METHODS FOR LIPID EXTRACTION AND IDENTIFICATION OF MICROBES USING SAME VIA MASS SPECTROMETRY

STATEMENT OF FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT

[0001] This invention was made with government support under Grant No. GM111066 awarded by the National Institutes of Health. The government has certain rights in the invention.

TECHNICAL FIELD

[0002] The present invention generally relates at least to the field of medicine and in particular medical diagnostics wherein a novel method for rapid and accurate bacterial identification based on lipid extraction and analysis is detailed.

BACKGROUND OF INVENTION

[0003] Infectious diseases are a continuing threat to public health due to the unpredictability and potential of explosive global effect [1]. Classical methods in clinical settings for bacterial identification rely on the gold standard, namely bacterial culture in broth or solid plates, followed by microscopy examination and biochemical testing [2,3]. However, it can take days or even weeks to obtain diagnostic results due to slow microorganism growth rates. Although modern PCR-based nucleic acids amplification and genome sequencing technologies have transformed the field of microbial diagnosis, these techniques suffer from limitations of high false positive rate, heavy informatics burden and high cost [2,4,5].

[0004] The need for alternative technologies brings mass spectrometry (MS) to the stage as it is a powerful analytical tool in biomedical research [6], particularly in rapid and accurate bacterial identification [7,8]. Matrix-assisted laser desorption ionization time of flight MS (MALDI-TOF MS) is playing a significant role in clinical diagnostics [7,9-13]. There are several commercial available platforms based on MALDI-TOF MS including: the VITEK MS (bioMerieux S.A., France; [14]) and the Bruker Biotyper (Bruker Daltonics, Billerica, MA; [15]). Both platforms utilize a strategy of generating protein MS profiles for an unknown species sample and then comparing it with a reference protein library for bacterial identification [16]. [0005] The success of MS-based commercial platforms enhances the rapidity and accuracy of diagnostics. However, several limitations of protein-based methodologies have been reported including: the need of pure colonies for identification, difficulties in distinguishing closely related species (e.g. Escherichia coli vs. Shigella flexneri), inability to detect antibiotic-resistant strains from the FDA-approved libraries, and incapability of identifying microorganisms directly from complicated biofluids.

[0006] To overcome the limitations of MS-based commercial platforms that utilize protein analysis, Leung et al. [17] reported a complementary approach that employs the use of essential bacterial membrane glycolipids and lipids for bacterial identification. They demonstrated that glycolipids, lipids and related molecules show species-specific characteristics, which can be used as a chemical barcode for identification of Gram-positive, Gram-negative and fungal species.

[0007] Complex and diverse, glycolipids and lipids are a major component of bacterial membranes. In Gram-positive bacteria, the cell wall consists of many layers of peptidoglycan sheet where cardiolipin and lipoteichoic acid (LTA) are arranged perpendicular to the sheet, which are unique to the Gram -positive cell wall [18,19].

[0008] In Gram-negative bacteria, two membranes form the cell wall in which a single layer of peptidoglycan (inner membrane) is surrounded by an outer membrane [19]. Lipid A, a unique class of glycolipids in Gram-negative bacterial membranes, is the hydrophobic anchor of lipopolysaccharide (LPS) in the outer leaflet of the outer membrane. Structurally, lipid A is comprised of a P-l ',6-linked diglucosamine backbone, flanked by terminal phosphate residues and attached fatty acyl chains from the backbone. Due to variation in fatty acid composition and terminal phosphate moiety modifications, lipid A is a highly diverse molecule [20].

[0009] Libraries of bacterial glycolipids and lipids exist, with one, for example, containing 50 entries, including molecules from clinically important ESKAPE pathogens such as

Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enter obacter spp [17]. Such libraries could be used as the basis for determining the identity of an unknown microbe in a sample, when bacterial glycolipids and lipids from the sample are isolated, analyzed and compared to the libraries.

[0010] Consistent means for extraction of glycolipids and lipids from microbial membranes is critical to ensure reliable identification of bacteria. One means for extraction using ammonium i sobutyrate (AI) readily generates a comprehensive reference library [21]. However, this protocol requires at least 18 hours to obtain extracts ready for MS analyses. Further, it requires the use of noxious chemicals and multi-step centrifugation, thereby prohibiting its easy application in clinical laboratory or hospital settings. There are other methods dedicated to lipids extraction but none of them permits high-quality mass spectra analysis from bacterial samples within a short period of time, e.g. one hour or less [22-27]. In addition, some methods are too complicated for healthcare practitioners and require additional laboratory equipment.

