensities. HCA is an important complement to PCA, as it examines abstract inter-point distances between samples and represents the information in 2D dendrograms, so that clusters of data can be identified by eye. The dendrograms are created through an iterative process of sample- specific cluster joining and identify those data subsets that are most alike, i.e. are clustered. PCA has become a staple in Raman spectroscopy and imaging generally, while HCA has had substantial impact in biological Raman imaging.

The following Examples are intended to illustrate the above invention and should not be construed as to narrow its scope. One skilled in the art can readily recognize that the Examples suggest many other ways in which the invention could be practiced. It should be understood that numerous variations and modifications may be made while remaining within the scope of the invention.

EXAMPLES

Example 1. Importance of Nutrient Composition upon Bacterial growth, survival, and exposure to traditional antibiotics

Numerous protocols are used currently to screen new compounds for antimicrobial activity. One common approach uses a microtiter-based assay to test some measure of bactericidal activity of a compound library in a 96-well plate format. The simplest outputs measure optical density where low values indicate growth inhibition or bactericidal activity. While it is generally understood that this approach can identify few potential candidates amongst many comparatively inert compounds, the limitations of this common strategy can be severe. Well-reasoned and meticulous screens can still fail to yield promising candidate compounds. As detailed by Payne et al. (Nat Rev Drug Discov., 2007, 6, 29), when using a high throughput approach with the intent of identifying compounds that could target specific proteins or actions of pathogenic bacteria, most investigation yielded no viable leads and several initial hits for activity were subsequently discarded when investigated in greater detail.

We demonstrate the importance of the nutrient composition importance and the importance of discerning specific chemical signals of our TAPS strategy while probing for bactericidal activity using two pathogenic bacteria as model systems: Pseudomonas aeruginosa and Streptococcus pyogenes (Group A Streptococcus; GAS). Both P. aeruginosa and GAS are responsible for numerous acute and chronic infections of the lung, skin, eye, nasopharynx, intestine, and bone.

It is widely understood that bacterial growth and pathogenesis are greatly influenced by the availability of nutrients. Yet, this knowledge is often conveniently applied selectively. Common practice of "high throughput screening" (HTS) tolerates use of poorly-chemically defined media for bacterial growth— these nutrient rich mixtures allow for rapid and high- yield growth but rarely represent the growth conditions of any specific host environment.

Pseudomonas aeruginosa is an opportunistic pathogen responsible for both acute and persistent infections. Many controlled in vitro studies have shown that basic phenotypes of P. aeruginosa are influenced by the constituents available in the growth environment (for a review of associated phenotypic behaviors see Shrout, et al., MRS Bull. 2011, 36, 367). Production of P. aeruginosa rhamnolipid, an agent with surfactant and cytotoxic properties is well established to increase when cultures are growing with excess glucose and nitrate as sources of carbon and nitrogen, respectively.

Group A Streptococcus pyogenes, GAS, is a clinically-relevant Gram-positive bacterial pathogen that causes an array of diseases, ranging from simple pharyngitis and impetigo to life-threatening toxic shock syndrome (TSS) and necrotizing fasciitis.

Approximately 700 million cases of GAS infection occur annually worldwide, with 18 million of these cases considered severe.

Aspects of carbohydrate utilization upon the behavior of GAS have also been established. Switching between organic acids and simple sugars affects growth and gene expression, and specifically, production of streptolysin S. In addition, transcriptome analysis of GAS has linked carbohydrate utilization and other metabolic pathways to distinct stages of bacterial community growth and virulence gene production.

Tunable Attribute Precision Screening (TAPS)

TAPS, is a microbial assessment platform developed to research and probe chemical activity effects upon microbial cells cultured under conditions that are actually relevant to an environment of interest.

Rationale and Need

Bacteria grow and function differently, even in pure culture, with changes to their chemical growth environment, (for example, see Figure 1).

Ex situ research of microbial function remains a useful approach to discern regulation and mechanism-but it is difficult to mesh "high throughput" with "realistic and relevant".

Considering the example of drug discovery, current strategies used to probe for antibiotic activity are coarse and inefficient to screen for compounds that limit bacterial function under relevant pathogenic conditions. Current high-throughput practices to screen potential antibiotics, for example, do not cell account for any specifics in growth or inhibition. Such approaches inefficiently identify many false positives and false negatives of novel compounds. TAPS Approach

The TAPS platform utilizes microfluidics to assess multiple scenarios by creating a vast number of well-defined growth environments to research cellular behavior and inhibitory action with precise detail. The readout and quantification of inhibitory action via the TAPS platform comes from a combined imaging method. Many possible outcomes besides cell death or rapid growth can signify important effects of chemical and radiation exposure.

Assessment of cellular behavior using TAPS is possible by multiple optical and chemical imaging strategies. See Figure 8 for differences in bacterial biomolecules measured using Confocal Raman Microspectroscopy. The TAPS platform allows for precision control of numerous parameters to test cellular growth and chemical effects in the same experiment.

