US20100227767A1

US20100227767A1 - Stochastic confinement to detect, manipulate, and utilize molecules and organisms

Info

Publication number: US20100227767A1
Application number: US12/670,739
Authority: US
Inventors: James Q. Boedicker; Rustem F. Ismagilov; Christian Kastrup; Cory Gerdts; Toan Huynh; Hyun Jung Kim; Matthew K. Runyon; Feng Shen
Original assignee: Individual
Current assignee: University of Chicago
Priority date: 2007-07-26
Filing date: 2008-07-28
Publication date: 2010-09-09
Also published as: WO2009048673A3; US20100317085A1; WO2009015390A2; WO2009015390A3; US20160145558A1; US11661577B2; US9090885B2; US20190218497A1; US10202571B2; WO2009048673A2

Abstract

Methods of detecting organisms e.g. bacteria using stochastic confinement effects with microfluidic technologies involving plugs are provided. Signal amplification methods for the detection of molecules are also disclosed.

Description

RELATED APPLICATIONS

The present patent document claims the benefit of the filing date under 35 U.S.C. §119(e) of Provisional U.S. Patent Application Ser. Nos. 60/962,426, filed Jul. 26, 2007, and 61/052,490 filed May 12, 2008, which are hereby incorporated by reference.

FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

This work was supported in part by Grant No. 0526693 awarded by the National Science Foundation (NSF) CRC and under grant numbers EB01903, GM075827, and GM074961 awarded by the National Institutes of Health (NIH). The government has certain rights in the invention.

BACKGROUND

Bacterial infections are a major health problem, leading to more than 130,000 deaths from sepsis annually in the United States alone. (G. S. Martin, D. M. Mannino, S. Eaton and M. Moss, N. Engl. J. Med., 2003, 348, 1546-1554) These deaths are often the result of nosocomial, or hospital acquired, infections and frequently involve drug resistant strains of bacteria. (B. M. Farr, Curr. Opin. Infect. Dis., 2004, 17, 317-322; G. J. Moran, A. Krishnadasan, R. J. Gorwitz, G. E. Fosheim, L. K. McDougal, R. B. Carey and D. A. Talan, N. Engl. J. Med., 2006, 355, 666-674) In addition, bacteremia, the presence of bacteria in the blood, is one of the major causes of sepsis and generally requires a minimum of a day or more to diagnose, increasing the chances of patient mortality. (S. D. Carrigan, G. Scott and M. Tabrizian, Clin. Chem., 2004, 50, 1301-1314) Patient mortality rates further increase when inappropriate antimicrobial treatment is administered, which is estimated to occur in 23-30% of cases. (S. D. Carrigan, G. Scott and M. Tabrizian, Clin. Chem., 2004, 50, 1301-1314)
Shortening the time necessary to detect and identify an effective antibiotic regimen to treat bacterial infections could significantly decrease the mortality rate and reduce the cost of treating patients with sepsis and other aggressive bacterial infections. (H. B. Nguyen, E. P. Rivers, F. M. Abrahamian, G. J. Moran, E. Abraham, S. Trzeciak, D. T. Huang, T. Osborn, D. Stevens and D. A. Talan, Ann. Emerg. Med., 2006, 48, 28-54) However, attempts to reduce the assay time of traditional diagnosis and characterization techniques are impeded by the necessity to incubate bacterial specimens for hours to days to increase the cell density of the sample to detectible levels. To overcome this challenge, new PCR-based detection methods enable diagnosis in the one to four hour time frame. (S. Poppert, A. Essig, B. Stoehr, A. Steingruber, B. Wirths, S. Juretschko, U. Reischl and N. Wellinghausen, J. Clin. Microbiol., 2005, 43, 3390-3397; K. P. Hunfeld, Int. J. Med. Microbiol., 2007, 297, 32-32.) However, these methods only provide a genetic profile of the infecting bacterial species and lack the ability to directly test the bacteria's function, such as susceptibility to particular antibiotics. Although some types of antibiotic resistance have genetic markers, such as the mecA gene for instance, (K. Murakami, W. Minamide, K. Wada, E. Nakamura, H. Teraoka and S. Watanabe, J. Clin. Microbiol., 1991, 29, 2240-2244) genetic markers have not been identified for all antibiotic resistant strains of bacteria. Therefore, antibiotic susceptibility is more accurately determined by a functional assay, especially for bacterial strains with unknown resistance mechanisms.
There are many people (10⁵-10⁶) affected every year, and there is no method of detection that works. Early diagnosis of the presence and type of bacteria in a patient's blood stream would help prevent the death of millions of people dying from sepsis. Currently, blood is drawn from a patient and cultures are done to grow the bacteria. However it usually takes days to weeks to grow enough bacteria to detect them, and by the time they grow in the culture, they also grow in the patient, and the patient becomes very sick.
The broth in the blood culture bottle is the first step in creating an environment in which bacteria will grow. It contains all the nutrients that bacteria need to grow. If the physician expects anaerobic bacteria to grow, oxygen will be kept out of the blood culture bottle; if aerobes are expected, oxygen will be allowed in the bottle.
The bottles are placed in an incubator and kept at body temperature. They are watched daily for signs of growth, including cloudiness or a color change in the broth, gas bubbles, or clumps of bacteria. When there is evidence of growth, the laboratory does a gram stain and a subculture. To do the gram stain, a drop of blood is removed from the bottle and placed on a microscope slide. The blood is allowed to dry and then is stained with purple and red stains and examined under the microscope. If bacteria are seen, the color of stain they picked up (purple or red), their shape (such as round or rectangular), and their size provide valuable clues as to what type of microorganism they are and what antibiotics might work best. To do the subculture, a drop of blood is placed on a culture plate, spread over the surface, and placed in an incubator.
If there is no immediate visible evidence of growth in the bottles, the laboratory looks for bacteria by doing gram stains and subcultures. These steps are repeated daily for the first several days and periodically after that.
When bacteria grow, the laboratory identifies it using biochemical tests and the Gram stain. Sensitivity testing, also called antibiotic susceptibility testing is performed as well. The bacteria are tested against many different antibiotics to see which antibiotics can effectively kill it.
All information is passed on to the physician as soon as it is known. An early report, known as a preliminary report, is usually available after one day. This report will tell if any bacteria have been found yet, and if so, the results of the gram stain. The next preliminary report may include a description of the bacteria growing on the subculture. The laboratory notifies the physician immediately when an organism is found and as soon as sensitivity tests are complete. Sensitivity tests may be complete before the bacteria are completely identified. The final report may not be available for five to seven days. If bacteria are found, the report will include its complete identification and a list of the antibiotics to which the bacteria is sensitive.
What is needed is a faster and better method of detecting organisms, including improving the accuracy and decreasing the man-hours associated with standard blood culturing, and shortening the time necessary to detect and identify an effective antibiotic regimen to treat bacterial infections.

BRIEF SUMMARY

In one embodiment, a method of detecting an organism is provided. The method comprises flowing at least two plugs in a carrier fluid through a microchannel; introducing a sample optionally comprising the organism into the first and second plugs; and analyzing the at least two plugs for the detectable signal. Each plug comprises a plug fluid that is substantially immiscible with the carrier fluid and the organism produces a detectable signal.
In a second embodiment, a method of detecting bacteria in a patient is provided. The method comprises flowing at least two plugs in a carrier fluid through a microchannel; introducing a patient sample optionally comprising bacteria into the at least two plugs; and analyzing the at least two plugs for the detectable signal. Each plug comprises a plug fluid that is substantially immiscible with the carrier fluid and the bacteria produce a detectable signal.
In a third embodiment, a method of detecting bacteria is provided. The method comprises flowing at least two plugs in a carrier fluid through a microchannel; introducing a sample optionally comprising bacteria into the first and second plugs, wherein the bacteria produce a detectable signal; and analyzing the plugs for the detectable signal. Each plug comprises a plug fluid substantially immiscible with the carrier fluid. The first plug comprises a means for detecting a first species of bacteria and the second plug comprises a means for detecting a second species of bacteria different from the first species of bacteria.
In a fourth embodiment, a method of detecting bacteria is provided comprising flowing at least two plugs in a carrier fluid through a microchannel; introducing a sample optionally comprising bacteria into the first and second plugs; and detecting the presence of bacteria bound beads. Each plug comprises a plug fluid substantially immiscible with the carrier fluid. The first plug comprises a first antibody bound bead, the first antibody bound bead comprising a first bead and a first antibody that binds to a first bacteria. The second plug comprises a second antibody bound bead, the second antibody bound bead comprising a second bead and a second antibody that binds to a second bacteria different than the first bacteria. The bacteria bind to the antibody bound beads to form bacteria bound beads.
In a fifth embodiment, a method of screening for antibiotic activity is provided. The method comprises flowing at least two plugs in a carrier fluid through a microchannel. Each plug comprises a plug fluid substantially immiscible with the carrier fluid and an antibody bound bead, the antibody bound bead comprising a bead and an antibody that binds bacteria; introducing a sample which comprises bacteria into the first and second plugs; and detecting the presence of either antibody bound beads or bacteria bound beads. The first plug comprises a first antibiotic candidate and the second plug comprises a second antibiotic candidate. The bacteria bind to the antibody bound beads to form bacteria bound beads.
In a sixth embodiment, a method of screening for antibiotic activity is provided comprising flowing at least two plugs in a carrier fluid through a microchannel; introducing a sample which comprises bacteria into the first and second plugs; detecting the presence of either antibody bound beads or bacteria bound beads. Each plug comprises a plug fluid substantially immiscible with the carrier fluid and an antibody bound bead, the antibody bound bead comprising a bead and an antibody that binds bacteria. The first plug comprises a first antibiotic candidate and the second plug comprises a second antibiotic candidate. The bacteria bind to the antibody bound beads to form bacteria bound beads.
In a seventh embodiment, a method of screening for antibiotic activity is provided comprising flowing at least two plugs in a carrier fluid through a microchannel; introducing a sample which comprises bacteria into the first and second plugs; and detecting the presence of bacterial growth. Each plug comprises a plug fluid immiscible with the carrier fluid and media capable of supporting bacterial growth. The first plug comprises a first antibiotic candidate and the second plug comprises a second antibiotic candidate.
In an eighth embodiment, a method of detecting molecules is provided comprising providing a first set of molecules that constitute an autocatalytic loop, capable of amplification of one of the components; providing a second set of molecules that modulate the autocatalytic loop when the autocatalytic loop reacts with the target molecules; and analyzing for the presence of the target molecules.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates a method of detecting bacteria using stochastic confinement of bacteria into plugs reduces detection time. (a) A schematic drawing illustrating the increase in cell density resulting from the stochastic confinement of an individual bacterium in a nanoliter-sized plug. (b) A schematic drawing illustrates the experimental procedure to compare the detection of bacteria incubated in nanoliter-sized plugs and bacteria incubated in a milliliter-scale culture. (c) An illustrative graph of decreased detection time vs. the log of the plug volume. (d) An illustrative graph of detection times vs. cell density for bacteria incubated in plugs (circles) and bacteria incubated in 96 well plates (crosses) with similar initial cell densities.

FIG. 2. illustrates a method for screening many antibiotics against a bacterial sample using a combination stochastic confinement with microfluidic hybrid and/or cartridge methods. (a) A schematic drawing illustrating the formation of plugs of bacteria, viability indicator, and antibiotic from a preformed array of plugs of different antibiotics. (b) An illustrative graph of fluorescence intensity of the control plugs with no antibiotic (+, blank1, positive control) and vancomycin (A, VCM, negative control). (c) An illustrative bar graph shows the results of the antibiotic screen against the Methicillin Resistant S. aureus (MRSA), indicating that this strain of MRSA was resistant to four antibiotics, but sensitive to two. (d) An illustrative chart shows the agreement between the susceptibility profiles (S, sensitive and R, resistant) of MRSA determined by the plug-based microfluidic screen and the control susceptibility screen using Mueller Hinton plates.

FIG. 3 illustrates detecting active and inactive particles. (a) a solution of target (large circles) and non-target (small circles) particles; (b) active particle decorated with antibodies or small peptides; (c) decorated target particle separated from non-target particles and concentrated by stochastic confinement; (d) rapid detection and optical readout.

FIG. 4 illustrates stochastic confinement of particles by (a) encapsulation in droplets, (b) placement in pores of a membrane, and (c) confinement on materials with restricted transport.

FIG. 5 is an illustrative method for analyte detection. (a) particles stochastically confined into plugs which undergo two-stage amplification on chip to give a macroscale readout; (b) particles confined on membranes which translate their activity to a signal seen by the naked eye on an upper layer; (c) particles trapped in a gel with amplification cascades incorporated localize the output signal over the active particle.

FIG. 6 is an illustrative method used to identify the minimal inhibitory concentration (MIC) of cefoxitin (CFX) for Methicillin Sensitive S. aureus (MSSA) and Methacillin Resistant S. aureus (MRSA). (a) a schematic drawing illustrates formation of plugs of bacteria, viability indicator, and an antibiotic at varying concentrations; (b and c) Using 24 mg/L CFX as the baseline, graphs show the average change in intensity of plugs greater than (solid) and less than (striped) 3 times the baseline for MRSA (b) and MSSA (c).

FIG. 7 is an illustrative combination of stochastic confinement with the plug-based microfluidic assay used to determine susceptibility of bacteria to an antibiotic in a natural matrix, blood plasma. (a) a schematic drawing illustrating formation of plugs of bacteria, viability indicator, antibiotic, and plasma/LB mixture; (b and c) Images and linescans of four representative plugs made from a 1:1 blood plasma/LB sample inoculated with MRSA without (left) and with (right) the addition of AMP; (d and e) Images and linescans of four representative plugs made from a 1:1 blood plasma/LB sample inoculated with MSSA without (left) and with (right) the addition of AMP

FIG. 8 is an illustrative schematic description of a test strip with amplification system.

FIG. 9 is an illustrative schematic drawing of a test strip with both detection region and control “timer” region.

FIG. 10 is an illustrative chemical amplification process involving a one-step positive feedback.

FIG. 11 is an illustrative chemical amplification process involving a two-step positive feedback.

FIG. 12 is an illustrative chemical amplification process with a cascade of two positive feedback loops.

FIG. 13 is an illustrative two-step amplification cascade involving blood coagulation enzymes.

FIG. 14 is a graph of time to get response versus amount of input obtained by simulation.

FIG. 15 is a graph of blood clotting time varied with size of patch of tissue factor.

FIG. 16 is an illustrative example of combining amplification cascades with stochastic confinement.

FIG. 17 is an illustrative example of selective detection of particles by using stochastic confinement.

DETAILED DESCRIPTION OF THE DRAWINGS AND THE PRESENTLY PREFERRED EMBODIMENTS

Definitions

The term “organism” refers to any organisms or microorganism, including bacteria, yeast, fungi, viruses, protists (protozoan, micro-algae), archaebacteria, and eukaryotes. The term “organism” refers to living matter and viruses comprising nucleic acid that can be detected and identified by the methods of the invention. Organisms include, but are not limited to, bacteria, archaea, prokaryotes, eukaryotes, viruses, protozoa, mycoplasma, fungi, and nematodes. Different organisms can be different strains, different varieties, different species, different genera, different families, different orders, different classes, different phyla, and/or different kingdoms.
Organisms may be isolated from environmental sources including soil extracts, marine sediments, freshwater sediments, hot springs, ice shelves, extraterrestrial samples, crevices of rocks, clouds, attached to particulates from aqueous environments, involved in symbiotic relationships with multicellular organisms. Examples of such organisms include, but are not limited to Streptomyces species and uncharacterized/unknown species from natural sources.
Organisms included genetically engineered organisms.
Further examples of organisms include bacterial pathogens such as: Aeromonas hydrophila and other species (spp.); Bacillus anthracis; Bacillus cereus; Botulinum neurotoxin producing species of Clostridium; Brucella abortus; Brucella melitensis; Brucella suis; Burkholderia mallei (formally Pseudomonas mallei); Burkholderia pseudomallei (formerly Pseudomonas pseudomallei); Campylobacter jejuni; Chlamydia psittaci; Clostridium botulinum; Clostridium botulinum; Clostridium perfringens; Coccidioides immitis; Coccidioides posadasii; Cowdria ruminantium (Heartwater); Coxiella burnetii; Enterovirulent Escherichia coli group (EEC Group) such as Escherichia coli-enterotoxigenic (ETEC), Escherichia coli-enteropathogenic (EPEC), Escherichia coli —O157:H7 enterohemorrhagic (EHEC), and Escherichia coli-enteroinvasive (EIEC); Ehrlichia spp. such as Ehrlichia chaffeensis; Francisella tularensis; Legionella pneumophilia; Liberobacter africanus; Liberobacter asiaticus; Listeria monocytogenes; miscellaneous enterics such as Klebsiella, Enterobacter, Proteus, Citrobacter, Aerobacter, Providencia, and Serratia; Mycobacterium bovis; Mycobacterium tuberculosis; Mycoplasma capricolum; Mycoplasma mycoides ssp mycoides; Peronosclerospora philippinensis; Phakopsora pachyrhizi; Plesiomonas shigelloides; Ralstonia solanacearum race 3, biovar 2; Rickettsia prowazekii; Rickettsia rickettsii; Salmonella spp.; Schlerophthora rayssiae varzeae; Shigella spp.; Staphylococcus aureus; Streptococcus; Synchytrium endobioticum; Vibrio cholerae non-01; Vibrio cholerae O1; Vibrio parahaemolyticus and other Vibrios; Vibrio vulnificus; Xanthomonas oryzae; Xylella fastidiosa (citrus variegated chlorosis strain); Yersinia enterocolitica and Yersinia pseudotuberculosis; and Yersinia pestis.
Further examples of organisms include viruses such as: African horse sickness virus; African swine fever virus; Akabane virus; Avian influenza virus (highly pathogenic); Bhanja virus; Blue tongue virus (Exotic); Camel pox virus; Cercopithecine herpesvirus 1; Chikungunya virus; Classical swine fever virus; Coronavirus (SARS); Crimean-Congo hemorrhagic fever virus; Dengue viruses; Dugbe virus; Ebola viruses; Encephalitic viruses such as Eastern equine encephalitis virus, Japanese encephalitis virus, Murray Valley encephalitis, and Venezuelan equine encephalitis virus; Equine morbillivirus; Flexal virus; Foot and mouth disease virus; Germiston virus; Goat pox virus; Hantaan or other Hanta viruses; Hendra virus; Issyk-kul virus; Koutango virus; Lassa fever virus; Louping ill virus; Lumpy skin disease virus; Lymphocytic choriomeningitis virus; Malignant catarrhal fever virus (Exotic); Marburg virus; Mayaro virus; Menangle virus; Monkeypox virus; Mucambo virus; Newcastle disease virus (VVND); Nipah Virus; Norwalk virus group; Oropouche virus; Orungo virus; Peste Des Petits Ruminants virus; Piry virus; Plum Pox Potyvirus; Poliovirus; Potato virus; Powassan virus; Rift Valley fever virus; Rinderpest virus; Rotavirus; Semliki Forest virus; Sheep pox virus; South American hemorrhagic fever viruses such as Flexal, Guanarito, Junin, Machupo, and Sabia; Spondweni virus; Swine vesicular disease virus; Tick-borne encephalitis complex (flavi) viruses such as Central European tick-borne encephalitis, Far Eastern tick-borne encephalitis, Russian spring and summer encephalitis, Kyasanur forest disease, and Omsk hemorrhagic fever; Variola major virus (Smallpox virus); Variola minor virus (Alastrim); Vesicular stomatitis virus (Exotic); Wesselbron virus; West Nile virus; Yellow fever virus; and South American hemorrhagic fever viruses such as Junin, Machupo, Sabia, Flexal, and Guanarito.
Further examples of organisms include parasitic protozoa and worms, such as: Acanthamoeba and other free-living amoebae; Anisakis sp. and other related worms Ascaris lumbricoides and Trichuris trichiura; Cryptosporidium parvum; Cyclospora cayetanensis; Diphyllobothrium spp.; Entamoeba histolytica; Eustrongylides sp.; Giardia lamblia; Nanophyetus spp.; Shistosoma spp.; Toxoplasma gondii; Filarial nematodes and Trichinella. Further examples of analytes include allergens such as plant pollen and wheat gluten.
Further examples of organisms include fungi such as: Aspergillus spp.; Blastomyces dermatitidis; Candida; Coccidioides immitis; Coccidioides posadasii; Cryptococcus neoformans; Histoplasma capsulatum; Maize rust; Rice blast; Rice brown spot disease; Rye blast; Sporothrix schenckii; and wheat fungus.
Further examples of organisms include worms such as C. Elegans and pathogenic worms nematodes.
“Particle” as used herein refers to an organism, molecule, cell, a viral particle, spore, and the like.
“Patient sample” refers to a sample obtained from a patient or person and includes blood, feces, urine, saliva or other bodily fluid, preferably blood. Food samples may also be analyzed.
“Sample” refers to any sample potentially comprising an organism. Environments for finding organisms include, but are not limited to geothermal and hydrothermal fields, acidic soils, sulfotara and boiling mud pots, pools, hot-springs and geysers where the enzymes are neutral to alkaline, marine actinomycetes, metazoan, endo and ectosymbionts, tropical soil, temperate soil, arid soil, compost piles, manure piles, marine sediments, freshwater sediments, water concentrates, hypersaline and super-cooled sea ice, arctic tundra, Sargosso sea, open ocean pelagic, marine snow, microbial mats (such as whale falls, springs and hydrothermal vents), insect and nematode gut microbial communities, plant endophytes, epiphytic water samples, industrial sites and ex situ enrichments. Additionally, a sample may be isolated from eukaryotes, prokaryotes, myxobacteria (epothilone), air, water, sediment, soil or rock, a plant sample, a food sample, a gut sample, a salivary sample, a blood sample, a sweat sample, a urine sample, a spinal fluid sample, a tissue sample, a vaginal swab, a stool sample, an amniotic fluid sample and/or a buccal mouthwash sample.
Microfluidics is an attractive platform for rapid single-cell functional analysis. (M. Y. He, J. S. Edgar, G. D. M. Jeffries, R. M. Lorenz, J. P. Shelby and D. T. Chiu, Anal. Chem., 2005, 77, 1539-1544; A. Grodrian, J. Metze, T. Henkel, K. Martin, M. Roth and J. M. Kohler, Biosens. Bioelectron., 2004, 19, 1421-1428; D. B. Weibel, W. R. DiLuzio and G. M. Whitesides, Nat. Rev. Microbiol., 2007, 5, 209-218; Y. Marcy, T. Ishoey, R. S. Lasken, T. B. Stockwell, B. P. Walenz, A. L. Halpern, K. Y. Beeson, S. M. D. Goldberg and S. R. Quake, PLoS Genet., 2007, 3, 1702-1708; J. El-Ali, S. Gaudet, A. Gunther, P. K. Sorger and K. F. Jensen, Anal. Chem., 2005, 77, 3629-3636; A. Huebner, M. Srisa-Art, D. Holt, C. Abell, F. Hollfelder, A. J. Demello and J. B. Edel, Chem. Commun., 2007, 1218-1220; H. M. Yu, C. M. Alexander and D. J. Beebe, Lab Chip, 2007, 7, 726-730; C. J. Ingham, A. Sprenkels, J. Bomer, D. Molenaar, A. van den Berg, J. Vlieg and W. M. de Vos, Proc. Natl. Acad. Sci. U.S.A., 2007, 104, 18217-18222; R. D. Whitaker and D. R. Walt, Anal. Chem., 2007, 79, 9045-9053.) Plugs, for example, droplets of aqueous solution surrounded by a fluorinated carrier fluid, provide a simple platform for manipulating samples with no dispersion or losses to interfaces. (H. Song, D. L. Chen and R. F. Ismagilov, Angew. Chem.-Int. Edit., 2006, 45, 7336-7356; H. Song, J. D. Tice and R. F. Ismagilov, Angew. Chem.-Int. Edit., 2003, 42, 768-772.) Microfluidic plug-based assays provide the ability to reduce detection time by confining bacterium into nanoliter-sized plugs. This confinement, referred to as “stochastic confinement” decreases detection time by confining the sample into plugs that either have a single bacterium, or are empty. This approach increases the effective concentration of the bacterium, and allows released molecules to accumulate in the plug. Such stochastic trapping is commonly used for single-cell analysis in microfluidic devices, (M. Y. He, J. S. Edgar, G. D. M. Jeffries, R. M. Lorenz, J. P. Shelby and D. T. Chiu, Anal. Chem., 2005, 77, 1539-1544; Y. Marcy, T. Ishoey, R. S. Lasken, T. B. Stockwell, B. P. Walenz, A. L. Halpern, K. Y. Beeson, S. M. D. Goldberg and S. R. Quake, PLoS Genet., 2007, 3, 1702-1708; A. Huebner, M. Srisa-Art, D. Holt, C. Abell, F. Hollfelder, A. J. Demello and J. B. Edel, Chem. Commun., 2007, 1218-1220; S. Takeuchi, W. R. DiLuzio, D. B. Weibel and G. M. Whitesides, Nano Lett., 2005, 5, 1819-1823; P. Boccazzi, A. Zanzotto, N. Szita, S. Bhattacharya, K. F. Jensen and A. J. Sinskey, App. Microbio. Biotech., 2005, 68, 518-532; V. V. Abhyankar and D. J. Beebe, Anal. Chem., 2007, 79, 4066-4073) and similar techniques have been used for single molecule and single enzyme work. (H. H. Gorris, D. M. Rissin and D. R. Walt, Proc. Natl. Acad. Sci. U.S.A., 2007, 104, 17680-17685; A. Aharoni, G. Amitai, K. Bernath, S. Magdassi and D. S. Tawfik, Chem. Biol., 2005, 12, 1281-1289; O. J. Miller, K. Bernath, J. J. Agresti, G. Amitai, B. T. Kelly, E. Mastrobattista, V. Taly, S. Magdassi, D. S. Tawfik and A. D. Griffiths, Nat. Methods, 2006, 3, 561-570; J. Huang and S. L. Schreiber, Proc. Natl. Acad. Sci. U.S.A., 1997, 94, 13396-13401; D. T. Chiu, C. F. Wilson, F. Ryttsen, A. Stromberg, C. Farre, A. Karlsson, S. Nordholm, A. Gaggar, B. P. Modi, A. Moscho, R. A. Garza-Lopez, O. Orwar and R. N. Zare, Science, 1999, 283, 1892-1895; J. Yu, J. Xiao, X. J. Ren, K. Q. Lao and X. S. Xie, Science, 2006, 311, 1600-1603.) Microfluidics also enables simultaneous execution of numerous assays of bacterial function from a single bacterial sample in the same experiment, which is especially useful for rapid antibiotic susceptibility screening. Previously, gel microdroplets had been utilized for susceptibility screening. (Y. Akselband, C. Cabral, D. S. Shapiro and P. McGrath, J. Microbiol. Methods, 2005, 62, 181-197; C. Ryan, B. T. Nguyen and S. J. Sullivan, J. Clin. Microbiol., 1995, 33, 1720-1726) However, this method did not take advantage of the stochastic confinement effects in plugs or high-throughput screening methods of current microfluidic technologies. Confinement effects as described herein are increased if the gel droplets are surrounded by a barrier substantially impermeable to released products (for example, a fluorous liquid or a non-porous solid). An immiscible fluid surrounding the droplet will form a barrier to prevent or reduce loss of released products from a cell, enabling the released products to accumulate more rapidly and reach higher concentrations in the droplet Microfluidic technology offers two advantages over traditional bacterial detection and drug screening methods: 1) stochastic confinement of single cells from dilute samples concentrates the bacteria, eliminates the need for pre-incubation, and reduces detection time; 2) each assay can be performed by using an individual bacterium, enabling hundreds of assays to be performed using a single, low density bacterial sample without pre-incubation. This technology will reduce the time needed to diagnose bacterial infections and enable patient-specific antibiotic regimens. This technology also has the advantage of separating objects into individual and separate volumes of fluid by forming plugs. A further advantage is that by using techniques such as the hybrid method to perform high throughput screening of multiple reagents and conditions using only a small volume sample.
Examples of microfluidic technology including descriptions of uses, applications and techniques for plug-based methods of analysis, manipulation of plugs, hybrid plug merging, cartridge formation, use and handling, and holding component and loading component handling, use and formation, use of markers and fluid handling include U.S. Pat. No. 7,029,091; U.S. Published Patent Applications 2005/0087122 A1, 2005/0019792 A1, 2007/0172954 A1, 2007/0195127 A1, 2007/0052781 A1, 2006/0003439 A1, 2006/0094119 A1, 2006/0078893 A1; 2006/0078888 A1, 2007/0184489 A1, 2007/0092914 A1, 2005/0221339 A1, 2007/0003442 A1, 2006/0163385 A1, 2005/0172476 A1, 2008/0003142 A1, 2008/0014589 A1; and WIPO published international applications WO 07/081,386 A2, WO 07/081,387 A1, WO 07/133,710 A2, WO 07/081,385 A2, WO 08/063,227 A2, WO 07089541 A2, WO 07/030,501 A2, WO 06/096571 A2. These references are incorporated by reference in their entirety.
Approaches that may be used for bacterial detection include: 1) the use of cartridges (cartridges pre-loaded with reagents for different detection methods/growth media/antibiotics may be used to detect and identify bacteria); 2) the hybrid method to test many antibiotics/substrates/detection growth conditions or culturing conditions at different concentrations; and 3) screening/sorting.
In addition to concentrating the sample by stochastic confinement, other microfluidic on-chip approaches may be used to preconcentrate a sample before detection. One method is to flow the initial sample through a device which contains a structure (such as a filter) that would collect all of the objects (cells, particles, molecules or proteins bound to particles) needed to be detected. The structure traps the objects (through size exclusion such as a filter or through specific chemical or physical interactions with the objects) but enables the aqueous fluid in which the objects are suspended to pass through the structure. Once all of the objects have been collected in the structure, the objects may be resuspended by another aqueous flow such that the volume of aqueous fluid used to resuspend the objects is less than the volume in which the objects were originally suspended. The resuspended objects, now at a higher concentration, may then be loaded into plugs or droplets for further concentration due to stochastic confinement.
Other concentration techniques may be used such as centrifugation, attaching magnetic beads to the particles of interest, or having surfaces which selectively bind to the target and then release the target at a later time. For example, see S. Song and A Singh, Analytical and Bioanalytical Chemistry, Volume 384, Number 1/January, 2006, 41-43, and P. Gridzinski, J Yang, R H Liu, M D Ward, Biomedical Microdevices, Volume 5, Number 4/December, 2003, pg. 303-310.
The experiments may be done by slowly flowing plugs into a tube, incubating them as they flow (with heating/cooling for PCR if need be along the way), flowing them by the detector, and then dumping them. This may address the “storage of 1,000,000 plugs” problem.
The 1,000,000 plugs problem with stochastic confinement refers to the fact that to concentrate a sample 1,000,000 fold, 999,999 empty plugs must be formed for every 1,000,000 fold concentrated plug. One approach to this problem is do some initial screening using a less sophisticated method (such as flow cytometry or optical scanning) to sort plugs into occupied/unoccupied groups. These simple tests would not provide a detailed characterization of the object in the plug, but would rapidly determine whether or not the plug is occupied by an object. Then the advantage of stochastic confinement can still be realized by running more tests to rapidly characterize the occupied plugs. It is also possible to run an initial screen that does take advantage of stochastic confinement, by first running a general test for the presence of the object (a simple fluorescent viability assay for bacteria) and then sorting the plugs into occupied/unoccupied before adding additional sets of reagents to test for further characterization of the object. It is also possible to take a blood sample (for example 10 mL), make it into a bulk emulsion, apply the procedure above, and then sort the plugs or analyze the plugs using flow cytometry and the like.
A possible method for encapsulating large numbers of cells is similar to sequential merging of reagents to plugs flowing in a 1D microfluidic channel (straight channel). However, the cells are introduced into a 2D channel (width of channel much larger than width of cell), and the cells flow through regions in which various reagents are applied to the cells. The 2D channel should be approximately similar in height to the cells such that the cells flowing through the device form a monolayer. For detection, plugs may or may not be encoded. If the plugs are not encoded, plugs may be analyzed for those that respond for more complex tests like susceptibility tests. Plugs may be encoded by position, or by an internal marker (for example a fluorescent marker of different colors) if not encoded. The same tests may be done by performing multiple non-encoded tests in parallel.