[0011] Thus, there is an urgent need for a rapid protocol that can reduce the time-to- diagnosis within one hour, that are uncomplicated and that do not use dangerous chemicals.

BRIEF SUMMARY OF INVENTION

[0012] To address the limitations of known bacterial lipid and glycolipid extraction protocols, the present invention provides a novel and clinician-friendly extraction protocol that reduces the time from bacterial cultures to MS-ready samples to less than an hour. Indeed, in some circumstances, the minimum processing time can be reduced to less than 30 minutes.

Further, the lipid extraction protocol of the present invention can be used to isolate lipids, such as glycolipids, from fungi as well as bacteria without prior knowledge of microorganism type or lipid class.

[0013] As detailed herein, the protocol uses a low pH buffer to release lipids from microbial membranes, followed by a single step purification, which circumvents the use of noxious chemicals and a time-consuming overnight lyophilization step. As a shorthand, the lipid extraction protocol of the present invention is also termed "the sodium acetate method" and "the SA method".

[0014] In a first embodiment, the invention is thus directed to a method for extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-5.0 using a buffer to form a mixture;

(b) incubating the mixture of (a) at a temperature of between about 37°C and 130°C for up to 180 minutes; and

(c) collecting the lipids from the mixture of (b), thereby extracting lipids from a sample.

[0015] In certain aspects of the first embodiment, the method further comprises (d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution; and (e) collecting the lipids from the mixture of (d). In certain other aspects of the first embodiment, the method further comprises (d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution; (e) collecting the lipids from the mixture of (d); and (f) resuspending the collected lipids of (e) in a solution of containing at least 20% chloroform and/or at least 10% methanol.

[0016] In a second embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-5.0 using a buffer to form a mixture;

(b) incubating the mixture of (a) at a temperature of between about 37°C and 130°C for up to 180 minutes;

(c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution of containing at least 20% chloroform and/or at least 10% methanol, thereby extracting lipids from a sample.

[0017] In a third embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-4.5 using a buffer to form a mixture;

(b) incubating the mixture of (a) at a temperature of between about 85°C and 105°C for up to two hours;

(c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is ethanol having a concentration of about 90-100%;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution containing at least 20%

chloroform and/or at least 10% methanol, thereby extracting lipids from a sample. [0018] In a fourth embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 3.5-4.0 using a buffer to form a mixture, wherein the buffer is a sodium acetate, potassium acetate, phosphoric acid, or acetic acid solution;

(b) incubating the mixture of (a) at a temperature of about 100°C for up to about 30 minutes;

(c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with 95% ethanol;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution of containing

chloroform/methanol/water, 3 : 1.5:0.25, v/v/v, thereby extracting lipids from a sample.

[0019] In relevant embodiments of the invention, the samples used in the methods comprise one or more microbes, or portions thereof. The one or more microbes may be, but are not limited to, a bacterium or bacteria, a fungus or fungi, or combinations thereof. In certain aspects, the one or more microbes are: one or more species of bacteria; one or more species of fungi; or one or more species of bacteria and one or more fungi.

[0020] In relevant embodiments of the invention, the pH of the sample in (a) is adjusted using a buffer comprising a mixture of sodium acetate and acetic acid, a mixture of potassium acetate and acetic acid, or a mixture of sodium phosphate and phosphoric acid.

[0021] In relevant embodiments of the invention, the pH of the sample in (a) is adjusted using sodium acetate, potassium acetate, phosphoric acid, or acetic acid. In certain aspects, the sodium acetate has a concentration of about 50-150 mM.

[0022] In relevant embodiments of the invention, the pH of the sample in (a) is adjusted to 3.0-4.0.

[0023] In relevant embodiments of the invention, the lipids in (a) were collected from the sample prior to adjusting the pH of the sample. In certain aspects, the collecting is via one or more of centrifugation, filtration, and precipitation. In certain aspects, the collecting is via centrifugation and said centrifugation is at a relative centrifugal force (RCF) of about 3000-6000 g for about 10-20 minutes. [0024] In relevant embodiments of the invention, the mixture in (b) is incubated at a temperature of between about 90°C and 100°C for up to one hour.

[0025] In relevant embodiments of the invention, the mixture in (b) is incubated at a temperature of about 100°C for 10-30 minutes.