Example 2. Susceptibility to Traditional Antibiotics varies with nutrient composition

The influence of nutrient composition and availability extends directly to antibiotic activity against bacteria. For example, for the aminoglycoside gentamicin, it was noted that increases in magnesium in the growth medium correlated with decreased inhibition of P. aeruginosa. Oxygen availability has also been linked directly with antibiotic tolerance where P. aeruginosa shows increased resistance to tobramycin, ciprofloxacin, carbenicillin, ceftazidime, chloramphenicol, or tetracycline when oxygen is limited. Data from the Shrout Lab shows that antibiotic susceptibility diverges with other changes to the nutrient environment. We find that the aminoglycoside tobramycin and beta-lactam carbenicillin both show different inhibitory characteristics when growing on rich LB medium versus a minimal medium (FAB) containing glucose. Data in Figure 9A shows how 1.0 μg/mL of the antibiotic tobramycin nearly completely limits growth in FAB-glucose cultures but does not fully limit growth in LB broth. Data in Figure 9B shows that while 100 μg/mL carbenicillin is inhibitory for both growth conditions, cells growing in 50 μg/mL with LB survive well but not in FAB-glucose medium.

GAS susceptibility to antibiotics has traditionally been tested on nutrient rich broth culture or agar and the general potency of antibiotic determined using MIC or disk diffusion assays. Other genetic-based tests as well as patient serum-specific bactericidal screening have been implemented for several bacterial pathogens, however, antibiotic resistance profiles in human strains of GAS have generally not been tested using a variety of media conditions. Data from the Lee lab demonstrates that indeed GAS 5448 strain Ml exhibits varying levels of antibiotic susceptibility in distinct media conditions— kanamycin is not inhibitory to GAS in human cell-growth medium DMEM at 30 μg/mL (Figure 9D) as in Todd-Hewitt or a minimal medium (Figure 9C). Collectively, these results show that inhibitory effects of these differing antibiotics are influenced by the nutrient environment in alternative ways (requiring further exploration).

Amino acids as a specific nutrient

The importance of amino acids to virulence of P. aeruginosa and other bacteria comes from several lines of evidence. Changes to the gut microbiome population dependent upon carbohydrate and amino acid availability were linked to the malnutrition disease

Kwashiorkor. Proline-dependent responses have been linked to Campylobacter jejuni and Clostridium difficile pathogenesis.

Amino acid auxotrophs have been recovered from infections for several pathogens:

Staphlyococcus aureus, Listeria monocytogenes, Burkholderia cepacia, Streptococcus suis, and Salmonella typhi. When investigating the biology of upper airway infections for CF patients, P. aeruginosa amino acid auxotrophs are routinely isolated from lung sputum samples. Amino acids are a major source of carbon (and nitrogen) to support growth of P. aeruginosa in CF sputum. Proline, alanine, arginine, lactate, glutamate, and aspartate (i.e., five amino acids and one hydroxyl acid) have been identified as the preferred carbon sources in CF sputum.

For P. aeruginosa, the amino acid-dependent responses are less characterized than for E. coli and some other bacteria but several responses are known. P. aeruginosa shows increased resistance to Polymixin B and Colistin when growing on glutamate. More direct links with virulence are also established as amino acid-dependent phenotypes have been observed in addition to the isolation of auxotrophic mutants from CF sputum. Production of virulence protein Exotoxin A is enhanced by glutamate, pyocyanin production is increased by aromatic amino acids. The Shrout and Bohn groups have researched P. aeruginosa responses to amino acids during surface growth. As illustrated in Figure 8, we find that pyocyanin production varies with both growth environment and strain. Thus, the biosynthetic pathways utilized at the sub-species level may differ (in addition to nutrient-dependent responses), which is germane to efforts of characterizing P. aeruginosa during infection. Whereas the PAOIC strain is competent to synthesize phenazines (of which pyocyanin is the most abundant member) from glutamate but not glucose, the FRD1 strain synthesizes abundant pyocyanin with both glucose and glutamate.

Conclusions. This disclosure has the potential to disrupt the current paradigm of bactericidal screening by a strategy that can be used to identify antimicrobial compounds that have specific efficacy under pathogenic growth conditions. Antibiotic resistance is one of the most important health concerns today, as resistant strains of pathogenic bacteria are increasingly pervasive in the community. Furthermore, antibiotic discovery of novel compounds has historically been derived from static screening of bactericidal activity under general in vitro culture conditions that do not take into account the heterogeneity of in vivo host environments under which bacterial infections can persist. This disclosure represents a unified, multi-environmental testing system that can serve as a standard for comprehensive testing of antimicrobial compounds.

Each of the components is individually novel, which can further advance: 1) the understanding and characterization of responses of Pseudomonas aeruginosa and Group A Streptococcus to differing growth environments, nutrient limitations, and multi-component factors that have been shown (separately) to be important to pathogenesis; 2) identification of new antimicrobials that have been missed using conventional screening approaches; and 3) the detail of biomolecular signatures exhibited by bacteria under distinct conditions using chemical imaging; these signatures can serve as novel indicator patterns of bacterial stress, inhibition, and killing in response to active antimicrobials. Our approach also provides a uniform platform to rapidly screen multiple growth environments and multiple potential antimicrobials, for which our microfluidic device design strategy can be critical.