Blood Cultures

There are many variables involved in performing a blood culture. Before a person's blood is drawn, the physician must make several decisions based on knowledge of infections and the person's clinical condition and medical history.
Several groups of microorganisms, including bacteria, viruses, mold, and yeast, can cause blood infections. The bacteria group can be further broken down into aerobes and anaerobes. Most aerobes do not need oxygen to live. They can grow with oxygen (aerobic microbes) or without oxygen (anaerobic microbes).
Based on the clinical condition of the patient, the physician determines what group of microorganisms is likely to be causing the infection and then orders one or more specific types of blood culture, including aerobic, anaerobic, viral, or fungal (for yeasts and molds). Each specific type of culture is handled differently by the laboratory. Most blood cultures test for both aerobic and anaerobic microbes. Fungal, viral, and mycobacterial blood cultures can also be done, but are less common.
The physician must also decide how many blood cultures should be done. One culture is rarely enough, but two to three are usually adequate. Four cultures are occasionally required. Some factors influencing this decision are the specific microorganisms the physician expects to find based on the person's symptoms or previous culture results, and whether or not the person has had recent antibiotic therapy.
The time at which the cultures are to be drawn is another decision made by the physician. During most blood infections (called intermittent bacteremia) microorganisms enter the blood at various time intervals. Blood drawn randomly may miss the microorganisms. Since microorganisms enter the blood 30-90 minutes before the person's fever spikes, collecting the culture just after the fever spike offers the best likelihood of finding the microorganism. The second and third cultures may be collected at the same time, but from different places on the person, or spaced at 30-minute or one-hour intervals, as the physician chooses. During continuous bacteremia, such as infective endocarditis, microorganisms are always in the blood and the timing of culture collection is less important. Blood cultures should always be collected before antibiotic treatment has begun.
Bacteria are the most common microorganisms found in blood infections. Laboratory analysis of a bacterial blood culture differs slightly from that of a fungal culture and significantly from that of a viral culture.
Blood is drawn from a person and put directly into a blood culture bottle containing a nutritional broth. After the laboratory receives the blood culture bottle, several processes must be completed: providing an environment for the bacteria to grow; detecting the growth when it occurs; identifying the bacteria that grow; testing the bacteria against certain antibiotics to determine which antibiotic will be effective.
Given that a typical 5 mL blood sample from a patient with bacteremia contains a cell density of 100 CFU/mL, (L. G. Reimer, M. L. Wilson and M. P. Weinstein, Clin. Microbiol. Rev., 1997, 10, 444-7) the methods of the present invention are capable of performing dozens of functional tests on such a sample. Patient-specific characterization of bacterial species not only allows more rapid and effective treatment, but also enables in-depth characterization of bacterial infections at the population level. Such detailed characterization can aid in tracking and identifying new resistance patterns in bacterial pathogens. (S. K. Fridkin, J. R. Edwards, F. C. Tenover, R. P. Gaynes and J. E. McGowan, Clin. Infect. Dis., 2001, 33, 324-329; and R. T. Horvat, N. E. Klutman, M. K. Lacy, D. Grauer and M. Wilson, J. Clin. Microbiol., 2003, 41, 4611-4616) The principles of these methods, stochastic single-cell confinement and multiple functional assays without sample pre-incubation, can also be applied to other areas, including performing functional tests on field samples, detecting contamination of food or water, separating and testing samples with mixtures of species, measuring functional heterogeneity in bacterial populations, and monitoring industrial bioprocesses.
Other patient samples that may be collected include feces, urine or saliva. The latter would be useful for assessing oral health. Tears may be collected for diagnosing eye infections. Interstitial fluid (also referred to as tissue fluid or intercellular fluid) may be collected and use to diagnose infections in the pleural space around the lungs.
Examples of bacterial infections, include, but are not limited to those listed in Table 1.

TABLE 1

Types of Bacterial Infection

Type of Infection	Description	Examples

Inapparent	No detectable clinical	Asymptomatic
(subclinical)	symptoms of infection	gonorrhea in women
		and men
Dormant (latent)	Carrier state	Typhoid carrier
Accidental	Zoonosis or	Anthrax, cryptococcal
	environmanetal or	infection, and
	inadvertent exposures	laboratory exposure,
		respectively
Opportunistic	Infection caused by	Serrati or Candida
	normal flora or	infection of the
	transient bacteria	genitourinary tract
	when normal host
	defenses are
	compromised
Primary	Clinically apparent	Shigella dysentery
	(e.g. invasion and
	multiplication of
	microbes in body
	tissues, causing local
	tissue injury)
Secondary	Microbial invasion	Bacterial pneumonia
	subsequent to primary	following viral lung
	infection	infection
Mixed	Two or more	Anaerobic abscess (E. coli
	microbes infecting the	and Bacteroides
	same tissue	fragilis)
Acute	Rapid onset (hours or	Diptheria
	days); brief duration
	(days or weeks)
Chronic	Prolonged duration	Mycobacterial
	(months or years)	diseases (tus and
		leprosy)
Localized	Confined to a small	Staphylococcal boil
	area or to an organ
Generalized	Disseminated to many	Gram-negative
	body regions	bacteria
	(gonococcernia)
Pyogenic	Pus-forming	Staphylococcal and
		streptococcal infection
Retrograde	Microbes ascending	E. coli urinary tract
	in a duct or tube	infection
	against the flow of
	secretions or
	excretions
Fulminant	Infections that occur	Airborne Yersinia
	suddenly and	pestis (pneumonic
	intensely	plague)