[0026] In relevant embodiments of the invention, the mixture in (b) is cooled to about room temperature after the incubating.

[0027] In relevant embodiments of the invention, the collecting in (c) is via one or more of centrifugation, filtration, and precipitation. In certain aspects, the collecting is via centrifugation at a RCF of about 7000-9000 g for about 2-15 minutes.

[0028] In relevant embodiments of the invention, the solvent in (d) is ethanol.

[0029] In relevant embodiments of the invention, the washing solution in (d) further comprises an acid to lower the pH of the solution.

[0030] In relevant embodiments of the invention, the solvent in (d) is 95% ethanol.

[0031] In relevant embodiments of the invention, the collecting in (e) is via one or more of centrifugation, filtration, and precipitation. In certain aspects, the collecting is via centrifugation and said centrifugation is at a RCF of about 4000-6000 g for about 2-15 minutes.

[0032] In relevant embodiments of the invention, the collected lipids in (e) are resuspended in a solution containing chloroform/methanol/water, wherein the concentration of chloroform in the solution ranges from about 2.5-3.5, wherein the concentration of methanol in the solution ranges from about 1.0-2.0, and wherein the concentration of water in the solution ranges from about 0.1-0.5 (v/v/v). In certain aspects, the concentrations of chloroform, methanol, and water are about 3 : 1.5:0.25, v/v/v.

[0033] In relevant embodiments of the invention, the method further comprises washing the collected lipids of (c) with 50-70% ethanol and collecting the lipids, prior to the washing of (d). In certain aspects, the collecting is via centrifugation and said centrifugation is at a RCF of about 4000-6000 g for about 2-15 minutes.

[0034] In relevant embodiments of the invention, the method further comprises repeating the washing and collecting of (d) and (e) wherein the ethanol is 100% ethanol. In certain aspects, the collecting is via centrifugation and said centrifugation is at a RCF of about 4000-6000 g for about 2-15 minutes. [0035] In relevant embodiments of the invention, the solution in (f) further comprises an ion exchanger. In certain aspects, the ion exchanger is Dowex 50WX8.

[0036] In relevant embodiments of the invention, the method further comprises concentrating the solution of (f). In certain aspects, the concentration is via one or more of evaporation, induced evaporation, dialysis, and precipitation.

[0037] In a fifth embodiment, the invention is directed to a method for obtaining mass spectral data on lipids, comprising:

(a) extracting lipids from a sample via one of the lipid extraction protocols defined herein; and

(b) performing mass spectral analysis to obtain mass spectral data from the extracted lipids of (a).

[0038] In certain aspects, the mass spectral analysis comprises mass spectrometry, mass spectrometry combined with liquid or gas phase separation, ion mobility spectrometry, or mass spectrometry combined with ion mobility spectrometry.

[0039] In certain aspects, the mass spectral analysis is or includes mass spectrometry, and the mass spectrometry comprises MALDI mass spectrometry, electrospray mass spectrometry, ambient ionization mass spectrometry, liquid chromatography mass spectrometry, capillary electrophoresis mass spectrometry, gas chromatography mass spectrometry, or ion mobility mass spectrometry.

[0040] In a sixth embodiment, the invention is directed to a method for identifying one or more microbes in a sample, comprising:

(a) obtaining mass spectral data from lipids extracted from a sample via one of the methods for obtaining mass spectral data on lipids defined herein; and

(b) comparing the mass spectral data obtained in (a) to a library of lipid mass spectral data or one or more mathematical models not comprising a library, corresponding to known microbes;

wherein when the mass spectral data obtained in (a) substantially matches one or mathematical models or one or more lipid mass spectra in the library of (b), the microbe is identified.

[0041] In certain aspects, the identifying is at the level of species, strain, genus, determining the presence of bacteria in the sample, or determining the presence of fungi in the sample. [0042] In certain aspects, the identifying of a microbe comprises determining that the sample is polymicrobial, whether or not one or more microbes are specifically identified in the sample.

[0043] In a seventh embodiment, the invention is directed to a method for detecting a marker for antimicrobial resistance from one or more microbes in a sample, comprising:

(a) obtaining mass spectral data from lipids extracted from a sample via one of the methods for obtaining mass spectral data on lipids defined herein; and

(b) comparing the mass spectral data obtained in (a) to a library of lipid mass spectral data or one or more mathematical models not comprising a library, corresponding to known markers of antimicrobial resistance;

wherein when the mass spectral data obtained in (a) substantially matches one or mathematical models or one or more lipid mass spectra in the library of (b), the instance of antimicrobial resistance is identified.