While specific embodiments have been described above with reference to the disclosed embodiments and examples, such embodiments are only illustrative and do not limit the scope of the invention. Changes and modifications can be made in accordance with ordinary skill in the art without departing from the invention in its broader aspects as defined in the following claims.

All publications, patents, and patent documents are incorporated by reference herein, as though individually incorporated by reference. No limitations inconsistent with this disclosure are to be understood therefrom. The invention has been described with reference to various specific and preferred embodiments and techniques. However, it should be understood that many variations and modifications may be made while remaining within the spirit and scope of the invention.

Claims

What is claimed is:

1. A method for precision throughput screening of biologically active compounds, the method comprising:

a) mixing a plurality of nutrients in a microfluidic chip, wherein the microfluidic chip comprises:

i) a plurality of microfluidic channels, wherein each microfluidic channel has an inlet for a nutrient that flows hydrodynamically to the microfluidic channel, and wherein the microfluidic channels are communicatively connected with a plurality of dilution channels for mixing of the nutrients; and

ii) an L x M array of cell culture chambers configured to receive the flow of nutrients from one or more microfluidic channels, one or more dilution channels, or a combination thereof, wherein L x M defines the size of the array, and the product of L x M is greater than the sum of inlets; b) forming an L x M array of varying media conditions relevant to growth of a microorganism by a spatial concentration gradient of cell culture growth media in the L x M array of cell culture chambers, wherein a distinct composition of cell culture growth media is formed by a combination of nutrients flowing into each cell culture chamber;

c) inoculating the microorganism in more than one cell culture chamber, thereby constituting a viable population of microorganisms that can grow at one or more distinct compositions of the varied media conditions in the L x M array of cell culture chambers;

d) treating the viable population with a biologically active compound; and e) imaging the viable population by an optical imaging technique or a chemical imaging technique, wherein biological activity of biologically active compounds is assessed by imaging the viable population;

wherein the biological activity of biologically active compounds is precisely screened in an L x M array of varying media conditions relevant to the microorganism.

2. The method of claim 1 wherein a microorganism enters the L x M array of cell culture chambers through a plurality of secondary micro-channels to avoid contamination of the microfluidic channels and the dilutional channels by the microorganism.

3. The method of claim 2 wherein the cell culture chambers have an outlet to avoid contamination of the microfluidic channels and dilutional channels by a microorganism from the cell culture chamber.

4. The method of claim 1 wherein a technique for imaging the viable population of microorganisms is non-invasive to the viable population.

5. The method of claim 1 wherein biological activity is assessed by a chemical imaging technique, wherein the chemical imaging techniques comprises infrared spectroscopy, Raman spectroscopy, or hyperspectral imaging.

6. The method of claim 5 wherein the chemical imaging technique comprises confocal Raman microscopy (CRM), wherein biological activity is assessed by a Raman biomolecular signature of the viable population.

7. The method of claim 6 wherein the Raman biomolecular signature of the colony distinguishes between living microorganisms and dead microorganisms in the viable population.

8. The method of claim 6 wherein the Raman biomolecular signature of the viable population is analyzed by principal component analysis (PCA), the Raman biomolecular signature of the viable population is analyzed by hierarchal cluster analysis (HCA), or a combination thereof.

9. The method of claim 1 wherein the microorganism is bacteria.

10. The method of claim 1 wherein the biologically active compound is an anti -infective, an antimicrobial, an antibacterial, an antibiotic, or a bactericide.

1 1. The method of claim 1 wherein the nutrients comprise a succinate, an amino acid, a carbohydrate, an ammonium ion, a nitrate ion, an essential metal ion, a host derived medium, molecular oxygen, or a combination thereof.

12. The method of claim 1 wherein the microfluidic chip is permeable to a gas, or is not permeable to gas.

13. The method of claim 12 wherein the microfluidic chip is constructed from gas permeable polydimethylsiloxane (PDMS).

14. The method of claim 1 wherein a biologically active compound hydrodynamically flows to the L x M array of cell culture chambers, and the biologically active compound enters the microfluidic chip through an inlet on a microfluidic channel.

15. The method of claim 14 wherein the hydrodynamic flow of nutrients is at positive pressure, and the hydrodynamic flow of a biologically active compound is at positive pressure.

16. The method of claim 1 wherein the microfluidic chip has a hydrodynamic flow rate ranging from about 1 nL/hr to about 1000 μί/τΐΓ.

17. The method of claim 1 wherein more than one biologically active compound is screened on the microfluidic chip.

18. The method of claim 1 wherein the product of L x M is greater than 1.

19. The method of claim 1 wherein the sum of inlets is less than or equal to 10.

20. The method of claim 1 wherein the mechanism of action of a biologically active compound is elucidated from the spatial concentration gradient of cell culture growth media in the L x M array of cell culture chambers.