Bacterial Detection
To monitor the presence and metabolically active bacteria in plugs, a fluorescent viability indicator alamarBlue® was added to the cultures. The active ingredient of alamarBlue is the fluorescent redox indicator resazurin. (J. O'Brien and F. Pognan, Toxicology, 2001, 164, 132-132.) Resazurin is reduced by electron receptors used in cellular metabolic activity, such as NADH and FADH, to produce the fluorescent molecule resofurin. Therefore, fluorescence intensity in a plug is correlated with the presence and metabolic activity of a cell, in this case, a bacterium. Because resazurin indicates cell viability, resazurin-based assays have been used previously in antibiotic testing. (S. G. Franzblau, R. S. Witzig, J. C. McLaughlin, P. Torres, G. Madico, A. Hernandez, M. T. Degnan, M. B. Cook, V. K. Quenzer, R. M. Ferguson and R. H. Gilman, J. Clin. Microbiol., 1998, 36, 362-366; A. Martin, M. Camacho, F. Portaels and J. C. Palomino, Antimicrob. Agents Chemother., 2003, 47, 3616-3619; K. T. Mountzouros and A. P. Howell, J. Clin. Microbiol., 2000, 38, 2878-2884; C. N. Baker and F. C. Tenover, J. Clin. Microbiol., 1996, 34, 2654-2659.) Resazurin may be used to detect both the presence of a live bacterium and the response of bacteria to drugs, such as antibiotics. Stochastic confinement decreases detection time because in a plug that has the bacterium, the bacterium is at an effectively higher concentration than in the starting solution, and the signal-to-noise required for detection is reached sooner since the product of reduction of resazurin accumulates in the plug more rapidly.
To demonstrate the ability of stochastic confinement to reduce detection time, a single sample of Staphylococcus aureus (S. aureus) containing the fluorescent viability indicator was split. Half of the culture was used to generate plugs of nanoliter volume, and the other half remained as a milliliter-scale culture. Both the nanoliter plugs and the milliliter-scale culture were incubated for 2.8 h at 37° C. After incubation, the milliliter-scale culture was used to form plugs. This experimental procedure is illustrated in FIG. 1 b. Line scans indicate that confining the bacteria at the beginning of incubation (t=0), led to a few occupied plugs with a high fluorescence intensity and many empty plugs with low fluorescence intensity (solid line). All plugs made from the milliliter-scale culture had an intermediate intensity (dotted line). Confining bacteria into plugs of nanoliter volume reduced the time required to detect a change in fluorescence intensity of the viability indicator. Bacteria confined to and incubated in nanoliter-sized plugs showed a greater change in fluorescence intensity after 2.8 h than the bacteria incubated in the “unconfined” milliliter-scale culture (FIG. 1 c). Line scans of the plugs of bacteria that were incubated in plugs showed many empty plugs with low fluorescence intensity and a few occupied plugs with high fluorescence intensity (FIG. 1 b, top). However, lines scans of plugs of bacteria that were incubated in the milliliter-scale culture have a lower, uniform fluorescence intensity (FIG. 1 b, bottom). Therefore, bacteria confined to nanoliter-sizes plugs may be detected earlier than bacteria in a milliliter-scale culture.
In plugs containing single bacterium, the detection time was proportional to plug volume. Detection time was defined as the time at which the increase in fluorescence intensity reached a maximum. When single bacterium were confined in plugs ranging from 1 mL to 1500 mL in volume, detection time increased with the log of plug volume (FIG. 1 c), implying that bacteria were dividing exponentially inside the plugs. This result is similar to previous estimates that detection time decreases by about 1.5 h for every order of magnitude increase in cell density. (P. Kaltsas, S. Want and J. Cohen, Clin. Microbiol. Infect., 2005, 11, 109-114) The detection times measured for bacteria incubated in plugs were similar to detection times measured for bacteria incubated in a 96 well plate from cultures with similar initial cell densities (FIG. 1 d). This result implies that incubation in plugs had no adverse effects on growth of bacteria.
Detecting low concentrations of species (down to single molecules and single bacteria) is a challenge in food, medical, and security industries. Plugs may allow one to concentrate such samples and perform analysis. For example, a sample containing small amounts of DNA of interest in the presence of an excess of other DNA may be amplified. Amplification may be detected if plugs are made small enough that some plugs contain single DNA molecules of interest, and other plugs contain no DNA molecules of interest. This separation into plugs effectively creates plugs with higher DNA of interest concentration than in the original sample. Amplification of DNA in those plugs, for example by PCR, may lead to higher signal than amplification of the original sample. In addition, localization of bacteria in plugs by a similar method may create a high local concentration of bacteria (1 per very small plug), making them easier to detect. For some bacteria that use quorum sensing, this may be a method to activate and detect them. Such bacteria may be inactive/non-pathogenic and difficult to detect at low concentrations due to lack of activity, but at a high concentration of bacteria, the concentration of a signaling molecule increases, activating the bacteria. If a single bacterium is localized in a plug, the signaling molecule produced by a bacterium cannot diffuse away and its concentration will rapidly increase, triggering activation of the bacterium, making it possible for detection. In addition, plugs may be used to localize cells and bacteria by creating gels or matrixes inside plugs. Bacteria and other species (particles and molecules) may be collected and concentrated into plugs by putting air through a plug fluid such as water, and then using that plug fluid to generate plugs. For example, by making smaller plugs from the initial plug, some of the newly formed smaller plugs will contain sample while other plugs will not contain the sample, but only buffer, for example. This results in concentrated sample containing plugs because some of the plugs do not contain any of the sample.
This method is not limited to liquid samples. Microorganisms and other particulate matter can be detected in gaseous samples, such as samples of air taken at airports or along train routes. There are numerous methods for collecting airborne particles in water, for example, as described in U.S. Pat. Nos. 7,201,878, 7,243,560, and 5,855,652 all incorporated by reference herein in their entirety. After collecting the airborne particulate matter in water, these samples can then be added directly to the fluorinated oil carrier fluid to form plugs.
PCR techniques are disclosed in the following published US patent applications and International patent applications: US 2008/0166793 A1, WO 08/069,884 A2, US 2005/0019792 A1, WO 07/081,386 A2, WO 07/081,387 A1, WO 07/133,710 A2, WO 07/081,385 A2, WO 08/063,227 A2, US 2007/0195127 A1, WO 07/089,541 A2, WO 07030501 A2, US 2007/0052781 A1, WO 06096571 A2, US 2006/0078893 A1, US 2006/0078888 A1, US 2007/0184489 A1, US 2007/0092914 A1, US 2005/0221339 A1, US 2007/0003442 A1, US 2006/0163385 A1, US 2005/0172476 A1, US 2008/0003142 A1, and US 2008/0014589 A1, all of which are incorporated by reference herein in their entirety.
Amplifications of nucleic acids have been performed via polymerase chain reactions (PCR). The key concept of PCR comprises genetic template (primer), thermostable DNA polymerase, and the circuit for regulating temperature. By combining these components in a miniaturized microfluidic device capable of implementing cartridge and/or hybrid methods, a high concentration of interesting DNA fragments from a very small amount of genetic sample via trivial PCR methods may be collected. It has been confirmed that PDMS is a heat-stable material, indicating it is a suitable material for the PCR process.
With respect to the cartridge, the plug-based microfluidic platform may be used to setup a huge number of reaction centers in nano- or pico-liter volume scale, extending into the femtoliter and microliter scales. With respect to the hybrid method, various conditions of reactions and samples may be incorporated, split, and merged in a microfluidic device.
For example, a sample comprising 1% of DNA of interest in the presence of 99% of background DNA would need to be amplified enough to harvest and use. However, amplification of 1% of DNA by a factor of 100 only increases the total amount of DNA by a factor of 2, resulting in detection difficulties. By making small enough plugs to comprise a single molecule of DNA of interest, and amplifying each plug by means of conventional PCR techniques in a microfluidic device, highly amplified PCR products of target 1% of DNA in a plug may be obtained. Assume that the probability of appearance of target DNA as a single molecule in a plug is one in every 10^thplug, then the ratio of target DNA/total DNA is 1:10 in such plugs. 1000-Fold amplifications of target DNA molecule in those plugs, affords a 100:10 ratio of target to background DNA, resulting in a 10 fold increase of the total amount of DNA. Conclusively, the target DNA from the very low concentration may easily be detected under the presence of large background signals. In the microfluidic device, a platform cartridge incorporating a carrier fluid channel, a sampling channel, and one or two PCR reagent channels may be made. By regulating each fluid, individual plugs containing a single molecule of target DNA or different molecules of DNA from the sample may be generated. Along the channel containing plugs, heating regions are placed in the channel region by repeating heating and cooling process for the denaturation and renaturation of DNA samples.
Stochastic confinement has applications in the isolation and screening of rare particles or cells from a sample. When samples are stochastically confined, the result is a set of isolated volumes of fluid, most with either no particles or 1 particle. In this way, rare particles are segregated from ubiquitous particles. The separation of rare particles enables the direct assay of the function and detection of rare particles without interference from other particles in the system.
Even if a rare particle has high activity, since it is at low concentration in the bulk sample it may not be detectable due to dilution of the signal and low signal to noise ratio as a result of low background reaction from the ubiquitous particles. Once rare particles are isolated and concentrated through stochastic confinement, other plug based microfluidic technologies can be used to screen the rare particle.
For example, if the original sample contains 100 rare particles, since stochastic confinement enables the separation and screening of individual particles, this enables the rare particle to be screened against up to 100 different conditions. After confining all of the particles into droplets, those droplets may be merged with thousands of different conditions. If the droplets are randomly merged with screening conditions, then up to 100 different conditions will be screened against the rare particles. If the rare particle is allowed to divide, then the plug containing the rare particle can be split into several smaller plugs and each plug assayed or screened independently, for example, using the hybrid and/or cartridge method. Examples include combining the plug containing many rare particles with plugs of various reagents and conditions in order to determine the function and optimal conditions for the rare particle. As an additional example, several stochastically confined organisms may be allowed to grow inside an array of plugs. After the array is split into four daughter arrays, each daughter array may be interrogated by a different technique or reagent, while preserving the identity of plugs and their relationship in the daughter array (for example, 37^thplug in the first array corresponds to the 37^thplug in the second, third, and forth array). The results may be combined to provide information on the response of corresponding daughter plugs to each of the techniques. In addition, some of the daughter arrays may be retained as a reference culture. When the results from the other three arrays are known, the reference culture arrays may be used for further manipulation, characterization, assaying, and isolation of organisms.
Isolation of rare particles through stochastic confinement may also be combined with the hybrid and/or cartridge method for screening growth/virulence activation/assay conditions against the rare cells. For example, if a sample contains 10,000 cells, 100 of which belong to a rare cell type, stochastic confinement may be used to isolate the 100 rare cells into plugs containing only a single rare cell and no other types of cells. Then the plugs containing the rare cell types may be used in hybrid and/or cartridge screening by the following methods.
The plugs generated from the stochastic confinement may be combined with screening conditions using the hybrid and/or cartridge method. Plugs containing the rare cells are randomly distributed throughout all of the plugs generated (many of which do not contain a rare cell). If many plugs (100's, 1000's) are merged with a single screening condition, it is likely that at least one of the plugs for that condition will contain a rare cell. In this way, multiple growth/assay/virulence activation conditions may be screened against the rare cell type. A separate test may be needed to separate plugs containing rare cells from the other plugs generated by confinement. It may be best to sort the plugs into rare cell/common cell/empty after merging with the hybrid and/or cartridge screen to reduce detection time. Instead of sorting for rare cells first (which may take time) and then merging with screening conditions, the screening and determining the presence of rare cell may be done simultaneously.
Alternatively, it may be desirable to first separate out plugs containing rare cells from plugs containing common cells or no cells. This may be done using antibodies, binding assays, testing for function specific to the rare cell combined with automated sorting mechanisms (optical, magnetic, FACS). Once the rare cell type plugs have been isolated, a hybrid and/or cartridge screen can be used to screen for growth/assay/virulence activating conditions or to run a multitude of functional and genetic tests on the rare cell type.
Another application of stochastic confinement is to accurately count populations of cells. Since confinement isolates 1 cell per plug and may be used to perform tests to identify the type of cell in each plug, confinement may be used to determine the density (number of cells of type X per volume of sample) or the ratio (100 cells of type X for every 1 of type Y) of cells in a sample. This may also be used to find ratios of phenotypes of cell populations that are genetically identical (25% of Staphylococcus cells are resistant to oxacillin or 30% of cells will induce virulence in response to host protein X).
Screening for growth conditions is an important application because an estimated 99% of all microbes cannot be cultured by standard techniques. Unculturability of these organisms may be due to: 1) nutrient levels of media are too high; 2) requirement of specific ion concentrations; and 3) requirement for additional factors (unusual compounds not found in most media). Confinement has two effects: 1) it reduces competition from other microbes, giving rare and slow growing cell types time to reproduce; and 2) it can be combined with the hybrid and/or cartridge method to screen for media additives and the concentration of the additive to find new growth conditions for a previously uncultured microbe or a microbe which is difficult to grow or slow growing under current conditions. In addition, one may also use control of surface chemistry provided by plugs to enhance growth of organisms. Compounds that modulate surface chemistries may be incorporated into the hybrid screen in addition to or along with compounds modulating growth conditions.
In addition, an organism might be releasing products such as quorum sensing molecules and will not initiate growth until a threshold concentration of released products have accumulated. Confinement will put microbial cells at higher initial cell density and enable cells to grow and optionally activate genes associated with high cell density. Some organisms may be able to grow in culture, which were previously believed to be not culturable by standard techniques due to their slow growth or growth to low densities. These organisms may still undergo a sufficient number of divisions when stochastically confined, and therefore allow further detection by less-sensitive techniques that require multiple copies of the organism to be present. When plugs are used to create stochastic confinement, growth of the organism allows further manipulation and analysis that cannot be done on a single plug (for example requiring mutually incompatible methods) by splitting the plugs, injecting reagents into them and monitoring results. These steps may be performed sequentially, where the results of the first experiment guides the design of the second experiment, or in parallel. Examples of mutually incompatible methods include reagents that produce similar signals (such as fluorescence in the same range of wavelengths), or methods that require different conditions (such as different solvents or pH values), or require different states of the organisms (such as a functional test that requires an alive organism, and a staining protocols that kills the organism).
Implementation of this type of screening may be done in many ways. For example, a known organism may be screened in the presence of many media and culturing conditions (varying ion concentrations, known autoinducers, amount of confinement, temperature, pH, protein additives, reaction oxygen species, stress inducers; changing carbon source and concentration of carbon source; changing nitrogen source and concentration of nitrogen source; changing availability of various trace metals (Mn, Mo, Cu, Pt, etc.), adding drugs known to interfere with specific cellular activities; adding transport and ion channel inhibitors, small molecules involved in cellular communication, virulence activators, etc.). After using the hybrid method to screen through many conditions, functional tests or other assays may be performed in plugs to determine if compounds have been generated with properties of interest (such as drug targets, antibiotic compounds, ion channel inhibitors, virulence activation, virulence inhibition, degradation of various compounds, binding affinity, etc.).
One idea to investigate molecules released by microorganisms is called OSMAC (one strain many compounds)), as described in Big Effects from Small Changes: Possible Ways to Explore Nature's Chemical Diversity by Helge Bjorn Bode, Barbara Bethe, Regina Höfs, Axel Zeeck, Chem Bio Chem Volume 3 Issue 7, Pages 619-627. Small changes in culturing conditions (for example, media composition, aeration, culture vessel, addition of enzyme inhibitors) drastically change the metabolites that are released from a cell. The molecules released may aid in detection of the organism, or may have functional uses such as antibiotics.
Therefore, even though strain B. subtilis can be cultured and a lot is known about its genome, useful compounds that it is capable of releasing may be missed simply because the organism has never been grown under specific conditions such as, for example, certain concentration of phosphate ions, addition of protease inhibitor, and addition of 10 uM autoinducer 2. Therefore, using hybrid and/or cartridge like approaches, a larger range of metabolites, released compounds, and drug leads may be probed simply by running high throughput screens of various media conditions/additives. Even changes of a single component in the media (for example, phosphate from 20 mM to 1 uM) may activate the production and release of a previously unknown metabolite or compound.
Release of compounds is highly dependent on culturing conditions. It is known that phosphate levels, temperature, nutrient availability all influence the production and release of various metabolites (J. E. González-Pastor, E. C. Hobbs, R. Losick, Science, 2003, 301, 510). Hybrid and/or cartridge methods developed previously may enable the screening of media conditions and concentrations of additives both for communities, common species, and rare species of microbes.
Implementation of this type of screening may be done by taking a known organism and screening many media and culturing conditions (ion concentrations, known autoinducers, amount of confinement, temperature, pH, protein additives, reaction oxygen species, stress inducers, carbon sources, concentration of carbon source, nitrogen sources, concentration of nitrogen source, availability of various trace elements and their chemical form (Mn, Mo, Cu, Pt, V, B etc. and corresponding ions), and/or by adding drugs known to interfere with specific cellular activities, adding transport and ion channel inhibitors, and/or small molecules involved in cellular communication, virulence activators, and the like.). After using hybrid and/or cartridge method to screen through many conditions, functional tests or other assays may be performed in plugs to determine if compounds have been generated with properties of interest (such as drug targets, antibiotic compounds, ion channel inhibitors, virulence activation, virulence inhibition, degradation of various compounds, binding affinity, etc.). Alternatively, this method may also be used with rare cells isolated from natural samples such as soil, aquatic environments including sea water and marine sediments and surfaces, an animal's digestive tract, environmentally contaminated sites including soil, water or air, sludge used in environmental remediation, etc. Cells which are unculturable (no known conditions cause them to divide/reproduce outside of natural environment) may also be used. If small volumes are used, activity may be detected even from a single cell without division.
It is known that some strains of microbes will not initiate growth or some cellular functions unless cell density is above a minimal threshold. It is also possible that although the process is occurring, the rate at low cell densities may be so slow that it would take weeks or months to observe growth or the function. Therefore, by placing single cells in very confined spaces with small volumes it is likely that they will initiate high density processes and that many processes will have increased rates. This is especially important in the case of rare cells, since the sample may only control 1 or a few copies of the rare cell. Small volume confinement makes it possible to achieve high cell densities of rare cell types.
Stochastic confinement may be used to isolate rare organisms from various sources, including: soil extract from various types of soil environments and soil layers (including the surface layer, subsoil, substratum), ice shelves, marine and freshwater sediments, naturally occurring biofilms, hot springs, hydrothermal vents, extraterrestrial samples, crevices of rocks, attached to particulates from aqueous environments, growing on or inside of manmade structures, clouds, gastrointestinal tract, and found forming a symbiotic relationship inside of a host organism. Specifically, stochastic confinement may be used to isolate rare cells from soil extract. Once rare cells have been obtained, the plugs containing the cells may be incubated overnight to allow for growth or secretion of molecules. The plugs containing the rare cells may then be used as an input into the hybrid and/or cartridge method. For example, to find rare cells or their secreted molecules that may stimulate production of antibiotics and other compounds by Streptomyces species, each plug containing the rare cell may then, for example, be merged with 1000's of plugs containing cells of Streptomyces species. After plugs containing streptomyces have been merged with rare cell supernatant and incubated, screening for antibiotic production is performed. In this way, compounds in the supernatant of the rare cell, or rare cells directly, may induce the production of new antibiotic compounds.
In another example, stochastic confinement may be used to isolate cells from ocean sediments. Once plugs containing cells have been collected, the hybrid and/or cartridge method may be used to screen various media conditions such as phosphate concentration (from 0 to 100 μM), autoinducer 2 concentration (from 0 to 100 μM), and glucose concentration (from 0 to 10 mM). The cells are then incubated in the new media conditions. Various functional/genetic tests are performed in the plugs to determine which media conditions yield growth and or production of compound with desired properties
Alternatively, cells which are unculturable (no known conditions cause them to divide/reproduce outside of natural environment) may be used. If small volumes are used, one may be able to detect activity even from a single cell without division.
Plug based methods may be used to collect the lysate or cell free supernatant from various types of cells and merge these solutions with other cells to elicit metabolite/compound production. Lysate/supernatant may be diluted during the screen (concentration screen using hybrid method). In this way, uncharacterized/unknown compounds/combinations of compounds may be screened to elicit production of useful compounds.
It should be noted that using stochastic confinement to enumerate particles including organisms and cells would also be useful for counting the occurrence of rare cell types in a sample. Stochastic confinement has applications in isolation of rare particles from samples with many other ubiquitous particles besides a few particles of interest. When samples are stochastically confined, the result is a set of isolated volumes of fluid, most with either no particles or one particle. Rare particles are therefore separated from ubiquitous particles. The separation of rare particles enables the direct assay of the rare particles without interference from other particles in the system. Even if a rare particle has high activity, since it is at low concentration in the bulk sample it may not be detectable due to dilution of the signal and low signal to noise or low signal to background (potentially low background reaction from the ubiquitous particles). Once rare particles are isolated and concentrated through stochastic confinement, other plug based microfluidic technologies may be used to screen the rare particle. If the rare particle is allowed to divide, then the plug containing the rare particle can be split into several smaller plugs and assayed or screened using the hybrid and/or cartridge method (combining the plug containing many rare particles with plugs of various reagents and conditions in order to determine the function and optimal conditions for the rare particle). In addition, if the original sample contains 10 rare particles, since stochastic confinement enables the separation and screening of individual particles, this enables the rare particle to be screened against up to 10 different conditions. When rare particles are bacteria, care must be taken with bacteria to avoid forming biofilms during incubation which might grow and adhere to the wall, interfering with enumerations. Thus it is preferable to use inert surfaces in plugs.
Two important purposes when dealing with the particles of interest are detection and harvest, both of which are enabled or greatly enhanced by stochastic confinement when the particles of interest are rare particles mixed with many other ubiquitous particles. Without stochastic confinement, the background is too high compared to the signal for detection, and the probability of isolating the particles of interest is too low for harvest.
Occasions when detection is needed include, but are not limited to, when the particles are rare cells such as cancer cells in general, cancer stem cells, or fetal cells in maternal blood, or when the particles are bacteria or viruses causing some diseases, or when the particles are some toxic materials released by some industrial procedure, or particles are results of some military, civil, or natural event that needs detection.
On the other hand, if the particles of interest have some special functions that are useful, isolating and possibly multiplying them is important in efficiently utilizing these special functions. For example, stem cells isolated and multiplied from adults may be used for treatment, bypassing the need for embryonic stem cells. Natural products used in medicine may be produced by isolating and multiplying natural cells that produce them. Bacteria with novel functions such as cleaning up hydrocarbon waste, degrading other environmental pollutants including halogenated compounds, converting biomass into more easily utilizable fuels such as ethanol or butanol or methane, oxidizing methane, or fixing nitrogen may be isolated, multiplied and used for appropriate purposes.
Therefore, after the particles are separated and the plugs containing the particles of interest are distinguished from others by a primary assay, one or multiple further assays may be done to detect particles of interest and/or one or multiple types of particles detected from the primary assay may be used alone or in combinations for appropriate functions.
The basis for detection of the particles include but are not limited to: 1) surface properties including functional groups on the surface of particles and signaling molecules on cells (antigens, receptors, sugar groups, lipids, etc); 2) materials inside the particles (chemicals enclosed in materials, DNA, RNA in general, microRNA, signaling molecules in general, proteins, and the like)—these materials may require further processing to the particles to be exposed and used for detection; and 3) chemical exchange with the environment (production and/or consumption of chemicals by material, uptake and/or secretion of molecules such as food, waste, signaling molecules in general, ions, novel molecules, etc. by cells). These chemical exchanges may occur naturally or with human intervention for example by stimulation with reagents.
Specific applications
Detection of diseases by examination of fetal materials in maternal blood
Prenatal diagnosis of genetic diseases plays an important role in pregnancy, at least in informing the parents about the possibilities. Women of 35 years of age or more have high risk of abnormalities. However, since there are also many more pregnancies in the “low-risk” 26 year-old group, most (about 70%) abnormalities occur in this group. (Daniilidis, A.; Kouzi-Koliakou, K., Fetal cells in maternal circulation—potentials for prenatal control. Journal of Biological Research-Thessaloniki 2006, 6, 119-130.)
Most effective current methods are invasive, where samples are taken directly from the fetus. These procedures have a risk of miscarriage (1-2%). (Daniilidis, A.; Kouzi-Koliakou, K., Fetal cells in maternal circulation—potentials for prenatal control. Journal of Biological Research-Thessaloniki 2006, 6, 119-130.) Thus, these invasion methods are usually applied to those in the high-risk group only. Because of the risk of miscarriage and because of the high collective occurrence of abnormalities in the “low-risk” group that are usually not screened, a non-invasive and reliable method to detect or predict genetic disorders is in high demand.
One promising possibility is using fetal materials in maternal blood. If the cells and free DNA in the mother's blood may be used effectively to detect genetic disorders, the risk of miscarriage by invasive procedure is diminished and virtually any expecting mother may have a blood draw to check for possible genetic disorders. However, the biggest challenge is the small number of fetal cells (1-6 cells/mL (Daniilidis, A.; Kouzi-Koliakou, K., Fetal cells in maternal circulation—potentials for prenatal control. Journal of Biological Research-Thessaloniki 2006, 6, 119-130.)) and DNA in maternal blood.
Previously developed methods to isolate such cells involve enrichment methods such as density gradient centrifugation and selective lysis, and sorting methods such as fluorescence-activated cell sorting (FACS) and magnetic-activated cell sorting (MACS). Methods to detect genetic disorders include (PCR) and fluorescent in situ hybridization (FISH).
FACS and MACS depend on tagging the fetal cells with fluorescent or paramagnetic antibodies and using fluorescence intensity as a signal to separate the cells with flow cytometry or using magnets to separate the cells. These techniques rely on specific antibodies. In a recent study in which blood samples spiked with fetal nucleated red blood cells were used to check the sorting procedure which used density-gradient centrifugation, MACS, and selective lysis, only 37% were recovered. (Ponnusamy, S.; Mohammed, N.; Ho, S. S. Y.; Zhang, H. M.; Chan, Y. H.; Ng, Y. W.; Su, L. L.; Mahyuddin, A. P.; Venkat, A.; Chan, J.; Rauff, M.; Biswas, A.; Choolani, M., In vivo model to determine fetal-cell enrichment efficiency of novel noninvasive prenatal diagnosis methods. Prenatal Diagnosis 2008, 28, (6), 494-502.)
Prenatal diagnosis of fetal physiology, non-genetic diseases, or genetic disorders by fetal cells and possible use of fetal cells.
Stochastic confinement and autocatalytic kinetics with threshold, as discussed above and in the amplification section, may allow one to detect reliably plugs containing fetal cells in maternal blood and separate them out to use techniques in further detection or application. The two important advantages of this system versus current techniques are:
1) Because of the threshold kinetics, the result in each plug is binary or pseudo-binary. In other words, there is a large contrast between plugs containing fetal cells and other plugs (which contain other cells or no cells). Therefore, the signal used to mechanically sort the cells is clear and error in this step is avoided. For example, in conventional FACS background fluorescence and photobleaching may make the results deviate from being ideal.
2) Stochastic confinement allows for reliable detection even if the specificity of an antibody label is not ideal. As long as binding to the fetal cells (such as fetal nucleated red blood cell) is at least two orders of magnitude stronger than undesired binding to non-fetal cells, the kinetic threshold may be adjusted to lie in between the two concentrations of antibodies in plugs. The adjustment may be carried out by choosing an appropriate amplification method and tuning the concentration of such method (see section about amplification). There is a need to determine properties of fetal cells (such as having a particular disease or not). Using antibodies that selectively bind to fetal cells of interest, stochastic confinement may be used to detect such properties. Even though the contrast provided by the specificity of the antibodies may not be much, stochastic confinement and/or an amplification method may greatly enhance this contrast.
The general steps of this method include:
obtaining about 10-20 mL of blood sample from the expecting mother (the volume is chosen because even in such cases when amplification methods may detect single cells, a 1 mL sample of blood still has a significant probability to not contain the cells of interest when the concentration of the fetal cells is 1-6 cells/mL);
coarse enrichment by density gradient centrifugation or other methods (optional);
stochastic confinement into plugs;
primary detection (such as by using antibodies (such as monoclonal antibody against H315 for trophoblasts, (Daniilidis, A.; Kouzi-Koliakou, K., Fetal cells in maternal circulation—potentials for prenatal control. Journal of Biological Research-Thessaloniki 2006, 6, 119-130) monoclonal antibody against transferring receptor for erythroblass, (Bianchi, D. W.; Flint, A. F.; Pizzimenti, M. F.; Knoll, J. H. M.; Latt, S. A., Isolation of Fetal DNA from Nucleated Erythrocytes in Maternal Blood. Proceedings of the National Academy of Sciences of the United States of America 1990, 87, (9), 3279-3283.) etc.));
optionally if an amplification method is needed to enhance the contrast: merging with plugs containing chemicals or materials needed in the amplification method (if these chemical and materials are added with the antibody this step is not needed);
using the plugs containing fetal cells of interest for further assays (such as FISH or PCR to look for genetic disorders, or any other possible methods to detect certain properties of interest of fetal cells) if necessary; and/or
using the plugs containing fetal cells of interest for applications if there are functions associated with such fetal cells in need, with or without multiplying these cells.
Fetal materials (including DNA) in maternal blood and markers of disorder after stochastic confinement may also be detected with methods described in the section entitled “amplification.”
Detection
The following articles, describing methods for concentrating cells and/or chemicals by making small volume plugs with low numbers of items to no items being incorporated into the plugs, with specific applications involving PCR, are incorporated by reference herein: Anal Chem. 2003 Sep. 1; 75(17):4591-8. Integrating polymerase chain reaction, valving, and electrophoresis in a plastic device for bacterial detection. Koh C G, Tan W, Zhao M Q, Ricco A J, Fan Z H; Lab Chip. 2005 April; 5(4):416-20. Epub 2005 Jan. 28. Parallel nanoliter detection of cancer markers using polymer microchips. Gulliksen A, Solli L A, Drese K S, Sörensen O, Karlsen F, Rogne H, Hovig E, Sirevåg R.; Ann N Y Acad. Sci. 2007 March; 1098:375-88. Development of a microfluidic device for detection of pathogens in oral samples using upconverting phosphor technology (UPT). Abrams W R, Barber C A, McCann K, Tong G, Chen Z, Mauk M G, Wang J, Volkov A, Bourdelle P, Corstjens P L, Zuiderwijk M, Kardos K, Li S, Tanke H J, Sam Niedbala R, Malamud D, Bau H; Sensors, 2004. Proceedings of IEEE 24-27 Oct. 2004 Page(s):1191-1194 vol. 3. A microchip-based DNA purification and real-time PCR biosensor for bacterial detection. Cady, N.C.; Stelick, S.; Kunnavakkam, M. V.; Yuxin Liu; Batt, C. A.; Science. 2006 Dec. 1; 314(5804):1464-7. Microfluidic Digital PCR Enables Multigene Analysis of Individual Environmental Bacteria. Elizabeth A. Ottesen, Jong Wook Hong, Stephen R. Quake, Jared R. Leadbetter; Electrophoresis 2006, 27, 3753-3763. Automated screening using microfluidic chip-based PCR and product detection to assess risk of BK virus-associated nephropathy in renal transplant recipients. Govind V. Kaigala, Ryan J. Huskins, Jutta Preiksaitis, Xiao-Li Pang, Linda M. Pilarski, Christopher J. Backhouse; Journal of Microbiological Methods 62 (2005) 317-326. An insulator-based (electrodeless) dielectrophoretic concentrator for microbes in water. Blanca H. Lapizco-Encinas, Rafael V. Davalos, Blake A. Simmons, Eric B. Cummings, Yolanda Fintschenko; Anal. Chem. 2004, 76, 6908-6914. Electrokinetic Bioprocessor for Concentrating Cells and Molecules. Pak Kin Wong, Che-Yang Chen, Tza-Huei Wang, and Chih-Ming Ho; Lab Chip, 2002, 2, 179-187. High sensitivity PCR assay in plastic micro reactors. Jianing Yang, Yingjie Liu, Cory B. Rauch, Randall L. Stevens, Robin H. Liu, Ralf Lenigk and Piotr Grodzinski; Anal. Chem. 2005, 77, 1330-1337. High-Throughput Nanoliter Sample Introduction Microfluidic Chip-Based Flow Injection Analysis System with Gravity-Driven Flows. Wen-Bin Du, Qun Fang, Qiao-Hong He, and Zhao-Lun Fang; Science Vol 315 5 Jan. 2007, 81-84. Counting Low-Copy Number Proteins in a Single Cell. Bo Huang, Hongkai Wu, Devaki Bhaya, Arthur Grossman, Sebastien Granier, Brian K. Kobilka, Richard N. Zare; Nature Biotechnology Vol 22 (4), April 2004. A nanoliter-scale nucleic acid processor with parallel architecture. Hong J W, Studer V, Hang G, Anderson W F, and Quake S R; Electrophoresis 2002, 23, 1531-1536. A nanoliter rotary device for polymerase chain reaction. Jian Liu, Markus Enzelberger, and Stephen Quake; Biosensors and Bioelectronics 20 (2005) 1482-1490. Microchamber array based DNA quantification and specific sequence detection from a single copy via PCR in nanoliter volumes. Yasutaka Matsubara, Kagan Kerman, Masaaki Kobayashi, Shouhei Yamamura, Yasutaka Morita, Eiichi Tamiya; US Patent Application 2005/0019792 A1, “Microfluidic device and methods of using same”; and Nature Methods 3, 541-543 (2006) “Overview: methods and applications for droplet compartmentalization of biology” John H Leamon, Darren R Link, Michael Egholm & Jonathan M Rothberg.
Plugs offer advantages over cuvettes. For example, with plugs, the optical path may be made very thin, so if a bacterium is labeled (for example, with fluorescent antibodies) it is easy to detect. Alternatively, the same effect may be obtained by squeezing a blood sample between two coverslips separated by a 20 μm gap, which may be created by placing a thin metal (gold or Pt) wire between the cover slips. Thus, the cover slips may be covered with something to which bacteria stick, and factors that makes them grow biofilms or multiply aggressively. Probes may be added for detecting these activities or colonies.
Another advantage is diffusion control in a plug versus a coverslip or cuvette. For example, in a cuvette, a material produced by bacteria can diffuse away and become diluted. In a plug, the material produced by bacteria builds up to a high enough concentration that it is easy to detect. The material produced or the multiplying bacteria may be detected. Once the bacteria start growing, they grow until nutrients in the plug are depleted. Monitoring the decrease in a nutrient attributable to the presence of the bacteria provides evidence of the bacteria. For example, the decrease in O₂attributable to the presence of the bacteria may be done using agglutination beads.
Oxygen or other gases or mixtures of gases may need to be provided to encourage growth. Gases may be introduced by various methods including by dissolving them in the fluorocarbon carrier fluid. Care may be taken to avoid evaporation of oil/media particularly for incubation at 37° C. The plugs may be sealed in glass or placed in a Teflon capillary that is permeable to gases. The capillary is then placed in a vial of media that is the same (isotonic but not comprising indicators, cells and other components which will not partition across tubing). In short, the oxygen concentration of blood in any sepsis assay may need to be controlled in a way other than what is typically done with blood tests.
Many bacteria and bacterial components can directly activate individual coagulation factors. However, direct initiation of the coagulation cascade and the formation of a propagating clot is not typically observed when humans are infected. These bacterial components usually activate low levels of coagulation factors, but this activation does not result in the amplification and positive feedback necessary to form a clot that can grow and propagate. For example, Staphylococcus aureus (S. aureus) produces coagulase, a protein that binds prothrombin stoichiometrically and leads to cleavage of fibrinogen to fibrin. However, this conversion simply precipitates fibrin and does not result in production of thrombin, feedback, or amplification of the coagulation cascade. Escherichia coli (E. coli) that express the protein Curli are also known to activate coagulation factors, such as factor XII. This process was shown to cause slower initiation of coagulation due to depletion of factor XII. Bacteria do initiate coagulation in some organisms, such as horseshoe crabs, but this mechanism of controlling infection is believed to have been lost during evolution of vertebrates.
Crab blood is known to clot rapidly on contact with bacteria. Plugs with grown bacteria may be merged with plugs that contain crab blood. Clotting occurs rapidly when the crab blood contacts the bacteria, and secondary indicators such as fluorescent indicators or dyes for “crab thrombin” may be used to detect bacteria. Other clotting systems are disclosed in Kastrup et al. Acc Chem. Res. “Using chemistry and microfluidics to understand the spatial dynamics of complex biological networks.” 2008 April; 41(4):549-58.
Bacteria export materials out of the cell, for various reasons including to attack the host, to digest food, to fight, to signal, etc. These chemical messengers become more concentrated inside a plug than inside the original sample of blood. If the chemical messenger is an enzyme, a fluorogenic substrate for the enzyme would indicate the presence of the bacteria. Human blood has esterases, etc, that may interfere, but the use of enzymes specific to bacteria would allow detection of the bacteria. In fact, if one had a panel of 30 substrates, one may set up 30 tubes with 1000 plugs each, and looks for specific substrates lighting up (if there is substrate for one bacterium), or one may look for patterns of substrates lighting up if there is a more complicated relationship (each bacterium has a pattern of substrates associated with it).
Another type of molecule exported by bacteria are signaling molecules. If a single bacterium produces activators, these activators can accumulate in the plug, turning a single bacterium ON into the attack mode, and making it easier to detect.
Many types of cells react to high cell density by activating behaviors and specific genes through the process of quorum sensing. For instance, the human opportunistic pathogen Pseudomonas aeruginosa releases signaling molecules (homo serine lactones or other signals known as autoinducers) which accumulate in the space surrounding the cells and enable the cells to measure their local cell density. Many pathogens (such as Pseudomonas aeruginosa) activate virulence behaviors in response to the activation of quorum sensing. Because detection of cells in a virulent state is of interest, stochastic confinement would place cells at a high enough density that they should activate virulence mechanisms which may then be detected. In this way, stochastic confinement can detect cells with the potential for virulence, even if they currently have not yet activated virulence in the patient. Additionally, virulence activation often involves the upregulation of enzymes involved in infection, such as lipases, coagulases, and proteases that may be used as detection targets, specifically as detection targets of virulent species. The virulence enzymes and other released molecules may be used to activate detection mechanisms.
Another possibility is to lyse the cells in the media to detect something that is only produced by bacteria.
PCR may also be applied to plugs. PCR may be overwhelmed by the background DNA of human cells. In the case of plugs, if plugs are made small enough there would be only one cell per plug (human or bacterial) which eliminate background issues (if bacterial-specific primers are used background from other cells would not interfere). Binding of bacteria to human cells, or bacteria hiding inside of human cells would not be problematic due to the minimal background. Plugs may also be spiked to stimulate bacteria to make them easier to detect.
If bacteria can be detected in people at low concentration (before infection becomes a real threat), these methods may be used for general clinical practice to screen high risk patients including babies and the elderly for bacteria. Viral infections may also be monitored by this method by using vesicles that a virus would try to enter, and detecting the entry by fluorogenic substrates, or by the destruction of vesicles (detected, for example, by a Ca/Fluo4 system).
Methods of detecting bacteria using microfluidic based techniques comprise: 1) screening a sample against various reagents which results in a detectable signal in the presence of bacteria; and 2) screening these samples in parallel or in series against different drugs to determine which drug is best suited for killing the bacteria in the sample. Part 1) of this technique may include screening a sample against preloaded, plug-based cartridges that comprise bacteria-specific reagents. If the sample comprises bacteria that are specific for a reagent in one of the plugs a signal will be detected in that plug. Examples of the detection methods include, but are not limited to: i) magnetic based detection, ii) optical detection, and iii) oxygen detection. Part 2) of this technique may include preloading plug-based cartridges with various antibiotics that are known to kill certain types of bacteria. WO 05-056826 A1 which discloses methods is incorporated by reference herein in its entirety.
In some aspects of the present invention, the method of detecting bacteria comprises confinement of a bacterium into a small volume plug, wherein the confinement induces virulence activation of the bacteria through a quorum sensing type mechanism. The activation of virulence in the bacteria may induce the upregulation of various virulence factors such as proteins, enzymes, and small molecules. The upregulation and release of these virulence factors may be used as a detection target for the presence of virulent bacteria, may be used to detect a bacterium with the potential to activate virulence mechanisms, or may be used as targets to detect a specific species or type of bacteria. The upregulation and release of these virulence factors such as lipases, proteases, and coagulases may be incorporated as the initial step in an enzymatic detection cascade.
In some aspects of the method of detection of bacteria, the plug comprises a substance capable of lysing the bacteria. Lysis may be accomplished by standard detergent-based bacterial cell lysis. For example, frozen and thawed cell pellets are incubated with lysis buffer that is supplemented with lysozyme, which help disrupt cell walls (Lysozyme hydrolyzes β(1→4) linkages between N-acetylmuramic acid and N-acetyl-D-glucosamine residues in peptidoglycan and between N-acetyl-D-glucosamine residues in chitodextrin). Gram-negative bacteria may be hydrolyzed in the presence of EDTA that chelates metal ions in the outer bacterial membrane. Cells are incubated with lysis buffer for about 30 min, on ice. If the target is a nucleic acid, proteinase should be supplemented, whereas if the target is a protein, nucleases should be supplemented. To separate cell debris and insoluble protein (e.g., inclusion bodies), the sample is centrifuged (14,000 g, 30 min, 4° C.) and the supernatant collected. The supernatant comprises the soluble protein fraction, which can be further purified or directly analyzed, for example by SDS-PAGE.
Methods for Detecting Bacteria
In some aspects, the method for detecting bacteria comprises detecting bacteria in various samples using microfluidic based techniques, and screening bacteria using a hybrid and/or cartridge-based method to detect microorganisms by detecting their binding to beads.
Binding bacteria to magnetic beads may be accomplished by using specific (antibodies, chemical link between bacteria and bead) or non-specific (charge of bacteria, general molecule expressed on outside of bacteria) methods. Bound and unbound magnetic beads may be separated by microfluidic sorting techniques which rely on differences in diffusion based on size or changes of magnetic potential due to bound bacteria. Bound and unbound magnetic beads may also be separated using a magnetic field because migration in the field will be reduced for bound beads with increased drag. This method may be used to isolate specific types of bacteria from a sample or to remove all bacteria from other parts of the sample matrix. Liquid containing bacteria bound to beads may then be used to make plugs. Various techniques may be used to detect the presence of a magnetic bead in a plug, including measuring the electric current induced by a moving magnetic particle. Another detection method incorporates a hard-drive head to detect magnetic particles. Another detection method takes a relative measurement of the magnetic beads to measure relative orientation rate in the field. Another detection method relies on bacteria binding to multiple beads and the detection method being able to distinguish between single unbound beads and groups of bound beads (for example by changes in the amplitude of the spike in the detector).
Magnetic beads may also be used to detect proteins or other molecules by either chemically linking the target to the bead, or by first attaching an antibody to the target. The antibody then recognizes and attaches to the magnetic bead.
Optical detection schemes may also be used. Beads may have an optical signal, and a detection method which differentiates single unbound beads vs. groups of beads may be used to detect the target.
If the beads-based method provides means to count the number of detected objects, then bead detection schemes may be used to monitor the proliferation of bound objects after exposure to a compound, such as an antibiotic. The bead method may be used to isolate the bacteria from the sample and then introduce the antibiotic to the bacteria. Addition of more beads and enumeration of bound beads after incubation with the antibiotic may be used to determine whether or not the antibiotic inhibited proliferation of the bound bacteria.
Other detection schemes may be used to detect the bacteria themselves, such as oxygen or carbon dioxide detectors. Oxygen detection schemes include formation of Prussian blue as a function of oxygen presence, and fluorescence based oxygen sensors.
The sample may be detected attached to a bead, free in solution, or after deposition on the wall of the channel. Instead of a bead, bacteria may be agglutinated.
In some aspects, an array of pre-formed nanoliter sized droplets, or plugs is generated. Each plug comprises one or more beads. In each plug, all beads are substantially similar. Each bead has a specific binding affinity, provided by an antibody for example, for a microorganism or a subset of microorganisms, for example, bacteria, viruses, or fungi. The binding affinity can also be for a specific small particle, such as a pollen grain or a spore. The beads themselves are detectable, for example a magnetic bead or a fluorescent bead. Ideally, the binding event is also detectable.
If the binding event of one or more beads with a bacterium is directly detectable, the assay may be performed by the steps of injecting the sample into each of the plugs, incubating to allow the detection, and performing the detection. Direct detection may be accomplished by taking advantage of the difference in the magnitude and frequency of the signals produced by many unbound beads inside of a plug vs a microorganism that carries with it the same number of beads bound. Many unbound beads would generate several smaller signals, while an organism with the same number of beads bound would generate a single signal of higher amplitude.
If the binding event is not detectable, then the detection can be performed by separating bound beads from unbound beads, and detecting bound beads. The separation may be performed by a range of methods including a diffusive filter, a Brownian ratchet, or in the case of magnetic beads, a magnetic filter that applies a magnetic field to bias motion of the beads. Separation may require separating the plug fluid from the carrier fluid, flowing through the filter, and re-forming the plugs by adding carrier fluid to the plug fluid that passed the filter and contains predominantly bound beads. Detection is performed by a range of methods, including scanning the detector over the array of plugs, or flowing the array of plugs past the detector. Magnetic detection may be performed by detecting currents generated by moving magnetic beads, and may also incorporate technologies used in hard disk drives.
Spacer and index plugs may be used in the original array of plugs as disclosed in WO 08-079274 A1, the entirety of which is hereby incorporated by reference. Index plugs may contain markers detectable by the same method used to detect binding of bacteria to beads.
This method is discussed in the context of microorganisms and particles, but it is applicable to the detection of molecules and other objects by using a larger bead carrying antibodies against the molecule, and a set of detectable beads carrying another antibody against the same molecule. In the absence of the molecule, the detectable beads do not bind larger beads and remain dispersed. In the presence of the molecule, the detectable beads bind to the molecules and therefore to the larger beads, and become detectable by the methods described here.
To perform an antibiotic screen, plugs may contain antibiotics, and the plugs injected with the sample may be allowed to incubate to permit growth of microorganisms. Antibiotics are recognized and are substances which inhibit the growth of or kill microorganisms. Examples of antibiotics include, but are not limited to, chlorotetracycline, bacitracin, nystatin, streptomycin, polymicin, gramicidin, oxytetracyclin, chloramphenicol, rifampicin, cefsulodin, cefotiam, mefoxin, penicillin, tetracycline, chloramphenicol, minocycline, doxycycline, vancomycin, bacitracin, kanamycin, neomycin, gentamycin, erythromycin, cephalosporins, geldanamycin, and analogs thereof. Examples of cephalosporins include cephalothin, cephapirin, cefazolin, cephalexin, cephradine, cefadroxil, cefamandole, cefoxitin, cefaclor, cefuroxime, cefonicid, ceforanide, cefotaxime, moxalactam, ceftizoxime, ceftriaxone, and cefoperazone. Additional examples of antibiotics that may be used are in US 2007/0093894 A1, hereby incorporated by reference in its entirety. Detection of differences in growth and microbial populations in the absence and presence of each antibiotic would provide information on antibiotic susceptibility. First the bacteria in the sample are counted. Then, the bacteria sample is merged into the cartridge containing plugs of different growth media and different antibodies along with some as “blank” media and “blank” antibiotics plugs. Recount is preformed by merging with magnetic beads to see which one had bacteria reproducing.
In some aspects, the detectable signal is produced by the growth of the bacteria. Optical detection comprising optical methods for detection may use fluorescent nanoparticles instead of magnetic ones. One may also use control of surface chemistry provided by plugs to enhance growth of organisms. One method comprises merging the sample with a cartridge, the cartridge may optionally have growth tablets; monitoring bacterial growth by a change in oxygen concentration; and recording the readout. The readout may be but is not limited to a change in volume of oxygen bubbles the device, a change in optical signal due to the presence of an oxygen sensitive substrate (the substrate may be incorporated in many ways including in solution, on beads, or immobilized on a film that coats the device), or calorimetric reactions for detecting oxygen.
Bacteria detection by using agglutination may also be employed. This method comprises merging sample with plugs that comprise substance that induces agglutination in the presence of bacteria including for example, antibody labeled beads; forming agglutination screens on a large scale for bacteria detection; filling preformed plugs with beads covered with antibodies specific for different bacteria; and monitoring which plugs result in clumps of beads indicating presence of bacteria. Monitoring may be done by eye or some other detection technique. Secondary agglutination may also be used.
In addition, multiple assays may be performed in one device, for example by splitting a sample into a plurality of samples using techniques described above for splitting plugs.
One strategy is to change the hydrophobicity of the channel. Aqueous plugs with fluorinated oil as the carrier fluid will not wet hydrophobic channel walls. Therefore, to have a plug stick to the channel wall it is necessary to create a region of the channel in which the surface chemistry of the channel is hydrophilic.
A hydrophilic patch may be useful in 2 ways: i) a hydrophilic patch may capture a plug and hold it in place; and ii) a hydrophilic patch may temporarily come into contact with a plug as it is flowing by. This brief contact may result in a small portion of the plug fluid being deposited on the hydrophilic region of the channel wall. Capturing a plug (i) or depositing a portion of a plug onto the channel wall (i) may enable the assays to be run on the plug fluid.
Methods which require a surface such as ELISA, or an oxygen sensor incorporated into the surface, etc. may then be performed at the hydrophilic patch of channel wall. There are some papers which suggest that measurements performed on surfaces are more sensitive than measurements in bulk fluid. More generally, the concept of controlling channel wall surface chemistry can be implemented in the trafficking and measurement of plug fluids.
Another strategy for manipulating plugs is to control pressure drops in the channels by channel constriction. From Hagen-Poiseuille's law, if a channel gets narrower, the pressure drop quickly increases. If the pressure drop in one branch of a microfluidic network becomes narrow, plugs will not be able squeeze through the constriction because the capillary pressure required for the plug to squeeze through will be higher than the pressure drop over the device. In this way, channel constriction may be used to collect plugs and stop them from moving.
By designing a device with specified pressure drops in each channel in a network of branching channels, it may be possible to automatically sort the large number of plugs needed for some stochastic confinement applications. For example, a main channel carrying plugs splits into 10 different smaller channels. If each channel has a different diameter/length (i.e. pressure drops for each channel are different) this would create a bias in the loading of plugs. A system may be designed such that the 1st 100 plugs would be loaded into channel 1, then the pressure drop in channel 1 becomes greater than channel 2 due to the presence of many plugs, therefore the next 100 plugs would flow into channel 2, etc. This type of bulk sorting of plugs can be achieved simply by designing the channels with specific pressure drops and does not rely on turning on and off pumps, opening and closing valves, or other “active” plug sorting mechanisms.
Certain embodiments of the invention may be used to detect sepsis in 3 to 4 hours from a blood sample from a patient, and in other embodiments detection times may be reduced to 20 minutes or less. In a 5 to 10 mL blood sample from an infected patient, there might be 100 to 1,000 bacteria. Allowing the sample to culture overnight can increase those numbers by 10 to 100-fold. It would also be useful to know patterns of antibiotic resistance, and current methods are very tedious.
Possible means for detecting bacteria in a sample include stains and dyes, as used in flow cytometry, and which are well known in the art. Alternatively, one may look for a uniform change in color, appearance, scattering or optical density across a plug. almarBLUE™ (resazurin) is a fluorescent redox sensitive dye that can be used to detect living cells.
Certain embodiments of the invention may be used to detect different strains of bacteria, including Pseudomonas, Staphylococcus, E. Coli, etc.
When testing whole blood for bacteria one can use known methods to get rid of white blood cells, which would also be metabolically active, before testing for the presence of bacteria. For example, there are lysing agents well known in the art that selectively lyse eukaryotic cells.
A severely infected patient can have 10⁶bacteria (CFU)/mL in the blood. There are known methods for detecting 10²to 10³bacteria/mL, but only with culturing overnight. For example, the PCR-based LightCycler™ can detect 10³, but only can only detect bacteria with known, predetermined, target gene sequences, and it does not give any functional information, for example concerning antibiotic resistance, and it is relatively slow, taking 6 to 8 hours. Other methods, using 96 well plates have demonstrated the detection of 10⁶bacteria in a 200 μl sample in 2 to 3 hours, 10⁵bacteria in approximately two hours, and 10³bacteria in 8 to 9 hours. However, it is desirable to be able to detect low concentrations of clinically relevant bacteria in 3 to 4 hours or less.
Another means of detecting and typing cells is PCR amplification of 16S-23S rRNA., as described in Vliegen, I., et al., “Rapid identification of bacteria by real-time amplification and sequencing of the 16S rRNA gene” Journal of Microbiological Methods 66 (2006) 156-164, and patent application WO 96/119585, hereby incorporated by reference in its entirety. This can be accelerated by using a rapid microchip PCR method described recently that uses infrared light to achieve a 12 minute PCR reaction. See “On-chip pressure injection for integration of infrared-mediated DNA amplification with electrophoretic separation” Christopher J. Easley, a James M. Karlinseya and James P. Landers, Lab Chip, 2006, 6, 601-610.
Fluorogenic media, which change color in the presence of specific bacteria, can also be used to detect cells. Chromogenic media include, for example, Difco mEl agar, Merck/EMD Chromocult™ Coliform Agars, Chromocult™ Enterococci Agar/Broth, or Fluorocult® LMX Broth, BL ml agar, IDEXX Colilert, CPI ColiTag and Merck/EMD ReadyCult®. Typical enzyme substrates linked to chromogens or fluorogens include ONPG, CPRG, and MUG. These are also available in ready-to-use format, e.g. BBL ml agar and ‘convenience’ packs, e.g. IDEXX Colilert, CPI ColiTag and Merck/EMD ReadyCult®.
Bacteria can also be detected using simple growth and density measurements of plugs. Such measurements may aid detection and characterization of specific antibiotics that block the ability of the bacteria to grow, after combining plugs with specific antibiotics.
Microchannel PCR is described in U.S. Pat. No. 6,990,290, rapid bacterial PCR is described in U.S. Pat. No. 6,673,578, and a sepsis detection chip is described in US 2005-130185 A1, all incorporated by reference herein in their entirety.
Confining bacteria in small spaces might influence their phenotype, and gene regulation. For example, one bacterium inside of a small volume may respond as if it is in a culture with high cell density because communication molecules that it secretes, such as homoserine lactones, activate quorum sensing. This can be used to an advantage of this to decrease detection times or limits, and/or trigger virulence.
Staining
Typically to monitor bacterial growth, blood sample bottles are placed in an incubator and kept at body temperature. They are watched daily for signs of growth, including cloudiness or a color change in the broth, gas bubbles, or clumps of bacteria. When there is evidence of growth, the laboratory does a gram stain and a subculture. To do the gram stain, a drop of blood is removed from the bottle and placed on a microscope slide. The blood is allowed to dry and then is stained with purple and red stains and examined under the microscope. If bacteria are seen, the color of stain they picked up (purple or red), their shape (such as round or rectangular), and their size provide valuable clues as to what type of microorganism they are and what antibiotics might work best. To do the subculture, a drop of blood is placed on a culture plate, spread over the surface, and placed in an incubator.
Aggregation of a signal can be used to detect the presence of a single organism in a plug. With reference to FIG. 3, a single bacterium can simultaneously bind to many antibodies. If each of these antibodies has a signal, then the signal would be localized and therefore become detectable. For instance if each antibody was tagged with a fluorescent marker, then although single fluorescent markers may not be detected, the co-localized signal from many markers would be above the threshold for detection. Similarly, if each antibody was tagged with an enzyme, then many antibodies close together would create a high local concentration of the enzyme. Many enzymatic cascades (such as initiation of blood coagulation) require a threshold local concentration of enzyme. In this way, single antibodies would not be detected, but a local cluster of antibodies all bound to the same bacteria would create a detectable local concentration of enzymes. In a more specific example, if the bacteria sample was in blood or in a solution that contained the coagulation cascade, a detection method may involve antibodies which bind to the bacteria but also are tagged with the metalloprotease InhA (expressed by Bacillus species which induce blood clotting). A bacterium that is recognized by the antibody will bind to many antibodies and create a cluster of antibodies and therefore a localized cluster of InhA. The coagulation cascade will respond to the local cluster of InhA and initiate coagulation.
An additional effect of aggregation of a signal in the decrease in the background signal. For instance, if there are 100 molecules of signal in a 1 mL sample, then before clustering the background is 100 molecules/mL. After clustering 80 of the molecules together followed by stochastic confinement (assume in a 1 femtoliter space), the signal is now 80 molecules/fL and the background is now 20 molecules/mL because 80 molecules of signal have been removed from the bulk solution. In this way, aggregation of a signal increases the signal to noise ratio.
Another advantage of stochastic confinement is to segregate the signal/analyte from background material. For example, in the detection of bacterial strain “A” in a sample that contained numerous contaminant strain “C”, assays in the bulk may experience interference from the presence of strain “C”. This interference may be in the form of “C” influencing the behavior of “A”, or “C” influencing the components of the assay itself. Interference problems may be increased if the sample contains an abundance of “C” and few “A”. By forming plugs which separate species “A” from “C”, the assay may be run successfully for the presence of “A” as some plugs will contain only strain “A” and none of strain “C”. As long as the threshold is tuned properly, the amplification cascade can selectively respond to only the active, target particles even in the presence of a large excess of interfering particles.
In addition to stochastic confinement using plug-based microfluidics, there are other methods that may be used to achieve confinement of individual organisms or molecules as shown in FIG. 4. One such method is to generate an array of microwells. Microwells can be defined as a small compartment with volumes of nanoliters or less. The well is open on the top and the walls are impermeable to either the particles to be detected, molecules used in the detection system, or both. Once the wells are loaded, a top barrier may be placed over the wells in order to completely confine the particle(s) and molecules in the well. Even without complete confinement, a well with a single open wall will still experience some confinement effects due to decreased flux of particles and molecules into and out of the well. Another confinement method is to trap the particles in the matrix of a gel or polymer. Confinement effects will be achieved by reducing the diffusion of both the particles and the molecules used in the detection scheme. In this way, high local concentrations of molecules used for detection will accumulate around the particle. Adding an impermeable boundary to the bottom and/or top of the matrix (such as glass, plastic, etc.) will further increase confinement effects due to decreased flux of the accumulating signal away from the particle.
Stochastic confinement of individual bacteria into plugs of nanoliter volume or smaller volumes reduces detection time.
To reduce the time required to detect bacteria in a sample, a microfluidic device was designed to confine single bacterium into plugs of nanoliter volume. In principle, when generating plugs with a small volume from a solution with a low concentration of bacteria, much of the volume of the initial solution forms plugs that contain no bacteria. There are a few occupied plugs, each occupied by single bacterium. As a result, the concentration of bacteria in the occupied plugs is greater than the concentration of bacteria in the initial solution. For example, if plugs of nanoliter volume were made from a culture with an initial bacterial concentration of 10⁵CFU/mL, one in ten plugs would receive a single bacterium, as illustrated in FIG. 1 a. The concentration of cells in these occupied plugs would be one bacterium per nanoliter or 10⁶CFU/mL. In other words, 10⁵CFU/mL corresponds on average to 0.1 bacterium per 1 mL, and confining this solution into nanoliter plugs creates plugs with 0 bacteria per nanoliter plug and with 1 bacterium per nanoliter plug. CFU/mL refers to the colony forming units (CFU), a measure of live bacteria, per milliliter. Confinement effects are increased as the volume of plugs is decreased, therefore when confinement effects could be achieved in nanoliter plugs, picoliter plugs or femtoliter plugs would have increased confinement effects.
Targeting virulence factor for screening antibacterial drugs is a potential way to develop novel drugs. To escape cross-resistance of current drugs, targeting to virulence factors involved in human pathogenesis is considered essential. It has also been suggested that using surrogate infection model systems to screen novel drugs is a key issue for detecting and investigating the virulence factor.
Bacterial virulence may also be used to identify new therapeutics. Targeting virulence helps to preserve the many symbioses between the microorganism and host that contribute to human health. Targeting bacterial adhesion and toxin production and function are good approaches to developing the antivirulence drugs which will prevent the assembly of adhesive machinery or toxin expression or secretion. Quorum-sensing systems, two-component response systems, and biofilm formation can additional targets for virulence factors control.
The following methods are the ways in which pathogens cause disease in humans: adhesion, colonization, invasion, immune response inhibitors, and toxins. Since pathogenic bacteria have different methods of inducing virulence, conditions for inducing/monitoring/evaluating virulence are useful targets for drug discovery.
A representative example of inducing virulence is the generation of reactive oxygen species (ROS). Inflammatory cells have defense mechanisms against invading microorganisms and are known to exert their antimicrobial actions by releasing reactive oxygen species (ROS), proteolytic enzymes and other toxic metabolites. While these ROS and other toxic compounds damage pathogens and host cells together, antioxidant defense system such as superoxide dismutase (SOD), catalase, glutathione, and glutathione peroxidase also affect the host cells specifically.
Standard Method Vs Stochastic Confinement
The standard procedure for a patient presenting a bacterial infection in the blood, is to take a blood sample from the patient, the blood sample having a cell density of 100 CFU/mL. If the patient has an infection of methicillin resistant Staphylococcus aureus, and traditional culture based methods are used, detection of the Staphylococcus aureus takes about 12 hours. At that point the susceptibility of the pathogen to antibiotics would not yet be known; antibiotic testing would require another 6 hours.
If the blood sample is stochastically confined into 0.1 mL plugs, detection takes less than 4 hours. Because stochastic confinement has the ability to screen individual bacterium, up to 100 different antibiotic conditions may be screened from a 1 mL blood sample without a preincubation step. Detection time for bacteria with S. aureus decreases by 1.5 hours for every order of magnitude increase in cell density at time zero (i.e. 100 CFU/mL takes 12 hours, 1000 CFU/mL take 9.5 hours). Therefore with confinement in nL plugs, concentration can be increased from 100 CFU/mL to 10⁷CFU/mL which decreases detection time by 7.5 hours. (P. Kaltsas, S. Want and J. Cohen, Clin. Microbiol. Infect., 2005, 11, 109-114.) In 23-30% of all cases, an inappropriate antibiotic is initially administered (S. D. Carrigan, G. Scott and M. Tabrizian, Clin. Chem., 2004, 50, 1301-1314.)
In the standard method, the infection is identified after 12 hours and then it takes 6 additional hours to screen antibiotics. Many patients would not receive appropriate antibiotic treatment for over 18 hours, which means that the infection has worsened and the chance of mortality is greatly increased.
In the stochastic confinement method, infection and antibiotic sensitivity of infection is known after 4 hours or less. Antibiotic treatment can begin much sooner. Other advantages of stochastic confinement are the ability to run multiple tests on the sample without preincubation, therefore the clinician can have an in depth characterization of the pathogen (antibiotic sensitivity, serotype, strain, genetic information, other functional tests like propensity for virulence) within a few hours. Since confinement involves testing individual cells, heterogeneity in the activity/phenotype within the population of cells may be tested. For instance, it may be detected that 1% of bacteria in the sample are resistant to oxacillin even though 99% of the cells are sensitive. A traditional method might not detect this difference. Consider starting with a sample at 100 CFU/mL in which only 1% of cells are resistant. Detection time of resistant cells would be for a density of 1 CFU/mL. In the standard method, protocols for running the test might not be long enough (an additional 3 hours or more) to even detect this type of resistance in the standard method. In addition, subpopulations of resistant cells typically grow more slowly, increasing the chances of not detecting the cells in traditional tests.
Stochastic confinement is useful for large scale monitoring of resistant strains in the population and offers many benefits. Bacterial infections in the hospital setting do not routinely undergo in depth characterization of the infecting strain. Stochastic confinement provides a cheap and efficient method of characterizing pathogens so that hospital bacterial infections can be characterized routinely in the hope of better controlling them. In addition, functional tests are a better way to characterize pathogens than genetic tests because the genetic marker for a new resistance mechanism can be found immediately. Therefore there is less of a time delay between identification a new resistance mechanism and the discovery of a genetic marker useful for the diagnosis of the resistant strain. Stochastic confinement also enables increased tracking of resistance patterns leading to early recognition of resistant strains and enables health care agencies to track the spread of resistant strains.
Functional detection is possible as well wherein target or host cells and potentially infectious organisms are added into plugs and monitored for infection. Human/mammalian, bird, live stock, plants/crops can all be used as the target or host cells for these kinds of infectivity assays. The same techniques can be used to look for ways of reducing infectivity of microbes in any of these systems. The term virulence is used to mean direct infectivity, or killing remotely, any kind of pathogenicity, or negatively affecting the host cell in other ways or sporulating of bacterial spores, or activation of viral particles, or transition of bacterial cells from dormant to active form (relevant to tuberculosis). “Particle” refers to a cell, a viral particle, a spore, and the like.
The same ideas can be applied to general screening of cells and their activity such as switching from one state to another which may depend on the concentration of soluble or surface-bound factors, or presence of other cells. Examples include stem-cell differentiation and cancer cell activation.
The hybrid approach is described in Liang Li, Debarshi Mustafi, Qiang Fu, Valentina Tereshko, Delai L. Chen, Joshua D. Tice, and Rustem F. Ismagilov, “Nanoliter microfluidic hybrid method for simultaneous screening and optimization validated with crystallization of membrane proteins”, PNAS 2006 103: 19243-19248 as well as in publications incorporated by reference above.
An example of chemical systems capable of amplification is described in Kastrup et al. PNAS 2006 Oct. 24; 103(43):15747-52 as well as in publications incorporated by reference above. This simple chemical model system, built by using a modular approach, can be used to predict the spatiotemporal dynamics of complex chemical networks. Microfluidics is used to create in vitro environments that expose both the complex network and the model system to surfaces patterned with patches presenting stimuli. Such chemical model systems, implemented with microfluidics, may be used to predict spatiotemporal dynamics of complex biochemical networks.
Reducing virulence is a good strategy to fight microbial infections. Targeting virulence factors for screening antibacterial drugs is a potential way to develop novel drugs. To escape cross-resistance of current drugs, targeting to virulence factors involved in human pathogenesis is considered essential. To find compounds that reduce virulence, conditions need to be presented that induce virulence so that the virulence and the effects of drugs on virulence can be monitored. There are a number of ways that pathogens cause disease in humans including adhesion, colonization, invasion, immune response inhibition, and toxins. Since pathogenic bacteria have different ways of inducing virulence, conditions of inducing, monitoring, and evaluating virulence should be selected as the target.
Conditions that lead to virulence are often not known. Components may include the presence of host's cellular signals, accumulation of secreted microbial factors, presence of surface-bound cellular components and signals, presence of host cells, presence of molecular species associated the host's environment, presence of molecular species present on surfaces of host cells, reactive oxygen species, or reactive nitrogen species.
One representative example of inducing virulence is the generation of reactive oxygen species (ROS). Inflammatory cells show defense mechanism against invading microorganisms and are known to exert their antimicrobial actions by releasing reactive oxygen species (ROS), proteolytic enzymes and other toxic metabolites. While these ROS and other toxic compounds damage pathogens and host cells together, antioxidant defense system such as superoxide dismutase (SOD), catalase, glutathione, and glutathione peroxidase protect the host cells. Bacteria may detect the presence of ROS to control their virulence.
The bacterial detection method may be used to determine conditions that induce virulence using the hybrid method. Any of the factors that induce virulence may be introduced into plugs containing microbial cells and their concentration varied. In addition, surface chemistries may be used to incorporate molecules present on host's cell surfaces. These surface chemistries may also be screened in the context of the hybrid and/or cartridge method. A combination of several cartridges, each containing a different set of reagents (e.g. solution-based and surface-based) may be used to screen combinations of reagents at different concentrations. Activation of virulence may be observed as a function of solution and/or surface conditions, leading to determination of optimal conditions and/or conditions that are most physiologically relevant. These methods may be used in combination with confinement.
An array of host cells may be introduced using the hybrid and/or cartridge method to determine the kinds or types of cells that a microorganism may be virulent against. The hybrid and/or cartridge approach may be used to screen a wide range of microorganisms against a particular host to determine which microorganisms may be virulent against the host.
Confinement, alone or in combination with other factors, may be used to induce virulence. For example, confinement may lead to accumulation of secreted microbial factors turning on virulence. Virulence induced under confinement may be more physiologically relevant.
Using the methods described herein, microbial cells or particles isolated from a patient may be tested for their ability to induce virulence. Hybrid and/or cartridge methods may be especially attractive for such determination because they allow variation in the concentration of the inducing factors.
Co-confinement of microbial cells and host cells (human/mammalian or bird/live stock/plants/crops hosts) may further improve induction of virulence, as both microbial and host factors may accumulate within the confined volume. Such co-confinement may be created, for example, by either forming plugs from at least two streams (one containing a suspension of microbes and the other a suspension of target cells), or by creating plugs containing one type of cells or particles (e.g., microbial cells or viral particles), and injecting them with a suspension containing the other type of particle (e.g., target cells).
Changes in virulence may be monitored using a variety of methods including assaying for the secretion of surface expression of molecules/proteins/enzymes associated with virulence including lipases, virulence factors, iron siderophores, and the like. The assay may be tailored for the specific type of virulence mechanisms. For example, for adhesion, the ability of cells to stick to a host cell and detecting secretion of molecules known to promote adhesion to other cells is important. In colonization and invasion, disruption of cell membranes and promotion of endocytosis is important and thus, injection systems to inject protein/genetic material into another cell (for example via the type III secretion system) may be useful. For immune response inhibitors, molecules that bind to antibodies, formation of capsules around the cell, induction of fibrin formation to surround the bacterial cell and prevention of recognition by the host are important. For toxins, molecules are released that may be detected using fluorescence assays, agglutination assays, light producing reactions, color change reactions and the like. Also, production of certain factors associated with virulence (e.g. the lethal factor associated with B. anthracis virulence) may be detected, or functional tests of the effects on the target cell may be used.
In terms of ROS, P. aeruginosa induction of oxidative stress in the host cells may be assessed by measuring changes in lipid peroxidation and glutathione contents in the host cell line or tissues. In addition, the activity of the antioxidant systems can be evaluated by measuring activities of superoxide dismutase, catalase, and glutathione peroxidase. (Microbial Pathogenesis, Volume 32, Issue 1, January 2002, pages 27-34).
If conditions leading to activation of virulence are known, or for example once they are identified using methods described in this application, a screen may be conducted for agents that reduce virulence. These agents may be small molecules, proteins, antibodies, etc. There are numerous applications to humans, livestock, plants, and the like. The method is especially useful for infections which remain dormant for long periods of time and then cause problems periodically when virulence is induced such as TB, malaria, and herpes. Some P. vivax and P. ovale sporozoites do not immediately develop into exoerythrocytic-phase merozoites, but instead produce hypnozoites that remain dormant for periods ranging from several months (6-12 months is typical) to as long as three years. After a period of dormancy, they became reactivate and produce merozoites. Hypnozoites are responsible for long incubation and late relapses in these two species of malaria.
Drugs may be compounds or entities that alter, inhibit, activate, or otherwise affect biological or chemical events. For example, drugs may include, but are not limited to, anti-AIDS substances, anti-cancer substances, antibiotics, immunosuppressants, anti-viral substances, enzyme inhibitors, including but not limited to protease and reverse transcriptase inhibitors, fusion inhibitors, neurotoxins, opioids, hypnotics, anti-histamines, lubricants, tranquilizers, anti-convulsants, muscle relaxants and anti-Parkinson substances, anti-spasmodics and muscle contractants including channel blockers, miotics and anti-cholinergics, anti-glaucoma compounds, anti-parasite and/or anti-protozoal compounds, modulators of cell-extracellular matrix interactions including cell growth inhibitors and anti-adhesion molecules, vasodilating agents, inhibitors of DNA, RNA or protein synthesis, anti-hypertensives, analgesics, anti-pyretics, steroidal and non-steroidal anti-inflammatory agents, anti-angiogenic factors, anti-secretory factors, anticoagulants and/or antithrombotic agents, local anesthetics, ophthalmics, prostaglandins, anti-depressants, anti-psychotic substances, anti-emetics, and imaging agents.
Possible applications include screening for drugs that reduce virulence/infectivity of a strain of bacteria and testing bacterial isolates from patients for virulence. For example, a hybrid screen and/or cartridge system pre-loaded with several drugs may be injected into plugs at the same time as combining infectable host cells with bacteria. Preventing organisms from virulence is less likely to cause evolution of resistance than simply killing them. Importantly, confinement may change virulence. Confinement may induce virulence not observed in well-plates, so the method may have advantages over traditional well-plate methods.
The sensitivity of a bacterial strain to many antibiotics can be screened in a single experiment by using plug-based microfluidics. A single bacterial sample can be combined with many antibiotics to generate an antibiogram, or chart of drug susceptibilty. A pre-formed array of plugs of six antibiotics—two beta-lactams (ampicillin, AMP, and oxicillin, OXA); a cephalosporin (cefoxitin, CFX); a fluoroquinolone (levofloxicin, LVF); vancomycin, VCM; and a macrolide (erythromycin, ERT)—was generated by aspiration. Antibiotics were tested at the breakpoint concentration, (British Society for Antimicrobial Chemotherapy, BSAC Methods for Antimicrobial Susceptiblity Testing, 2007) the accepted concentration of antibiotic at which bacterial susceptibility is determined (FIG. 2 d). Since stochastic confinement of the bacterium into nanoliter-sized plugs generates many empty plugs, 50 plugs of each antibiotic were generated such that it was statistically likely that each condition would contain several plugs each occupied by a single bacterium. In total, 400-500 plugs were formed for each screen, which consisted of 6 drug conditions and 2 blank conditions. All 400-500 plugs were collected in the same coil of tubing. A blank condition was located at the beginning and end of the array to ensure that the position in the array did not affect assay results. The plugs in this antibiotic array were merged with Methicillin Resistant S. aureus (MRSA, ATCC#43300) at an initial cell density of 4×10⁵CFU/mL and the viability indicator on-chip to form plugs approximately 4 mL in volume, as illustrated in FIG. 2 a. The merged plugs were collected and incubated for 7 h at 37° C. After incubation, the fluorescence intensity of the plugs was measured.
Occupied plugs containing an antibiotic to which the bacterial strain was resistant showed increased fluorescence intensity, whereas plugs containing an antibiotic to which the bacterial strain was sensitive showed no significant increase in fluorescence intensity (FIG. 2 c). Plugs containing VCM were used as a negative control, because VCM inhibited this S. aureus strain in macro-scale experiments, in agreement with expectations (B. T. Tsuji, M. J. Rybak, C. M. Cheung, M. Amjad and G. W. Kaatz, Diagn. Microbiol. Infect. Dis., 2007, 58, 41-47). The average increase in fluorescence from all plugs containing VCM was used a baseline to which the increase in fluorescence intensity of all other plugs was compared (FIG. 2 b, Δ VCM). Four out of 49 (12%) control plugs with no antibiotic (FIG. 2 b, +blank 1) showed an increase in fluorescence intensity more than three times greater than the VCM baseline, indicating that they were occupied by bacteria. However, the other plugs with no antibiotic showed an increase in fluorescence intensity similar to the baseline, indicating that they were unoccupied (FIG. 2 b, +blank 1).
By comparing the fluorescence increase in each plug to the VCM baseline, it can determine which antibiotics were toxic to the bacteria. Plugs occupied with a viable bacterium showed an increase in fluorescence intensity greater than three times the VCM baseline. FIG. 2 c shows the average intensities of plugs that showed an increase in fluorescence intensity greater than 3 times the baseline (black bars) and plugs that showed an increase in fluorescence intensity less than 3 times the baseline (hatched bars). No plugs containing VCM or LVF had a fluorescence increase greater than 3 times the baseline, indicating that MRSA was sensitive to these antibiotics. Poisson statistics (Eq. 2) can be used to predict the probability of not loading a bacterium into any of plugs in the conditions LVF or VCM. In other words, Eq. 2 predicts the possibility of the LVF or VCM results being false-negative.
$\begin{matrix} f (k, λ) = \frac{λ^{k} e^{- λ}}{k!} & (2) \end{matrix}$
In Eq. 2, f is the probability of having k bacteria in a plug given an average bacterial loading of λ bacteria per plugs. The experimentally determined λ was 0.12, as 12% of control plugs with no antibiotics received bacteria (FIG. 2 b, +blank plugs). For k=0 and λ=0.12, we calculated the probability of having an unoccupied plug to be 0.887. The probability of having 49 unoccupied plugs is 0.887⁴⁹, or 0.0028. Given that LVF and VCM had at least 49 plugs, the probability of a false-negative due to loading is less than 0.3%.
The results from the MRSA antibiotic screen were used to make the antibiogram in FIG. 2 d. The antibiotics were tested at the breakpoint concentration, and the fluorescence data was used to determine if the bacterial strain was sensitive (S) or resistant (R) to the antibiotic. Sensitive means that no plugs containing a specific antibiotic showed an increase in fluorescence intensity greater than 3 times the VCM baseline. Resistant means that at least one plug containing a specific antibiotic showed increased fluorescence intensity greater than 3 times the VCM baseline. The susceptibility profile generated for MRSA by using the microfluidic screen was identical to the profile generated by using Mueller-Hinton agar plate tests and similar to previous reports in the literature for MRSA. (B. T. Tsuji, M. J. Rybak, C. M. Cheung, M. Amjad and G. W. Kaatz, Diagn. Microbiol. Infect. Dis., 2007, 58, 41-47) However, antibiotic sensitivity testing is influenced by many factors, including bacterial load, culturing conditions, temperature, bacterial strain, and type of assay used to detect sensitivity. In addition, a cell population may contain sub-populations of cells with variable sensitivity to a given antibiotic. All of these factors should be considered and further characterized before formulating guidelines for implementing plug-based antibiotic sensitivity assays.
Plug-based methods can also be used to determine the minimal inhibitory concentration (MIC) of an antibiotic against a bacterial sample.
Next, this microfluidic approach was used to determine the MIC of the antibiotic cefoxitin (CFX) for MRSA and MSSA (FIG. 6). This assay was similar to the antibiotic screening assay described above, except that the pre-formed array of antibiotic plugs all contained the same antibiotic and the concentration of that antibiotic in each plug of the pre-formed array was different. Again, plugs containing saline solution were included at the beginning and end of the array to serve as negative controls and to ensure that the first and last plugs of the array gave similar assay results. The positive control plugs consisted of CFX at a concentration of 24 mg/L, as both strains are known to be inhibited by CFX at this concentration. Plugs of the antibiotic array were merged with bacteria and the fluorescent viability indicator as illustrated in FIG. 6 a and incubated at 32° C. Plugs with MRSA were incubated for 6.75 h and plugs with MSSA were incubated for 6.5 h. It should be noted that temperature can affect the results of antibiotic sensitivity assays. Here, the difference in MIC of MRSA and MSSA was discerned by assays conducted at 32° C.
After incubation, the fluorescence intensity of the plugs was measured. Here, the average increase in fluorescence intensity of plugs containing 24 mg/L CFX was used as the baseline to which the increase in fluorescence intensity of other plugs was compared. Because MRSA is resistant to many beta-lactam antibiotics, CFX should be less effective against the strain MRSA. As expected, the MIC of CFX was higher for MRSA (<8 mg/L) than the MIC of CFX for MSSA (<4.0 mg/L) (FIG. 6 b). These results validate the use of this plug-based technology for screening both the susceptibility and the minimal inhibitory concentration of many antibiotics against a single bacterial sample.
The issue of screening media conditions along with drugs is of interest. The outcome of a drug screen can be heavily influenced by the media and culturing conditions used for the screen. To eliminate false negatives in a drug screen, it would be useful to screen the same drug condition in a variety of media and culturing conditions. For instance, an unknown sample of bacteria may be screened against the drug oxicillin at 30 and 37° C., 100 and 150 mM NaCl, and in Luria Bertani media and soy trypticase media. The media may: 1) influence the growth rate of bacteria regardless of drug condition (if the media condition is such that bacteria grow very slowly, the assay would falsely determine that the drug is killing the bacteria); or 2) influence the interaction between the bacteria and the drug.
Stochastic confinement combined with plug-based microfluidic handling methods accelerates bacterial detection and enables rapid functional antibiotic screening. By using this method, assays may be performed on a single bacterium, potentially eliminating the need for pre-incubation. By confining and analyzing single bacterium in plugs, detection time is now determined by plug volume. We were able to achieve detailed functional characterization of a bacterial sample in less than 7 hours. We also demonstrated that a bacterium in a 1 mL plug may be detected in as little as 2 hours. The detection time is limited by the formation and measurement of plugs of small volume and is less dependent on the initial concentration and growth rate of bacteria in the sample. This feature may be potentially important for accelerated detection of slowly-growing species such as M. tuberculosis, a pathogen of significant importance world-wide. (E. Keeler, M. D. Perkins, P. Small, C. Hanson, S. Reed, J. Cunningham, J. E. Aledort, L. Hillborne, M. E. Rafael, F. Girosi and C. Dye, Nature, 2006, 444 Suppl 1, 49-57) Here, we have demonstrated a screen with 400-500 mL plugs. High-throughput screens with more conditions and increased concentration of the sample would require methods that can handle larger numbers of smaller plugs, including methods for automated sorting and analysis. (Y. C. Tan, Y. L. Ho and A. P. Lee, Microfluid. Nanofluid., 2008, 4, 343-348; D. Huh, J. H. Bahng, Y. B. Ling, H. H. Wei, O. D. Kripfgans, J. B. Fowlkes, J. B. Grotberg and S. Takayama, Anal. Chem., 2007, 79, 1369-1376; K. Ahn, C. Kerbage, T. P. Hunt, R. M. Westervelt, D. R. Link and D. A. Weitz, Appl. Phys. Lett., 2006, 88, 024104; M. Chabert and J.-L. Viovy, Proceedings of the National Academy of Sciences, 2008, 105, 3191-3196) Upon incorporating such methods for handling and sorting large numbers of plugs of small volume, this technique may be used for the detection of bacteria in a sample at a cell density much lower than 10⁵CFU/mL. Since the activity of single cells is being measured, it is conceivable that detecting the presence of even a single bacterium in a sample may be feasible.
Given that a typical 5 mL blood sample from a patient with bacteremia contains a cell density of 100 CFU/mL, (L. G. Reimer, M. L. Wilson and M. P. Weinstein, Clin. Microbiol. Rev., 1997, 10, 444-7) this method is capable of performing dozens of functional tests on such a sample. Patient-specific characterization of bacterial species would not only lead to more rapid and effective treatment, but such an advance would also enable in-depth characterization of bacterial infections at the population level. Such detailed characterization may aid in tracking and identifying new resistance patterns in bacterial pathogens. (S. K. Fridkin, J. R. Edwards, F. C. Tenover, R. P. Gaynes and J. E. McGowan, Clin. Infect. Dis., 2001, 33, 324-329; R. T. Horvat, N. E. Klutman, M. K. Lacy, D. Grauer and M. Wilson, J. Clin. Microbiol., 2003, 41, 4611-4616) The principles of these methods, stochastic single-cell confinement and multiple functional assays without sample pre-incubation, may also be applied to other areas, including performing functional tests on field samples, detecting contamination of food or water, separating and testing samples with mixtures of species, measuring functional heterogeneity in bacterial populations, and monitoring industrial bioprocesses.
Other Applications
Confinement affects may be of particular use in the detection of slow growing bacterial strains such as Mycobacterium tuberculosis (the strain which causes tuberculosis, TB). To grow a culture of M. tuberculosis using traditional methods takes several weeks. Current tests for TB are based on an immune response, but cannot distinguish between someone with an active infection and someone who has previous immunization. Other current tests involve PCR based methods. Stochastic confinement should increase the sensitivity of detection based assays and also remove the need for a lengthy preincubation step before running the detection assay since detection may be done with a single confined cell.
Confinement effects may be used to screen natural sources for candidate organisms or their genes that perform functions of interest or generate molecules of interest. Functions of interest include nitrogen fixation, carbon recycling, hydrocarbon production, pollutant degradation, solar energy conversion, forming a symbiotic relationship with other microbes, producing a toxin that kills other organisms, produces light, produces an odor, generates electricity. Molecules of interest include drug candidates, small molecule inhibitors, enzymes which degrade cellulose, enzymes which degrade pollutants, adhesives, electron transport molecules, metal chelators, selective inhibitors of small molecules, catalysts, plasticizing agents, and proteases.
Some of these functions may not occur in large volume (low density of cell type which performs function) samples or samples with mixtures of microbes. One advantage of confinement is that individual microbes will be able to function under high density conditions. This would be useful for rare cell types in a sample, as high density functions would not be occurring in the original sample due to the rare cells being at a low cell density. Such functions would not be occurring in a macroscale sample. Confinement may also enable rare cell types of initiate growth or increase the growth rate due to the high concentration of rare cells in a small volume plug after confinement. This type of approach can be extended to rare cell types from mixed species samples by confining rare cell at high densities to enable more rapid growth of the rare cells due to both increased inoculation density and eliminating competition with other strains in the original mixed sample. The effect of confinement in small volumes is increased if the molecules accumulating in the plug are immiscible in the carrier fluid (the fluid surrounding the plug). Using immiscible fluids around the plug will prevent released microbial products from diffusing out of the plug and hence the released products will more quickly accumulate and achieve a higher concentration in the plug.
Other applications include detecting bacteria for applications in homeland security and safety of the food chain and water. It is also possible to apply these methods of detection to the areas of sepsis, bioenergy, proteins, enzyme engineering, blood clotting, biodefense, food safety, safety of water supply, and environmental remediation.
The following patents and patent applications are hereby entirely incorporated by reference in their entirety: WO 05-010169A2, U.S. Pat. No. 6,500,617, WO 2007-009082 A1.
A process of collecting a useful product from stochastically confined cells may comprise: confining organisms, cells or particles (by themselves, or with their enemies such as other bacterial or other cells), or with addition of stimulating chemicals; accumulating their products (antibiotics or other potentially valuable enzymes); using these products for detection; using these products for further screening (for example, dilute and different concentrations and merging with suspensions of other bacteria to see if those bacteria get killed off, or drip those dilutions into standard growth assay plates); using the accumulated enzyme and assaying for function (as described in this application, including but not limited to cellulose degradation, catalysts for synthesis, disruption of biofilms); and adding other assays. When plugs are used for stochastic confinement, allowing an organism to multiply inside a plug and then splitting the plug into daughter plugs (where at least two daughter plugs contain daughter organisms) provides an opportunity to perform multiple assays on clones of the organism (including assays that cannot be performed on a single organism or in a single volume).
All of the methods and applications described herein may be done under controlled atmosphere using plugs/droplets, where fluorocarbon can enable transport of gases to the plug, and massively parallel small-scale incubations under controlled atmosphere can be performed. This may be useful for control of virulence, for hydrogen generation and for using organisms that require controlled atmosphere (anaerobes, organisms that consume or produce methane or other hydrocarbons or H₂S, etc).
The methods described herein may also be used to detect fungi, archaea and other organisms in a sample.
Test Strips
General Components of Test-Strips
A test strip comprises an amplification layer and may comprise one or more of the following layers: a filtration wetting layer with selectivity, a detection layer with threshold and a layer of substrate for signal output. A signal produced by target bacteria, for example an enzyme, will turn on the amplification reactions in the amplification layer. Multiple amplification layers may be applied to achieve a high magnitude of amplification. A substrate is used to detect the generation of control molecule or the output of the amplification in the system.
The amplification region may be regulated by a threshold mechanism. The methods for amplification may be chosen from those described in the section entitled “Amplification.” A detailed definition of threshold may be found below.
The techniques described in this section aim to detect a small number of molecules or particles in a short time with a high resistance to noise and background signals. To achieve such goals, each of these techniques consists of multiple modules. The two most important modules are the amplification process with positive feedback that produces a large amount of substances in short time and the inhibitory mechanism. The interplay between these two processes sets up a threshold or a threshold-like behavior. In an ideal system, a threshold is the concentration below which an input gives a background output and above which an input gives a the signal output, where the two outputs are easily distinguishable. To achieve most useful amplification, the signal output must be significantly (often by two orders of magnitude or more) different than the background output. The transfer function, the function of output versus input, may be a shifted ideal step function. However, in many cases, it is impossible to achieve an ideal threshold, but possible to achieve a threshold-like behavior, with which the transfer function is similar to a step function but has a finite slope at the transition region. A sigmoidal function or a similar function may be used to describe such threshold-like behaviors. Another way to look at threshold is the time to reach maximum possible output as a function of input. With an ideal threshold, this function is infinite when the input is below the threshold and reaches a constant small positive value when the input is above the threshold. With a threshold-like behavior, this function is very large when the input is below the threshold. As the amount of input increases from the threshold, this function decreases rapidly and reaches a very small value (such as 10% or less of that at the threshold). In the most ideal case, at threshold, a change in the number of input molecules of 1 unit leads to a drastic change in the output.
As long as the threshold is tuned properly, the amplification process may selectively respond to only the active, target particles (molecules) even in the presence of a large excess of interfering particles (molecules). One or multiple amplification layers with threshold may be added to increase the degree of amplification. Threshold response may be incorporated in the detection layer to limit false positives and false negatives. Diffusion of signal molecules and control molecules may be restricted on each layer by choosing the appropriate material. Thus, stochastic confinement may be applicable to test strips. For example, a bacterium on the strip is confined by limited diffusion (the reagents and products are not mixed on purpose), or the test strip may be structured to restrict diffusion, for example when based on alumina or track-etched membranes.
FIG. 8 is a schematic description of a test strip with an amplification system. Bacteria are brought into contact with a filtration wetting layer. Enzymes (E) produced by target bacteria, for example in the scale of pM or even lower, enter a detection layer with threshold and turn on the reactions to generate control molecules (C). After proceeding to amplification layer, the signal will be amplified and the concentration of control molecule increased to μM or mM. The control molecules will react with chromogenic, fluorogenic or other substrate in the substrate layer to give a strong output signal. Output signal can also be generated in any other ways.
A timer region may be added to the test strip as shown in FIG. 9. This timer region is an analogous reaction which is not detecting the analyte, but instead demonstrating the reaction is running correctly under the current conditions (age of strip, temperature, pressure, humidity, presence of certain impurities, etc.). It also gives the user the assurance that the strip is working and that they have waited long enough for the results. There may be one or multiple timer regions to indicate different parameters (age of strip, temperature, pressure, humidity, presence of certain impurities, etc.)
The timer region may have the same amplification system with reactions of the same sensitivity as the detection region. It may be used to test for false positive. However, the timer regions may utilize other techniques for specific purposes. Some amplification systems may be previously loaded with a known amount of analyte which may be activated upon the beginning of the test (e.g. by wetting).
Amplification
Amplification schemes using enzyme cascades are known. In particular, enzyme cascades may be modified to detect the type of molecule not naturally associated with the enzymes. For example, the crab blood cascade may be used to detect air pollutants by modifying enzymes at the beginning of the cascade to detect a new type of input.
Cascade assay formats may be used. These detect an analyte through a process wherein, a first signal generating compound (i.e., SGC #1) produces a product that may be utilized by a second SGC #2 to produce a product which, e.g., may be utilized by a third SGC #3. The subject cascade of products from SGC #1-3 results in amplification which results in a greater overall signal than may be achieved by any single SGC.
Many substances could be detected by a scheme which uses the same (generic) amplification mechanism. For different analyte, a different starter reaction is designed, but all starter reactions produce the same output. For example, a detection scheme with the first step which is designed to detect a specific activity and the second steps which is able to amplify the product of the first step. An example is using the coagulation cascade as the generic amplification mechanism, and a set of started reactions that all produce an activator of thrombin. One would only have to redesign the starter reaction to detect a new analyte
Alternatively, selection may be performed indirectly by coupling a first reaction to subsequent reactions that take place in the same plug. There are two general ways in which this may be performed. In the first method, the product of the first reaction is reacted with, or bound by, a molecule which does not react with the substrate of the first reaction. In a second, the coupled reaction will only proceed in the presence of the product of the first reaction. For example, a genetic element encoding a gene product with a desired activity may then be purified by using the properties of the product of the second reaction to induce a change in the detectable properties of the genetic element.
Alternatively, the product of the reaction being selected may be the substrate or cofactor for a second enzyme-catalyzed reaction. The enzyme to catalyze the second reaction may either be translated in situ in the plug or incorporated in the reaction mixture prior to incorporation into a plug. Only when the first reaction proceeds will the coupled enzyme generate a product which may be used to induce a change in the detectable properties of the genetic element. Stochastic confinement could be used to increase local concentration and signal to noise ratio and to give advantages described elsewhere herein. The product of the first reaction could be a substrate, enzyme, or cofactor of the second reaction, or promote the release of inhibition of the second reaction. More reaction steps could be used.
The concept of coupling may be elaborated to incorporate multiple enzymes, each using as a substrate which is the product of the previous reaction. This allows for selection of enzymes that will not react with an immobilized substrate. It may also be designed to give increased sensitivity by signal amplification if a product of one reaction is a catalyst or a cofactor for a second reaction or series of reactions leading to a selectable product. Furthermore an enzyme cascade system may be based for the production of an activator for an enzyme or the destruction of an enzyme inhibitor. Coupling also has the advantage that a common selection system may be used for a whole group of enzymes which generate the same product and allows for the selection of complicated chemical transformations that cannot be performed in a single step.
Previously developed methods of detection include those using nucleotide-based amplification (such as immuno-polymerase chain reaction (iPCR), (Adler, M.; Wacker, R.; Niemeyer, C. M., Sensitivity by combination: immuno-PCR and related technologies. Analyst 2008, 133, (6), 702-718.) amplification based on allosteric catalysis, (Zhu, L.; Anslyn, E. V., Signal amplification by allosteric catalysis. Angewandte Chemie-International Edition 2006, 45, (8), 1190-1196.) biobarcode, (Nam, J. M.; Stoeva, S. I.; Mirkin, C. A., Bio-bar-code-based DNA detection with PCR-like sensitivity. Journal of the American Chemical Society 2004, 126, (19), 5932-5933; Nam, J. M.; Thaxton, C. S.; Mirkin, C. A., Nanoparticle-based bio-bar codes for the ultrasensitive detection of proteins. Science 2003, 301, (5641), 1884-1886.) molecular beacon, (Li, J. W. J.; Chu, Y. Z.; Lee, B. Y. H.; Xie, X. L. S., Enzymatic signal amplification of molecular beacons for sensitive DNA detection. Nucleic Acids Research 2008, 36, (6).) liposome-based amplification, (Edwards, K. A.; Baeumner, A. J., Liposomes in analyses. Talanta 2006, 68, (5), 1421-1431.) and nanowire sensor. (Zheng, G. F.; Patolsky, F.; Cui, Y.; Wang, W. U.; Lieber, C. M., Multiplexed electrical detection of cancer markers with nanowire sensor arrays. Nature Biotechnology 2005, 23, (10), 1294-1301; Cui, Y.; Wei, Q. Q.; Park, H. K.; Lieber, C. M., Nanowire nanosensors for highly sensitive and selective detection of biological and chemical species. Science 2001, 293, (5533), 1289-1292.) Each of these techniques has one or multiple disadvantages: slow response, vulnerability to noise and environmental degradation, low detection limit, and requirement for highly sophisticated and expensive fabrication.
The techniques described in this section aim to detect a small number of molecules or particles in a short time with a high resistance to noise and background signals. To achieve such goals, each of these techniques consists of multiple modules. The two most important modules are the amplification process with positive feedback that produces a large amount of substances in short time and the inhibitory mechanism. The interplay between these two processes sets up a threshold or a threshold-like behavior. In an ideal system, a threshold is the concentration below which an input gives a background output and above which an input gives a the signal output, where the two outputs are easily distinguishable. To achieve most useful amplification, the signal output must be significantly (often by two orders of magnitude or more) different than the background output. The transfer function, the function of output versus input, may be a shifted ideal step function. However, in many cases, it is impossible to achieve an ideal threshold, but possible to achieve a threshold-like behavior, with which the transfer function is similar to a step function but has a finite slope at the transition region. A sigmoidal function or a similar function may be used to describe such threshold-like behaviors. Another way to look at threshold is the time to reach maximum possible output as a function of input. With an ideal threshold, this function is infinite when the input is below the threshold and reaches a constant small positive value when the input is above the threshold. With a threshold-like behavior, this function is very large when the input is below the threshold. As the amount of input increases from the threshold, this function decreases rapidly and reaches a very small value (such as 10% or less of that at the threshold). In the most ideal case, at threshold, a change in the number of input molecules of 1 unit leads to a drastic change in the output.
An elementary amplification method for detecting molecules includes: a) a set of molecules or materials that constitute an amplification process that is capable of amplifying one or multiple components (termed output); b) a set of molecules or materials that provides an inhibitory mechanism to stop or slow down amplification by the amplification process described in a) when the amount of the analyte is insufficient; c) an activating mechanism which, once sufficient analyte (termed input) is present, triggers the amplification process described in a); d) a readout process to provide an apparent signal (with visual signal as one example).
Inhibition in part (b) and the amplification process in part (a) set up a threshold or threshold-like behavior which gives very little or no output if the input is below the threshold and gives fast and abundant output if the input is above the threshold. This case is termed positive contrast.
Alternatively, the system may be set up with the input providing a mechanism to inhibit the amplification process. In such cases, with a below-threshold amount of input, a lot of the output is produced, while an above-threshold amount of input would give little or no output. This phenomenon is called negative contrast. The roles of part (b) and part (c) are swapped. The general feature is a big contrast (be it positive or negative) between the output of a below- or above-threshold input.
Elementary amplification methods may be coupled in such a way that outputs of one may be used as inputs for another. This is also called a cascade. The number of elementary amplification methods in a cascade may be varied depending on how much more amplification is needed in comparison to what is provided by each elementary amplification method.
The analyte may be any molecules such as enzymes, DNA, RNA, small molecules, or any other molecules. It may come from any sources including bacterial components, human fluids, water samples, etc.
The amplification processes may be any reaction that is autocatalytic or a reaction network with one or more positive feedback loops. The processes may involve enzymes such as those in the blood clotting cascade, apoptosis, and Limulus amebocyte (horseshoe crab) lysate (LAL), or any other enzymes. The processes may involve inorganic chemicals such as the Co(III)-Co(II)-oxone system, (Endo, M.; Abe, S.; Deguchi, Y.; Yotsuyanagi, T., Kinetic determination of trace cobalt(II) by visual autocatalytic indication. Talanta 1998, 47, (2), 349-353; Endo, M.; Ishihara, M.; Yotsuyanagi, T., Autocatalytic decomposition of cobalt complexes as an indicator system for the determination of trace amounts of cobalt and effectors. Analyst 1996, 121, (4), 391-394; Tsukada, S.; Miki, H.; Lin, J. M.; Suzuki, T.; Yamada, M., Chemiluminescence from fluorescent organic compounds induced by cobalt(II) catalyzed decomposition of peroxomonosulfate. Analytica Chimica Acta 1998, 371, (2-3), 163-170.) Ag⁺-Ag(0) system, systems in which a metal surface catalyzes reduction of Ag+ or other ions, producing more metal surface available for catalysis the chlorite-iodide system, (Dateo, C. E.; Orban, M.; Dekepper, P.; Epstein, I. R., J. Am. Chem. Soc. 1982, 104, 2, 504-509) or thiosulfate-chlorite-hydronium system. (Runyon, M. K.; Johnson-Kerner, B. L.; Ismagilov, R. F., Minimal functional model of hemostasis in a biomimetic microfluidic system. Angewandte Chemie-International Edition 2004, 43, (12), 1531-1536; Horvath, A. K.; Nagypal, I.; Epstein, I. R., Kinetics and mechanism of the chlorine dioxide-tetrathionate reaction. Journal of Physical Chemistry A 2003, 107, (47), 10063-10068; Horvath, A. K.; Nagypal, I.; Epstein, I. R., Oscillatory photochemical decomposition of tetrathionate ion. Journal of the American Chemical Society 2002, 124, (37), 10956-10957; Nagypal, I.; Epstein, I. R., Systematic Design of Chemical Oscillators 0.37. Fluctuations and Stirring Rate Effects in the Chlorite Thiosulfate Reaction. Journal of Physical Chemistry 1986, 90, (23), 6285-6292.). They may also involve organic reactions, such as those with acid as autocatalysts. (Ichimura, K., Nonlinear organic reactions to proliferate acidic and basic molecules and their applications. Chemical Record 2002, 2, (1), 46-55.) They may also be combinations of different types of chemicals.
Phenomena in which nucleation is involved may also be used as an amplification process, in which a crystal of aggregate produced by nucleation may serve as a nucleus to promote more production of such crystal or aggregate.
An amplification process may also be achieved by using materials that release substances that catalyze the release of more of such substances.
There are many inhibitory mechanisms. Each mechanism may occur directly or indirectly through more than one reaction or processes.
The first mechanism may use chemical inhibitors. These chemical inhibitors may be stoichiometric, such as those that bind to enzymes and block the active sites or ligands that bind to and sequester metal cations. The chemical inhibitors may also be catalytic, such an enzyme that cleaves an active enzyme that is important in positive feedback loops.
The second mechanism may use materials to mechanically separate important components of the amplification process. These materials include, but are not limited to, vesicles containing liquid, particles of solid or gel, and any combination of single-layer or multi-layer particles or vesicles of one type or multiple types.
There are many activating mechanisms. For example, the input may be one or multiple types of the substances that are inhibited chemically or mechanically by materials. The input may directly or indirectly produce more of the substances that are inhibited chemically or mechanically in the inhibitory mechanism. The input may directly or indirectly interfere with the chemical or mechanical inhibition by competition with the substances being inhibited. The input may directly or indirectly interfere with the chemical or mechanical inhibition by chemically modifying the chemical inhibitors or the materials used to mechanically separate the components of the amplification process.
The input generates a high local concentration of substances that may have functions described above. Stochastic confinement is an example of this category.
With such flexibility of activating mechanism, the methods may be designed to be applicable to many kinds of analytes with different desired degrees of specificity. For example, if an existing method needs modifying to be applicable for detection of a different analyte, the activating mechanism may be changed completely or may be adapted with single of multiple steps to use the analyte of interest to promote the production of the analyte of the existing method.

Readout Processes

One or more substances that are amplified or activated by the amplification process may promote the production of some form of easily detectable readout. This readout may be a visual signal based on color. The color may come from any reaction that can generate color when the input (over the threshold) is present. For example, an enzyme amplified in the amplification process cleaves a fluorogenic substrate or a chromogenic substrate to give a fluorescent signal or color. Some other examples are pH indicators, reduction-oxidation potential indicators, and indicators for specific cations.
Alternatively, the readout may be a visual signal based on production of aggregates (precipitates) or crystals from any method when the input (over the threshold) is present. These aggregates (precipitates) or crystals may be the result of any process such as chemical reaction or production of paramagnetic substances that come together as solid in a magnetic field.
Examples of reactions that produce paramagnetic solids that may be used in an amplification process include: 1) Ag⁺->Ag:Ag+ has electron configuration of d¹⁰, so it is diamagnetic. Ag has electron configuration of s¹d¹⁰, so it is paramagnetic. The autocatalyst is Ag (or its surface to be more detailed);
2) Guyard reaction: Mn(VII) (such as, KMnO₄) reacts with Mn(II) (some soluble salt such as nitrate or chloride) and makes MnO₂which is a black paramagnetic powder. The autocatalyst is MnO₂solid (its surface). (Polissar, M. J. Journal of Physical Chemistry 1935, 39, 1057.)
Further still, the readout may be any other detectable signal, such as an electrical signal that gives analog or digital readout or any other kind.
Amplification Process Involving Enzymes with Chemical Amplification Processes
The amplification process may be an autocatalytic reaction or reaction network that has a positive feedback. This positive feedback may be achieved through one step (FIG. 10). In such case, an autocatalytic enzyme may catalyze the cleavage of its precursor. (A chemical inhibitor may also used to inhibit this autocatalytic enzyme, and a fluorogenic substrate may be used to detect this autocatalytic enzyme). Positive feedback may also be achieved through two steps (FIG. 11). In such case, enzyme 1 may catalyze the cleavage of enzyme precursor 2 to produce enzyme 2, which may catalyze the cleavage of enzyme precursor 1 to produce enzyme 1. (Common or specific inhibitors and fluorogenic substrates for enzyme 1 and enzyme 2 may be included). Generally, positive feedback may be achieved through any number of steps, in which one substance, called substance 1, may catalyze directly the production of itself or indirectly by catalyzing the production of another substance that may directly or indirectly catalyze the production of the substance 1. Many positive feedback loops may also be coupled with each other into a cascade, in which the output of one positive feedback loop is used as the input of another (FIG. 12). In such case, the number of positive feedback loops with threshold must be at least 1 and may vary from 1 to all.
In certain embodiments at least one enzyme in the positive feedback loop is inhibited by an inhibitor. Not every enzyme is necessarily inhibited, but many may be. The inhibitors may be those found naturally or synthesized. Antibodies may also be used as inhibitors.
If there is only one enzyme involved in the positive feedback loops (FIG. 10), the input may be that enzyme. If there are multiple enzymes involved in the positive feedback loops (FIG. 11), the input may be single ones or combinations of the enzymes. The input may produce single or combinations of enzymes from the same or different precursors as in the amplification process, through one or many steps. The input may also be much more strongly binding substrate for the inhibitor, or promotes the production of much more strongly binding substrate for the inhibitor. The input may chemically alter the inhibitors and disable the inhibitory effect, or promote the production of substances with such function.
If fluorogenic substrates are used, one or multiple fluorogenic substrates may be cleaved by one or multiple enzymes in the amplification process, through one or multiple steps. In certain embodiments, there has to be at least one cleavage reaction but one cleavage reaction may or may not be sufficient. For methods with plasma or whole blood, blood clotting may also be used as a visual readout.
Amplification Using Materials
Amplification process: The unit of the material contains molecules that once released can promote the release of more molecules from other units of the material. The units may be vesicles containing liquid (such as liposomes) or particles (of solid or gel).
Amplification using materials may be coupled with chemical amplification. In other words, besides effects on the material, the released molecules may also undergo chemical amplification.
The molecules enclosed in the materials may be inactive and get activated when released by chemical reactions with substances in the bulk or by any other ways such as change in conformation, release of self-inhibition by an attached inhibitor, etc. They may also be in the same form after release, but the breakage of the material is very slow when they are enclosed, such as when the material is designed to have different reactivity with the molecules in the enclosed environment or in the bulk.
Inhibitory mechanism: These molecules may also be inhibited by some substances in the bulk when they are released.
Activating mechanism: The input may also be or produce (through one or multiple steps) a substance that releases the molecules from the materials (this substance may be similar to or different from the enclosed) or releases the molecules from the inhibitor (by competing with the inhibitor or by inactivating the inhibitor).
Readout process: readout processes described elsewhere herein may be used, depending on specific situations.
Activating mechanism via generating a high concentration of substances.
In certain embodiments, the input generates a high local concentration of substances.
This technique can be used in combination with amplification processes using chemical reactions or materials.
Stochastic Confinement
Stochastic confinement as described above is a powerful tool to generate high local concentrations from a solution of low bulk concentration. This technique may also be used to distinguish interfering particles (molecules) with much weaker activity but much larger number than those of interest. When interfering particles (molecules, etc.) are in a large excess, the total activity of interfering particles may be higher than the total activity of active particles (molecules) of interest. Even with kinetic amplification via a threshold response, it is impossible to detect the signal from the particles of interest in the presence of overwhelming background signal from the interfering particles unless separation is used to isolate the particles of interest. This situation may be commonly encountered in the microbiological analysis of environmental samples. For example, human samples, including skin, saliva, and stool samples, being analyzed for microorganisms are routinely contaminated with bacteria or other cells (blood). Stochastic confinement is a powerful method to compartmentalize the active particles separately from the interfering particles. Therefore, as long as the threshold is tuned properly, the amplification process can selectively respond to only the active, target particles (molecules) even in the presence of a large excess of interfering particles (molecules).
Immobilizing Molecules onto a Limited Surface
Surfaces may be the interfaces of a plug, surface of a particle, surface of a microbe, and patterns on a surface which allows it to adsorb specific molecules. Many techniques for immobilizing molecules onto a surface may be used. These techniques include but are not limited to using antibodies, biotin/avidin interaction, and His-tag/Ni²⁺/nitrolotriacetic acid (His-tag/Ni/NTA) interaction.
Using Particles Containing Molecules
In some embodiments, the molecules are enclosed in the particles. When needed the particles release the molecules providing a burst of high concentration of these molecules in solution. This process may or may not be autocatalytic. In other words, the released molecules do not have to promote the release of more molecules, although they may.
Stochastic confinement, immobilizing molecules onto a limited surface and using particles containing molecules, may be used in different combinations to generate even higher local concentrations. For example, if the analytes are bacteria (with or without background interfering bacteria), the specific antibodies for the bacteria of interest can be used to concentrate a reporter enzyme on the surface of the bacteria. The bacteria tagged with reporter enzymes are then stochastically confined into plugs. These plugs are then merged with plugs containing vesicles containing secondary reporter enzymes. The reporter enzymes tagged on the bacteria promotes the breakage of the vesicles, releasing an amplified number of secondary reporter enzymes. The reporter enzymes then act as input for one of the amplification processes.
Using Membranes
To generate a high local concentration, a membrane can be used to keep one particular component on one side. For example, a membrane that is impermeable to H⁺ but permeable to R⁻—H⁺ can be used. On one side of the membrane, an amplification process uses R⁻—H⁺ (which diffuses in from the other side of the membrane) as input to initiate the production of H⁺ and amplifies the amount of H. Other substances and amplification processes may be used with appropriate membranes as well.
Using Phase for Separation
Two sets of substances are separated into insoluble or immiscible phases (solid/solid, solid/liquid, or liquid/liquid). A shuttle substance may transfer between two phases and speeds up the reaction in one or both phases by fast reactions with positive feedback. For example, the system contains solid oxidant and iron particles in water. If the analyte is Fe²⁺ (or Fe³⁺) or produces Fe²⁺ (or Fe³⁺), Fe²⁺ is oxidized by the solid oxidant to Fe³⁺, which oxidizes Fe(0) (from the iron particles) to Fe²⁺. Eventually, the solution contains a lot of Fe²⁺ (if the metal particles are in excess) or Fe³⁺ (if the solid oxidant is in excess) which may be detected by strongly colored indicators for Fe²⁺ or Fe³⁺.

EXPERIMENTAL

Microfluidic Device Design and Fabrication

Microfluidic devices were fabricated by using soft lithography (Y. N. Xia and G. M. Whitesides, Annu. Rev. Mater. Sci., 1998, 28, 153-184.) as described previously. (L. S. Roach, H. Song and R. F. Ismagilov, Anal. Chem., 2005, 77, 785-796; H. Song and R. F. Ismagilov, J. Am. Chem. Soc., 2003, 125, 14613-14619; L. Li, D. Mustafi, Q. Fu, V. Tereshko, D. L. L. Chen, J. D. Tice and R. F. Ismagilov, Proc. Natl. Acad. Sci. U.S.A., 2006, 103, 19243-19248.) Except where noted below, plugs were collected in PFA or PTFE Teflon tubing (Zeus, Orangeburg, S.C.) with 150 μm or 200 μm inner diameter (I.D.). The tubing was cut at a 45 degree angle, inserted into the outlet of the microfluidic device up to the inlet junction, and sealed into the device by using PDMS prepolymer (10:1 elastomer to curing agent). To aid in imaging of the plugs, the Teflon tubing was wound in a spiral on a glass slide, and PDMS prepolymer was poured over the tubing to fix it in place. The device with attached tubing was then autoclaved at 135° C. for 10 min to sterilize. Once sterilized, the glass slide containing the tubing was transferred to a sterile Petri dish.
Flowing Solutions into the Microfluidic Devices
All solutions were loaded into 1700 series Gastight syringes (Hamilton, Reno, Nev.) with removable 27 gauge needles and 30 gauge Teflon tubing (Weico, Wire & Cable, Edgewood, N.Y.). To maintain sterility, the syringes were filled and attached to the device within a biosafety cabinet. Syringes were connected to the microfluidic devices by using 30 gauge Teflon tubing. Solutions where flowed into the microfluidic devices by using previously described methods. (L. Li, D. Mustafi, Q. Fu, V. Tereshko, D. L. L. Chen, J. D. Tice and R. F. Ismagilov, Proc. Natl. Acad. Sci. U.S.A., 2006, 103, 19243-19248) Flow rates were controlled by using PHD 2000 infusion syringe pumps (Harvard Apparatus, Holliston, Mass.).
Bacterial Cell Culture
Cells were obtained from ATCC (Staphylococcus aureus ATCC#25923 (MSSA) and Staphylococcus aureus ATCC#43300 (MRSA)). Stock solutions of the cells were made by using Luria-Bertani media Miller formulation (LB) (BD, Sparks, Md.) containing 30% (v/v) glycerol and stored at −80° C. For each experiment, a vial of frozen stock was brought to room temperature and streaked onto a Modified Trypticase Soy Agar (TSA II, BD, Sparks, Md.) plate and incubated overnight at 30° C. Colonies from the plates were transferred to LB and cultured at 37° C., 140 rpm for 3 h at which point OD₆₀₀was 1.5-2.0. Cell densities were then adjusted by diluting in LB. To maintain sterility, all procedures were performed in a biosafety cabinet and all tubing, devices, syringes, and solutions used were either autoclaved, sterilized by EtOH, packaged sterile, or filtered through a 0.45 μm PES or PTFE filter.
Antibiotic Preparation
Antibiotic stock solutions of ampicillin (AMP), oxacillin (OXA), cefoxitin (CFX), levofloxacin (LVF), vancomycin (VCM), erythromycin (ERT) were made by using 150 mM NaCl_aqat a concentration of 4000 times greater than the final concentration in the plugs, filter sterilized, and then frozen at −80° C. (AMP, Fisher Bioreagents, Fair Lawn. NJ; OXA, LVF, Fluka, Buchs, Switzerland; CFX, VCM, ERT, Sigma, St. Louis, Mo.). For example, AMP was tested at the breakpoint concentration of 0.25 mg/L, meaning that the stock solution was prepared at a concentration of 1000 mg/L. In the case of erythromycin (ERT), a stock solution was prepared at 1000 times the final concentration in plugs. Before each experiment, vials of the antibiotics were thawed and diluted 1000×(250× for ERT) with saline containing 80 μM fluorescein carboxylate. Fluorescein carboxylate was used to aid in indexing the resultant array of plugs. The plugs in FIG. 7 contain no fluorescein carboxylate, since indexing was not required. The blank conditions consisted of 150 mM NaCl. Antibiotic solutions were further diluted on chip 1:3 (v/v) during plug formation. 20 μM fluorescein carboxylate did not interfere with the viability assay, the activity of the cells, or effectiveness of antibiotic in tests performed on 96 well plates.
Antibiotic Testing on Plates
Plates were made from Mueller Hinton Agar (Fluka, Switzerland). After autoclaving, the agar was cooled and antibiotics were added and 20 mL plates were poured. For CFX and OXA testing, 50 μL of MRSA and MSSA bacterial culture at 4×10³CFU/mL was spread onto separate TSA plates. The plates were incubated at 30° C. After 16.5 h and 40 h the plates were examined for colonies. MRSA colonies appeared on CFX after 16.5 h and on the OXA plates after 40 h. Even after 40 h, MSSA colonies did not appear on the CFX or OXA plates. For AMP, ERT, LVF, and VCM, 5 μL of culture at 2×10⁴CFU/mL were spread onto plates, and the plates were incubated at 37° C. for 12 h. After 12 h, growth of colonies on the plates was considered resistance to the antibiotic and no colonies on the plates were considered sensitivity to the antibiotic. For all tests, control plates with no antibiotic were inoculated to ensure that each plate tested received many CFU during inoculation.
Comparing Detection Times of Bacteria in Nanoliter Plugs and Milliliter-Scale Culture
Plugs were formed by using the general methods described previously. (L. Li, D. Mustafi, Q. Fu, V. Tereshko, D. L. L. Chen, J. D. Tice and R. F. Ismagilov, Proc. Natl. Acad. Sci. U.S.A., 2006, 103, 19243-19248; D. N. Adamson, D. Mustafi, J. X. J. Zhang, B. Zheng and R. F. Ismagilov, Lab Chip, 2006, 6, 1178-1186) Plugs were formed in a 3 inlet PDMS device with 100 μm wide channels by flowing S. aureus culture in LB at 2×10⁵CFU/mL at 1 μL/min, a 20% alamarBlue solution in saline at 1 μL/min, and fluorinated carrier fluid at 5 μL/min. 25 plugs were collected in the channel. Inlets and the outlet were sealed with silicon grease and the device was placed in a Petri dish containing LB for incubation. The same aqueous solutions were mixed 1:1 (total volume 0.6 mL) in a 14 mL polypropylene round-bottom tube (BD Falcon, Franklin Lakes, N.J.). After 2.8 h, plugs were made from the milliliter-scale culture by using the same method, with the cell culture containing alamarBlue for both aqueous inlets. Both sets of plugs were immediately imaged by using a epi-fluorescence microscope (IRE2, Leica) with a Cy3 (Chroma 41007, Cy3) filter and a 10×0.3 NA objective for a 5 ms exposure time with binning set to 4 and gain set to 200. Fluorescent images of plugs were processed by subtracting the average background intensity from all images. Line scans (FIG. 1 b) with a width of 25 pixels were taken along the long axis of each plug.
Experiment to Compare Plug Size to Detection Time
PDMS devices with channel widths ranging from 200 to 800 μm were prepared. Teflon tubing with diameter similar to that of the channel was cut at a 45° angle, inserted into the device up to the inlet junction, and sealed in place using PDMS. For FIGS. 1 c and d, plugs were formed as above, with the exception that the ˜1500 mL plugs were made via aspiration by using a manual aspirator. In addition, 1 mL plugs were formed in PTFE tubing with an outer diameter (OD) of 200 μm and an inner diameter (ID) of 90 μm, 690 mL plugs were formed in PTFE tubing with an OD of 700 μm and an ID of 600 μm, 100 and 120 mL plugs were formed in PTFE tubing with an OD of 800 μm and an ID of 400 μm, and 1500 mL plugs were formed in PTFE tubing with an OD of 1100 μm and an ID of 1000 μm. Plugs were collected in the Teflon tubing, the tubing was sealed with wax, and the tubing was placed in a Petri dish containing LB for incubation and imaging. Incubation and imaging was performed in a microscope incubator (Pecon GmbH, Erbach, Germany). Time zero is defined as the time at which the sample entered the incubator, which was less than 20 min after sample preparation. Fluorescence measurements were taken with 5 ms exposure times with a 5×0.15 NA objective using a 1× camera coupler for plug sizes 1, 12.6, 100, and 690 mL plugs and a 0.63× camera coupler for 1500 mL plugs. Plugs 125 mL in volume were imaged with 10 ms exposure times with a 5×0.15 NA and a 0.63× camera coupler.
Plugs were analyzed by first separating them from the background by thresholding to exclude intensity below 250. The average intensity of the thresholded plugs was measured. Over time, the intensity of the plugs diverged into 2 groups, occupied plugs and unoccupied plugs. All occupied plugs had a change in intensity more than 2 fold unoccupied plugs. Detection time is defined as the time at which the fold change of the occupied plugs compared to the intensity of unoccupied plugs reaches a maximum. Fold change in intensity is defined as the change in intensity of an occupied plugs divided by the average change in intensity of unoccupied plugs (Eq. 1).
Fold change_(t=ti)=Occupied plug(I _t=ti −I _t=1)/Unoccupied plugs(I _t=1 −I _t=1) (1)
In Eq. 1, I_t=tiis intensity at time point i. The intensity of the empty plugs is the average of all empty plugs in each experiment.
Comparing Detection Times of Bacteria in Nanoliter Plugs and 96 Well Plates
For FIG. 1 d, 96 well plates results for FIG. 1 d were acquired in a Tecan Safire II plate reader (MTX Lab Systems, Vienna, Va.) with Ex/Em 560/630 nm, gain 25, and 40 μs integration time. 200 μL of cell culture suspended in LB with 10% alamarBlue was added to wells of a Costar 96 well assay plate with black sides and a clear, flat bottom (Corning, Corning, N.Y.). Each data point represents triplicate measurements taken at 37° C. Fold change in intensity from 96 well plate results were calculated by using Eq. 1 where the well with LB and alamarBlue only was the unoccupied plug condition.
Screening Susceptibility of Bacteria to Many Antibiotics
For antibiotic screening experiments (FIG. 2), an array of ˜50 nL antibiotic plugs was aspirated into Teflon tubing (200 μm ID) using a manual aspirator. Air spacers were included between each antibiotic plug to prevent merging of adjacent antibiotic plugs and to enable indexing of plugs in the output array. Plugs of saline solution were included as the first and last plugs in the preformed array to serve as positive controls. The Teflon tubing containing the array of antibiotic plugs was sealed into a device inlet by using wax (Hampton Research, Aliso Viejo, Calif.). To screen the susceptibility of MRSA and MSSA to each antibiotic, bacterial samples and indicator were merged with the preformed array of antibiotic. Bacterial samples were at a density of 4×10⁵CFU/mL in LB, and viability indicator solution was made by mixing 4 parts alamarBlue solution (AbD Serotec, Oxford, UK) with 6 parts 150 mM NaCl., The flow rate of the antibiotic array was 0.25 μL/min; the flow rate of the bacterial solution was 0.5 μL/min, and the flow rate of the viability indicator was 0.25 μL/min. The carrier fluid was FC40 (Acros Organics, Morris Plains, N.J.) with a flow rate of 1.6 μL/min. For each antibiotic plug in the preformed array, approximately 50 smaller plugs (4 mL in volume) were formed, each potentially containing a single bacterium. The resulting plugs were collected in the coil PTFE Teflon tubing (I.D.=150 μm).
After plug formation, the tubing was disconnected from the PDMS device, and the ends were sealed with wax. The Petri dish containing the tubing was filled with 20 mL of LB solution to prevent evaporation of the plugs during incubation. The plugs were immediately transferred to a microscope incubator (Pecon GmbH, Erbach, Germany). Time zero is defined as the time which the plugs entered the incubator, which was about 20 minutes after plugs were formed. Fluorescence measurements for plugs were recorded by using an inverted epi-fluorescence microscope (DMI6000, Leica, Bannockburn, Ill.) with a 10×0.3 NA objective (HCX PL Fluotar) coupled to a CCD camera ORCA ERG 1394 (12-bit, 1344×1024 resolution) (Hamamatsu Photonics) by using a 0.63× camera coupler. Images were taken of each plug using Metamorph Imaging Software (Molecular Devices, Sunnyvale, Calif.) every 30 min with exposure times of 5 ms. Plugs were analyzed by first separating them from the background by thresholding to exclude intensity below 250. The average intensity of the thresholded plugs was measured. The change in intensity at time point ti is I_t=ti−I_t=1. In the experiment described in FIG. 2 c, fluorescence intensity of plugs was normalized by setting the intensity of the brightest plug to 100.
Determining the Minimal Inhibitory Concentration of a Drug Against a Bacterial Sample
For MIC determination in plugs (FIG. 3), a procedure similar to screening susceptibility of many antibiotics was used. The input array of antibiotics consisted of plugs of CFX at a range of concentrations. Bacterial samples were MRSA or MSSA in LB at cell densities near 10⁶CFU/mL. In FIGS. 3 b and c, fluorescence intensity of plugs was normalized as described for FIG. 2 c.
Statistical Analysis of Antibiotic Screening Results
Unpaired t-tests were performed to compare antibiotic screening results to positive and negative controls. For FIG. 2 c: VCM and LVF are statistically different than positive controls and AMP, CFX, OXA, ERT, and blank conditions were all statistically different than the negative control. For FIG. 3 b: 8 and 24 mg/L CFX were statistically different than positive controls and 0, 0.2, 1, 2, and 4 mg/L were statistically different than the negative control. For FIG. 3 c: 4, 8, and 24 mg/L CFX were statistically different than positive controls and 0, 0.2, 1, and 2 mg/L were statistically different than the negative control. P values are two-tailed.
Detection and Drug Screening of MRSA and MSSA in Human Blood Plasma
For FIG. 4, cells were suspended in a 1:1 mixture of human blood plasma (Pooled normal plasma George King Bio-Medical, Overland Park, Kans.) and LB containing 40% alamarBlue. Plugs were formed and collected in Teflon tubing (200 μm ID). Images were taken with a 5×0.15 NA objective with a 0.63× camera coupler. Texas red pictures were taking every 10 minutes with exposure times of 25 ms. A bright-field image was taken at beginning and end of experiment. Linescans of original plug images were taken at time 0 and time 7.5 h. Adobe Photoshop was used to enhance contrast of plugs shown in FIG. 4.
Microfluidic bacterial detection and drug screening are applicable to complex, natural matrices, including human blood plasma.
To validate the applicability of this method to detecting bacteria in natural matrices, this method was used to detect bacteria in a sample of human blood plasma. Bacterial strains MSSA or MRSA were inoculated into pooled human blood plasma at a concentration of 3×10⁵CFU/mL. To test the sensitivity of the bacteria to beta-lactams, the antibiotic ampicillin (AMP) was added to the culture at the breakpoint concentration. The inoculated plasma was then combined on-chip with viability indicator as illustrated FIG. 4 a. After 7.5 h of incubation at 37° C., plasma samples infected with MRSA were distinguishable from samples infected by MSSA by screening the samples against AMP at the breakpoint concentration. While plugs containing MRSA and AMP showed a similar increase in fluorescence intensity to plugs containing MRSA and no AMP (FIGS. 4 b and c), plugs containing MSSA and AMP showed no increase in fluorescence intensity (FIGS. 7 d and e).
Amplification examples 1-6 involve blood coagulation proteins. Amplification example 7 involves proteins in apoptosis. Amplification example 8 involves Limulus amebocyte lysate from Horseshoe crabs.

Amplification Example 1

Amplification process: The enzyme precursor is engineered prothrombin that may be cleaved by thrombin to produce more thrombin. The potential of this engineering approach is supported by a method to make an engineered factor X that may be cleaved by thrombin. (Louvain-Quintard, V. B.; Bianchini, E. P.; Calmel-Tareau, C.; Tagzirt, M.; Le Bonniec, B. F., Thrombin-activable factor X re-establishes an intrinsic amplification in tenase-deficient plasmas. Journal of Biological Chemistry 2005, 280, (50), 41352-41359.)
Inhibitory mechanism: The inhibitor is hirudin (or other inhibitors or antibodies of thrombin such as antithrombin III, heparin, or anophelin), which binds to thrombin and prevent the cleavage of prothrombin by thrombin.
Activating mechanism: The analyte, bacterial phosphatase, cleaves a tag attached to an inhibitor of hirudin (anti-hirudin), allowing this activated molecule to inhibit hirudin and release thrombin from hirudin. Alternatively, the bacterial phosphatase cleaves a tag attached to thrombin. In both cases, thrombin then cleaves prothrombin and produces more thrombin. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: Thrombin enzymatically cleaves a fluorogenic substrate (such as Boc-Asp(OBzl)-Pro-Arg-MCA) to give fluorescent signal.
This method is predicted to be feasible with concentration of prothrombin from 1 nM to 1 μM, be activated with concentration of thrombin input after concentrating techniques of as low as 10 μM, give response after 1-10 minutes, and give amplification gain of 3 to 9 orders of magnitude.

Amplification Example 2

Amplification process: The enzyme precursor is factor XII that may be cleaved by factor XIIa to produce more factor XIIa in the presence of dextran sulfate or negatively charge surface in general. (Tankersley, D. L.; Finlayson, J. S., Kinetics of Activation and Autoactivation of Human Factor-Xii. Biochemistry 1984, 23, (2), 273-279.)
Inhibitory mechanism: The inhibitor is ecotin (or other inhibitors or antibodies of factor XIIa).
Activating mechanism: the analyte, bacterial phosphatase, cleaves a tag attached to an inhibitor of ecotin, allowing this activated molecule to inhibit ecotin and release factor XIIa from ecotin. The system may also be designed so that the analyte cleaves a tag attached to factor XIIa or kallikrein. Factor XIIa or kallikrein then cleaves factor XII and produces more factor XIIa. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: Factor XIIa enzymatically cleaves a fluorogenic substrate (such as Boc-Gln-Gly-Arg-MCA) to give fluorescent signal.
This method is predicted to be feasible with concentration of factor XII from 0.1 μM to 10 μM, be activated with concentration of kallikrein input after concentrating techniques of as low as 1 nM, give response after 1-10 minutes, and give amplification gain of 3 to 5 orders of magnitude.

Amplification Example 3

Amplification process: The enzyme precursor is factor XI that may be cleaved by factor XIa to produce more factor XIa in the presence of dextran sulfate or negatively charge surface in general. (Gailani, D.; Broze, G. J., Factor-Xi Activation in a Revised Model of Blood-Coagulation. Science 1991, 253, (5022), 909-912; Naito, K.; Fujikawa, K., Activation of Human Blood-Coagulation Factor-Xi Independent of Factor-Xii-Factor-Xi Is Activated by Thrombin and Factor-Xia in the Presence of Negatively Charged Surfaces. Journal of Biological Chemistry 1991, 266, (12), 7353-7358.)
Inhibitory mechanism: The inhibitor is aprotinin (or other inhibitors or antibodies of factor XIa).
Activating mechanism: the analyte, bacterial phosphatase, cleaves a tag attached to an inhibitor of aprotinin, allowing this activated molecule to inhibit aprotinin and release factor XIa from aprotinin. The system may also be designed so that the analyte cleaves a tag attached to factor XIa. Factor XIa then cleaves factor XI and produces more factor XIa. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: Factor XIa enzymatically cleaves a fluorogenic substrate (such as Boc-Glu(OBzl)-Ala-Arg-MCA) to give fluorescent signal.
This method is predicted to be feasible with concentration of factor XI from 0.1 μM to 10 μM, be activated with concentration of factor XIa input after concentrating techniques of as low as 1 nM, give response after 1-10 minutes, and give amplification gain of 3 to 5 orders of magnitude.

Amplification Example 4

Amplification process: Factor XII and prekallikrein are precursors of factor XIIa and kallikrein, respectively. In the presence of dextran sulfate or a negatively charged surface in general, factor XIIa cleaves both factor XII and prekallikrein to produce factor XIIa and kallikrein, respectively, while kallikrein cleaves factor XII to produce factor XIIa. (Tankersley, D. L.; Finlayson, J. S., Kinetics of Activation and Autoactivation of Human Factor-Xii. Biochemistry 1984, 23, (2), 273-279.)
Inhibitory mechanism: Inhibitors or antibodies that are common to both factor XIIa and kallikrein (such as ecotin) or different inhibitors specific to each may be used.
Activating mechanism: the analyte, bacterial phosphatase, cleaves a tag attached to an inhibitor of ecotin, allowing this activated molecule to inhibit ecotin and release factor XIIa and kallikrein from ecotin. The system may also be designed so that the analyte cleaves a tag attached to factor XIIa and/or kallikrein, which then activate the amplification process. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: Factor XIIa or kallikrein or both enzymatically cleave fluorogenic substrates (such as Boc-Gln-Gly-Arg-MCA for factor XIIa and Pro-Phe-Arg-MCA for kallikrein) to give fluorescent signal.
This method is predicted to be feasible with concentration of factor XII from 0.1 μM to 10 μM and concentration of kallikrein from 10 μM to 10 nM, be activated with concentration of kallikrein input after concentrating techniques of as low as 1 nM, give response after 1-10 minutes, and give amplification gain of 3 to 5 orders of magnitude.

Amplification Example 5

Amplification process: The reaction network with positive feedback is shown in FIG. 13. This is an example of cases in which positive feedback loops may be achieved through multiple steps and many positive feedback loops may be coupled with each other to form a cascade. In this network, there are two positive feedback loops. The first one includes the production of factor Xa from factor X catalyzed by the input (InhA1), the production of factor VIIIa from factor VIII, the binding of factor VIIIa to factor IXa to form a complex, and the production of Xa from factor X catalyzed by the VIIIa:IXa complex. The output of this first loop is factor Xa. The second positive feedback loop includes the production of factor Va from factor V catalyzed by the input (factor Xa), the binding of factor Va to the input (factor Xa), the production of thrombin from prothrombin catalyzed by the Xa:Va complex or the input of the first loop (InhA1), and the production of factor Va from factor V catalyzed by thrombin. The output of this loop is thrombin. Thrombin is detected by the cleavage of a fluorogenic substrate to release fluorescent molecules catalyzed by thrombin. ATIII/heparin is used to inhibit factor Xa, the VIIIa:IXa complex, and thrombin.
Inhibitory mechanism: The chemical inhibitors are ATIII/heparin that may inhibit factor Xa, the VIIIa:IXa complex, and thrombin. Other common or specific inhibitors for these enzymes, of for factor VIIIa, Va, and IXa may be used as well.
Activation mechanism: Similar to examples 1-4, the analyte, bacterial phosphatase, may cleave tagged molecules and release them. These molecules may be the activated factors shown in FIG. 13 or inhibitors of the inhibitors of those activated factors used in the inhibitory mechanism. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: Thrombin enzymatically cleaves fluorogenic substrates (such as Boc-Asp(OBzl)-Pro-Arg-MCA) to give fluorescent signal as shown in FIG. 13. Other single of combinations of activated factors may be used to cleave fluorogenic substrates as well.
Using the techniques described in this application, this method is predicted to be feasible with concentration of enzyme precursors from 0.1 μM to 1 μM, be activated with concentration of InhA1 input after concentrating techniques of as low as 1 nM, give response after 1-10 minutes, and give amplification gain of 3 to 6 orders of magnitude.

Amplification Example 6

This example is similar to amplification example 5, but is more complicated. Here the amplification process contains most of the components in the natural blood clotting network, as shown in a review. (Kastrup, C. J.; Runyon, M. K.; Lucchetta, E. M.; Price, J. M.; Ismagilov, R. F., Using chemistry and microfluidics to understand the spatial dynamics of complex biological networks. Accounts of Chemical Research 2008, 41, (4), 549-558.) Positive feedback loops are achieved through multiple steps. Inhibitory mechanism, activating mechanism, and readout process are also similar to amplification example 5, and can be done with single of combinations of enzymes. Additionally, blood clotting may also be visualized by eyes. (Song, H.; Li, H. W.; Munson, M. S.; Van Ha, T. G.; Ismagilov, R. F., On-chip titration of an anticoagulant argatroban and determination of the clotting time within whole blood or plasma using a plug-based microfluidic system. Analytical Chemistry 2006, 78, (14), 4839-4849.)
Using the techniques described in this application and the reaction conditions previously described, (Kastrup, C. J.; Runyon, M. K.; Shen, F.; Ismagilov, R. F., Modular chemical mechanism predicts spatiotemporal dynamics of initiation in the complex network of hemostasis. Proceedings of the National Academy of Sciences of the United States of America 2006, 103, (43), 15747-15752.) the predicted response is in 1-5 minutes.
Results from simulations and experiments (Kastrup, C. J.; Runyon, M. K.; Shen, F.; Ismagilov, R. F., Modular chemical mechanism predicts spatiotemporal dynamics of initiation in the complex network of hemostasis. Proceedings of the National Academy of Sciences of the United States of America 2006, 103, (43), 15747-15752.) shown in FIGS. 14 and 15 support the ideas in amplification examples 1-6 discussed above. In general, they show a threshold-like behavior in which the time of response (time for amount of some certain substance to reach a detectable value) drastically reduces as the amount of input increases over a certain value. Although only amplification examples 1, 2, and 4 were considered in FIG. 14, and amplification example 6 in FIG. 15, the reaction network of amplification example 3 is similar to that of amplification example 1 and the reaction network of amplification example 5 has complexity between those of amplification examples 4 and 6. Therefore, amplification examples 3 and 5 is expected to work as well.
FIG. 14 is the time to get response versus amount of input obtained by simulation using previously found rate constants (Tankersley, D. L.; Finlayson, J. S., Kinetics of Activation and Autoactivation of Human Factor-Xii. Biochemistry 1984, 23, (2), 273-279; Kuharsky, A. L.; Fogelson, A. L., Surface-mediated control of blood coagulation: The role of binding site densities and platelet deposition. Biophysical Journal 2001, 80, (3), 1050-1074; Ulmer, J. S.; Lindquist, R. N.; Dennis, M. S.; Lazarus, R. A., Ecotin Is a Potent Inhibitor of the Contact System Proteases Factor Xiia and Plasma Kallikrein. Febs Letters 1995, 365, (2-3), 159-163. Kawabata, S. I.; Miura, T.; Morita, T.; Kato, H.; Fujikawa, K.; Iwanaga, S.; Takada, K.; Kimura, T.; Sakakibara, S., Highly Sensitive Peptide-4-Methylcoumaryl-7-Amide Substrates for Blood-Clotting Proteases and Trypsin. European Journal of Biochemistry 1988, 172, (1), 17-25; Stone, S. R.; Hofsteenge, J., Kinetics of the Inhibition of Thrombin by Hirudin. Biochemistry 1986, 25, (16), 4622-4628.)
FIG. 1( a) is the simulation of the amplification process used in amplification example 1, with set initial concentration of the engineered prothrombin (1.4*10⁻⁶M), thrombin (1.4*10⁻¹⁰M), and hirudin (1*10⁻⁸M), and varied concentration of input which is thrombin. Time of response for each concentration of input was defined as the time when concentration of thrombin reaches 80% of the initial concentration of prothrombin (if this time is larger than 10000 seconds, it was set to 10000 seconds). Rate of cleavage of engineered prothrombin by thrombin was taken to be the same as the rate of cleavage of factor V by thrombin.
FIG. 1( b) is the simulation of the amplification process used in example 2, with set initial concentration of the factor XII (1×10⁻⁶M), factor XIIa (1×10⁻¹⁰M), ecotin (1×10⁻⁷M), and Boc-Gln-Gly-Arg-MCA (fluorogenic substrate for factor XIIa), and varied concentration of input which is kallikrein. Time of response for each particular concentration of input was defined as the time when concentration of the fluorescent molecules reaches 80% of the initial concentration of the fluorogenic substrate. If this time is larger than 10000 seconds, it was set to 10000 seconds.
FIG. 1( c) Simulation of the amplification process used in example 4, with set initial concentration of the factor XII (1×10⁻⁶M), factor XIIa (1×10⁻¹⁰M), prekallikrein (1×10⁻¹⁰M), kallikrein (1×10⁻¹⁴M), ecotin (1×10⁻⁷M), and Boc-Gln-Gly-Arg-MCA (fluorogenic substrate for factor XIIa), and varied concentration of input which is kallikrein. Time of response for each particular concentration of input was defined as the time when concentration of the fluorescent molecules reaches 80% of the initial concentration of the fluorogenic substrate. If this time is larger than 10000 seconds, it was set to 10000 seconds.
FIG. 15 is the experimental results showing how blood clotting time varied with size of patch of tissue factor, an input for the blood clotting network discussed in example 6. (Kastrup, C. J.; Runyon, M. K.; Shen, F.; Ismagilov, R. F., Modular chemical mechanism predicts spatiotemporal dynamics of initiation in the complex network of hemostasis. Proceedings of the National Academy of Sciences of the United States of America 2006, 103, (43), 15747-15752.) The patch size in these experiments correlated with local concentration of tissue factor.

Amplification Example 7

Amplification process: This reaction network with positive feedback involves proteins in apoptosis. This network includes the production of caspase-9 from procaspase-9 catalyzed by cytochrome C and Apaf1, the dimerization of caspase-9, the production of caspase-3 from procaspase-3 catalyzed by caspase-9-dimer, the production of caspase-9 from procaspase-9 by catalyzed by either caspase-9 dimer or caspase-3.
Inhibitory mechanism: Inhibitors or antibodies for caspase-3 and/or caspase-9 are used.
Activation mechanism: Similar to examples 1-5, the analyte, bacterial phosphatase, may cleave tagged molecules and release them. These molecules may be caspase-3 and/or caspase-9, or inhibitors of the inhibitors used in the inhibitory mechanism. The specificity of this method may be designed to match expectation by changing the specificity of the cleavage of the tag by bacterial phosphatase. If detection of an enzyme different from phosphatase is needed, a different tag is used.
Readout process: caspase-3 and/or caspase-9 enzymatically cleave fluorogenic substrates to give fluorescent signal.
Using the techniques described in this patent, this method is predicted to be feasible with concentrations of enzyme precursors from 0.1 μM to 10 μM, be activated with concentration of input after concentrating techniques of as low as 1 nM, give response after 1-10 minutes, and give amplification gain of 3 to 5 orders of magnitude.

Amplification Example 8

Limulus amebocyte lysate (LAL) is known to coagulate when bacterial lipopolysaccharide (LPS) is present. One mechanism is the binding of LPS to an 82-kDa protein (termed LPS-binding protein (LBP)), which normally negatively regulates coagulation. (Roth, R. I.; Tobias, P. S., Lipopolysaccharide-Binding Proteins of Limulus Amebocyte Lysate. Infection and Immunity 1993, 61, (3), 1033-1039.)
Amplification process: The clotting network of LAL.
Inhibitory mechanism: A small excess of LBP.
Activating mechanism: Generally, the input may be or promote the production of some substance that binds to LBP. The input may be bacterial LPS or bacterial phosphatase that may cleave tagged (and inactive) LPS.
Readout process: Clotting or absorbance at 405 nm may be used.
The variations below can be applied individually or in combinations with other variations to all of amplification examples 1-8 shown above.
Variation 1:
Negative contrast is used instead of positive contrast.
Amplification processes: The processes used in examples 1-8 are used here.
Inhibitory mechanism: There is no inhibitory mechanism.
Activating mechanism: Inhibitory mechanism used in examples 1-8 are used as input to see the contrast.
Readout process: The processes used in examples 1-8 are used here.
Variation 2:
An enzyme precursor (inactive form) and an enzyme (active form) do not have to be two totally different molecules. They only need to have different reactivity.
The inactive form may have an inactive conformation while the active form has an active conformation. A conformation change may be facilitated by binding of a small molecule to an enzyme, an enzyme to and enzyme, a small molecule to a DNA or RNA molecule, an enzyme to a DNA or RNA molecule, or binding of more than two substances.
The inactive form may be tagged with an inhibitor, thus being self-inhibitory. When the linker to the inhibitor is cleaved, the molecule is now active. Positive feedback can be incorporated in these variations because the active enzyme can catalyze the change of conformation or the cleavage of the self-inhibiting tag of another enzyme of the same kind or of different kind.
Variation 3:
Detection of molecules other than enzymes may be achieved as well. To the systems in examples 1-8 without or with any single or combination of variations, one or multiple steps is added. The analyte, which may not be an enzyme, promotes the activation of an enzyme that may cleave the tag from the substrate used in activating mechanisms in examples 1-8, through one or multiple reactions. For example, activation of enzymes may be done through change of conformation (a result of binding of some molecule to an enzyme), linking two inactive components to make an active substance, or activating another enzyme that may activate the enzyme of interest.
Variation 4:
The techniques described above with or without any variations may be used in combinations with amplification using materials and activation mechanism involving generation of high local concentration.

Amplification Example 9

Amplification process: The reaction of the purple complex of Co(III) and 2-(5-bromo-2-pyridylazo)-5[N-n-propyl-N-(3-sulfopropyl)amino]phenol (5-Br-PAPS) and oxone has Co²⁺ (aq) as the autocatalyst. (Endo, M.; Abe, S.; Deguchi, Y.; Yotsuyanagi, T., Kinetic determination of trace cobalt(II) by visual autocatalytic indication. Talanta 1998, 47, (2), 349-353; Endo, M.; Ishihara, M.; Yotsuyanagi, T., Autocatalytic decomposition of cobalt complexes as an indicator system for the determination of trace amounts of cobalt and effectors. Analyst 1996, 121, (4), 391-394; Tsukada, S.; Miki, H.; Lin, J. M.; Suzuki, T.; Yamada, M., Chemiluminescence from fluorescent organic compounds induced by cobalt(II) catalyzed decomposition of peroxomonosulfate. Analytica Chimica Acta 1998, 371, (2-3), 163-170.)
Inhibitory mechanism: A ligand that can be used to capture Co²⁺. As a technical detail, the Co(III).(5-Br-PAPS) complex may be separated from oxone by immobilizing them into two different layers or keeping the mixture dry.
Activating mechanism: The input, which may be an enzyme or a small molecule, may cleave or compete for the ligand that captures Co²⁺ discussed in the inhibitory mechanism. Alternatively, the input may also be a reducing agent that reduces Co³⁺ to Co²⁺ rapidly but reduces oxone more slowly.
Readout process: The Co(III).(5-Br-PAPS) complex has a purple color that is lost when the reaction gets activated because of the conversion of Co(III) into Co(II) and the oxidation of the organic ligand.

Amplification Example 10

Amplification process: The reduction of aqueous Ag(I) is catalyzed by its product, Ag(0). Other metals (such as Pd, Au, and Pt) may also be used for the autocatalytic reduction of metal cations to lower oxidation states.
Inhibitory mechanism: The reductant may be chemically protected. For example, the reductant may be based on hydroquinone, with the hydroxyl groups protected by a group that is removed by the process of interest. The metal cation may be captured by a ligand. The autocatalyst metal may be coated with small molecules, gel or polymer.
Activating mechanism: The input (enzyme or small molecule) may deprotect the reductant by cleaving off the protective rings. The input may also destroy or compete for the ligand. The input may also remove the coating layer on the metal particle.
Readout process: Visual readout may be achieved via aggregates (precipitates) of metal, specific indicator for different oxidation states of the metals, or general reduction-oxidation indicator.

Amplification Example 11

Amplification process: The reduction-oxidation reaction between chlorite and iodide has iodine as the autocatalyst. (Dateo, C. E.; Orban, M.; Dekepper, P.; Epstein, I. R., Systematic Design of Chemical Oscillators 0.5. Bistability and Oscillations in the Autocatalytic Chlorite Iodide Reaction in a Stirred-Flow Reactor. Journal of the American Chemical Society 1982, 104, (2), 504-509.)
Inhibitory mechanism: A compound may be used to consume the autocatalyst, iodine, thus slowing down the reaction effectively to the degree that no significant output is visualized after a long time.
Activating mechanism: The input may be iodine, or produce iodine through one or multiple reactions. The input may also be a compound that inhibits the consumption of iodine, or may produce such compound after one or many steps.
Readout process: Starch may be used to give a strongly colored blue complex with the product iodine.

Amplification Example 12

Amplification process: The reduction-oxidation reaction between thiosulfate and chlorite is autocatalytic in both hydronium ion and chloride ion, (Runyon, M. K.; Johnson-Kerner, B. L.; Ismagilov, R. F., Minimal functional model of hemostasis in a biomimetic microfluidic system. Angewandte Chemie-International Edition 2004, 43, (12), 1531-1536.
Horvath, A. K.; Nagypal, I.; Epstein, I. R., Kinetics and mechanism of the chlorine dioxide-tetrathionate reaction. Journal of Physical Chemistry A 2003, 107, (47), 10063-10068; Horvath, A. K.; Nagypal, I.; Epstein, I. R., Oscillatory photochemical decomposition of tetrathionate ion. Journal of the American Chemical Society 2002, 124, (37), 10956-10957; Nagypal, I.; Epstein, I. R., Systematic Design of Chemical Oscillators 0.37. Fluctuations and Stirring Rate Effects in the Chlorite Thiosulfate Reaction. Journal of Physical Chemistry 1986, 90, (23), 6285-6292.) and may be activated by Ag⁺ which is predicted to oxidize thiosulfate and release hydronium or chloride (preliminary result).
Inhibitory mechanism: The amount of hydronium ion may be kept in check by using a pH buffer. Chloride anion may be sequestered by a cation that is a sufficiently weak oxidizer to not react with thiosulfate. Ag⁺ may be sequestered by ligands. Solids that produce an acidic environment or solids that are salts of Cl⁻ or Ag⁺ may be coated with materials such as gel or polymer.
Activating mechanism: The input may be acidic or promote the production of any acid through one or multiple steps. The input may also be chloride anion or promote the production of chloride anion, or compete with the cation that captures chloride anion, or react with the cation to disable its ability to capture chloride anion. The input may also be Ag⁺ or produce Ag⁺ by compete for or inactivate the ligands for Ag⁺. The input may also break the coating materials of the solid particles if such inhibitory mechanism is used.
Readout process: A pH or reduction-oxidation indicator may be used.
Using the techniques described in this patent and the reaction conditions previously described, (Kastrup, C. J.; Runyon, M. K.; Shen, F.; Ismagilov, R. F., Modular chemical mechanism predicts spatiotemporal dynamics of initiation in the complex network of hemostasis. Proceedings of the National Academy of Sciences of the United States of America 2006, 103, (43), 15747-15752.) one can get response in 1-2 minutes.

Amplification Example 13

Amplification process: The organic reactions developed by Ichimura and coworkers (Ichimura, K., Nonlinear organic reactions to proliferate acidic and basic molecules and their applications. Chemical Record 2002, 2, (1), 46-55.) have acids as autocatalysts.
Inhibitory mechanism: The amount of hydronium ion may be kept in check by using a pH buffer or a weak base.
Activating mechanism: The input may be acidic or promote the production of any acid through one or multiple steps.
Readout process: A pH or reduction-oxidation indicator is used.
The variations below can be applied individually or in combinations with other variations to all of amplification examples 9-13 shown above
Variation 1:
Negative contrast is used instead of positive contrast.
Amplification processes: The processes used in examples 1-5 described above are used.
Inhibitory mechanism: There is no inhibitory mechanism.
Activating mechanism: Inhibitory mechanism used in examples 1-5 described above are used as input to see the contrast.
Readout process: The processes used in examples 1-5 described above are used here.
Variation 2:
The techniques described above with or without any variation may be used in combinations amplification using materials and activation mechanism involving generation of high local concentration.

Amplification Example 14

Provided the analytes are particles that can be tagged with some activating molecules through various methods (such as antibodies, His-tag/Ni/NTA, biotin/avidin etc), a method can be used to detect small concentration of the analytes in which the amplification process may be any. The activating molecule is first concentrated on particles. Then stochastic confinement is performed to concentrate these particles into plugs. The concentrated activating molecules act as input for the amplification process. Because excess activating molecules are not concentrated on particles, even after stochastic confinement, the concentration is still not high enough to activate the amplification process.
For example, the amplification process may be chosen as the one described in example 5 of section 1 and in FIG. 13, the activating molecules may be chosen to be InhA1, as shown in FIG. 16. The particular example is predicted to be able to detect particles of concentration of as low as 1fM, which may hypothetically be amplified to 1 nM after stochastic confinement. To total time of detection is predicted to be 1-10 minutes.
FIG. 16 illustrates combining amplification cascades with stochastic confinement enables sensitive detection of single particles or more. The general idea is described in the text. The protease shown here represents an activating molecule in general. (A and B) Particles are added to a container containing the activating molecules. (C) The particles bind to the activating molecules and are stochastically confined in plugs. (D) In plugs loaded with particles, the concentration of activating molecules is above the detection threshold and produces a threshold signal, while plugs without particles do not. (E and F) In solutions with excess but sub-threshold concentrations of activating molecules, stochastic confinement of particles into plugs (G) results in a few plugs containing a particle with multiple copies of the activating molecules attached and many plugs with a few copies of activating molecules. (H) Plugs containing a particle initiate the cascade, because stochastic confinement of the particles has concentrated the activating molecules in occupied plugs, whereas plugs without a particle remain at the sub-threshold protease concentration of the bulk solution.

Amplification Example 15

The analytes are particles which bind to activating molecules. The reactions may be chosen from those described in sections 1 and 2. For example, if the system with example 2 in section 1 is chosen, the activating molecule is kallikrein.
Amplification process: The enzyme precursor is factor XII that may be cleaved by factor XIIa to produce more factor XIIa in the presence of dextran sulfate or negatively charge surface in general.
Suppose a sample containing low concentration of type-B particles (such as 100 CFU/mL) has a large excess of interfering type-A particles (such as 10⁵CFU/mL) that can bind much less activating molecules per particle (such as 100 fold) (with a specifically designed antibody or any other means), but much more collectively. Under bulk detection mechanisms, either with amplification or by classical methods for measuring enzyme concentrations, the activity due to the excess of type-A particles will dominate the response (FIGS. 17 A and B). When the tagged particles are stochastically confined in plugs, only those containing type-B particles will have enough activating molecules to activate the amplification process. In general, any kind of particles which can bind to any of activating molecules described in sections 1 and 2 through antibodies or any other method may be detected using this method. For example, this technique may be use to detect B. anthracis from a sample containing a lot of interfering bacteria such as B. circulans and other kinds.
FIG. 17 illustrates selective detection of particles by using stochastic confinement. (A and B) In a bulk solution containing both a high concentration of interfering type-A particles (such as the bacteria B. circulans) (small gray particles binding few activating molecules (such as kallikrein)) and a low concentration of target type-B particles (B. anthracis) (large gray particles binding many activating molecules), the activity of the excess type-A particles dominates. Therefore, solutions with excess type-A particles (A) and excess type-B particle with a low concentration of type-A particles (B) both trigger a detection response (dark gray) (C and D). When stochastically confined in plugs, the amount of activating molecule in each plug made from solution A remains below the detectible level, while the amount of activating molecules of plugs containing type-B particles from solution B results in readout of type-B particles (dark gray plug), not from interfering type-A particles (lighter gray plugs).

Amplification Example 16

This technique is a variation of example 2 in section 2. Ag particles are coated with tags to form Ag—(X—Y)_n. This coated particle cannot bind with bacteria. However, the input cleaves or produces some substance that cleaves X—Y, exposing X on the surface of the particles. These particles now can be locally concentrated using techniques shown in section 4ii.

Amplification Example 17

This technique is a variation of example 4 in section 2. This system contains AgCl particles coated with an inert shell (such as Ca₃(PO₄)₂) and protected ethylenediaminetetraacetic acid (EDTA) ligands (in the solution or in the particle), and thiosulfate and chlorite in the solution. The input deprotects EDTA, allowing it to complex with Ca²⁺, dissolving the inert shell, exposing AgCl to the thiosulfate/chlorite mixture and activates the amplification process. The threshold is set up by the thickness of the inert shell. To detect different kinds of input, the method to protect and deprotect EDTA may be customized. The major reaction in this amplification process was described previously. (Runyon, M. K.; Johnson-Kerner, B. L.; Ismagilov, R. F., Minimal functional model of hemostasis in a biomimetic microfluidic system. Angewandte Chemie-International Edition 2004, 43, (12), 1531-1536; Horvath, A. K.; Nagypal, I.; Epstein, I. R., Kinetics and mechanism of the chlorine dioxide-tetrathionate reaction. Journal of Physical Chemistry A 2003, 107, (47), 10063-10068; Horvath, A. K.; Nagypal, I.; Epstein, I. R., Oscillatory photochemical decomposition of tetrathionate ion. Journal of the American Chemical Society 2002, 124, (37), 10956-10957; Nagypal, I.; Epstein, I. R., Systematic Design of Chemical Oscillators 0.37. Fluctuations and Stirring Rate Effects in the Chlorite Thiosulfate Reaction. Journal of Physical Chemistry 1986, 90, (23), 6285-6292.)

Claims

1. A method of detecting bacteria in a sample, comprising:

creating an array of plugs by introducing a first plug fluid into a flow of carrier fluid in a microchannel; wherein the majority of plugs in the array do not contain a bacterium;

wherein the first plug fluid is substantially immiscible with the carrier fluid and comprises a concentration of the sample diluted such that at most 2 bacteria are present in any plug; and

analyzing the array for the presence of bacteria.

2. The method of claim 1, wherein the array is analyzed for a detectable signal produced by the bacteria, wherein the detectable signal is a substance produced by the bacteria or is produced when the bacteria consumes a substance in the plug.

3. The method of claim 2, wherein the substance is selected from the group consisting of oxygen, carbon, a protein produced by the bacteria, a molecule produced by a bacterial enzymatic reaction, and a redox/potential sensitive indicator.

4-5. (canceled)

6. The method of claim 1, wherein the sample is from human, soil or marine.

7. (canceled)

8. The method of claim 1, wherein the bacteria are at a higher concentration in the plugs than in the sample.

9. The method of claim 1, wherein the plugs contain different species of bacteria.

10. The method of claim 1, further comprising introducing a plug fluid comprising media capable of supporting bacterial growth into the plug.

11. The method of claim 10, wherein the detectable signal is produced by growth of the bacteria.

12. The method of claim 1, wherein the plugs comprise a substance capable of inducing virulence in the bacteria.

13. The method of claim 1, wherein the at least two plug comprise a substance capable of lysing the bacteria.

14. (canceled)

15. The method of claim 1, further comprising assaying the bacteria detected for a biological activity.

16. The method of claim 1, further comprising identifying the bacteria.

17. The method of claim 1, further comprising splitting a plug that has been determined to contain bacteria into multiple plugs, each containing at least one of the bacteria.

18. The method of claim 1, further comprising conducting a polymerase chain reaction on the contents of the at least two plugs prior to analyzing the at least two plugs for the detectable signal.

19. A method of detecting bacteria comprising:

flowing at least two plugs in a carrier fluid through a microchannel;

wherein each plug comprises a plug fluid substantially immiscible with the carrier fluid;

wherein a first plug comprises a means for detecting a first species of bacteria;

wherein a second plug comprises a means for detecting a second species of bacteria different from the first species of bacteria;

introducing a sample optionally comprising bacteria into the first and second plugs, wherein the bacteria produce a detectable signal; and

analyzing the plugs for the detectable signal.

20-28. (canceled)

29. A method of screening for antibiotic activity comprising:

flowing at least two plugs in a carrier fluid through a microchannel;

wherein each plug comprises a plug fluid immiscible with the carrier fluid and media capable of supporting bacterial growth;

wherein the first plug comprises a first antibiotic candidate;

wherein the second plug comprises a second antibiotic candidate;

introducing a sample which comprises bacteria into the first and second plugs; and

detecting the presence of bacterial growth.

30-44. (canceled)

45. A method of detecting bacteria in an aqueous sample, comprising:

dividing a sample into a plurality of subsamples in a microfluidic device,

where each subsample is separated from another by a fluorinated liquid, each subsample has a volume that is a nanoliter or less, the majority of the subsamples do not contain a bacterium, and each subsample has at most 2 bacteria; and

analyzing the volumes for the presence of bacteria.