[0044] In certain aspects, a single comparison in (b) both identifies a microbe and determines antimicrobial resistance.

[0045] In certain aspects, each of the one or more microbes is a bacterium or a fungus.

[0046] The foregoing has outlined rather broadly the features and technical advantages of the present invention in order that the detailed description of the invention that follows may be better understood. Additional features and advantages of the invention will be described herein, which form the subject of the claims of the invention. It should be appreciated by those skilled in the art that any conception and specific embodiment disclosed herein may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes of the present invention. It should also be realized by those skilled in the art that such equivalent constructions do not depart from the spirit and scope of the invention as set forth in the appended claims. The novel features which are believed to be characteristic of the invention, both as to its organization and method of operation, together with further objects and advantages will be better understood from the following description when considered in connection with the accompanying figures. It is to be expressly understood, however, that any description, figure, example, etc. is provided for the purpose of illustration and description only and is by no means intended to define the limits of the invention. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1. Optimization of Extraction and Purification. A. Optimization of Extraction. Contour plot of lipid A ions ratio vs pH, Time. Each black point in the figure indicates a mean (n = 3) of output response from each extraction condition. The darker the grey indicates the higher the ratios of lipid A ions, m/z 1796 and 1919 were determined as canonical and mcr-1 modified E. coli lipid A ions, respectively. Lipid A ions ratio was calculated as sum of ratios of intensity of (1796-1717+1919-1839)/1744. B. Optimization of Purification. Screening of different purification solvents for lipid A extraction. E. coli mcr-1 strain was used for lipid A extractions. Signatural lipid A ions observed at m/z 1716, 1796, 1839, and 1919 were summed. To examine whether the solvents wash out the debris and other lipids, dominant peaks observed in mass range m/z 1300-1400 were normalized to MPL and subtracted from summed lipid A ions' ratio. Data are presented as mean ± SD (n = 3) with one-way ANOVA, Tukey's Multiple Comparison (** significant at p<0.01; *** significant at p<0.001; ns: not significant).

[0048] Figure 2. E. coli mcr-1 extracted by different incubation time.

[0049] Figure 3. LOD comparison between SA and AI methods (dry cell). E. coli mcr-1 strain was used for dry cell LOD study. The LOD for both methods were determined at around 0.82 μg on a MALDI spot.

[0050] Figure 4. Pearson's correlation analysis of SA and AI method sample mass spectra analysis for ESKAPE pathogens differentiation. Glycolipids and/or lipids were extracted using the optimized SA method. Pearson's correlation coefficients were calculated to estimate mass spectra similarity between the SA and AI methods. This resulted a correlation coefficient, r, in the range of [-1,1]. Positive correlations are displayed in circles and coefficients are labeled (the circle on -0.2 for baumaniilE. cloacae is a negative correlation). A coefficient of 0.0 indicates that there is no correlation. To understand strength of the correlation, the guide from Evans [39] suggests for absolute value of r: 0.20-0.39 "weak", 0.40-0.59 "moderate" 0.60-0.79 "strong" and 0.80-1.0 "very strong". The size and color intensity of circle are proportional to the correlation coefficients. Non-statistically significant coefficients of circles were removed and left blank. (* and **) indicate colistin-resistant strains with phosphoethanolamine (PEtN) and aminoarabinose (AraN) addition to the lipid A, respectively.

[0051] Figure 5. Pearson's correlation analysis of ESKAPE pathogens mass spectra between samples produced using the SA method with 10-minute (A.) or 20-minute incubation (B.) and samples produced using the AI method using the same conditions as reported for Figure 4.

[0052] Figure 6. Pearson's correlation analysis of SA method sample mass spectra from 10- minute incubation. Mass spectra were acquired from ESKAPE pathogens, inherently colistin- resistant strains and clinical relevant fungi. Species were compared by calculating a Pearson's correlation coefficient between mass list of ions from triplicate averaged 10-minute incubation mass spectra. Correlation coefficient is considered as a spectrum similarity score, where 1.0 represents an identical match (black Square). White squares represent a score of 0.0 where there is no correlation. (*) indicate mcr-1 colistin-resistant strains with PEtN addition on lipid A and (**) indicate inherently colistin-resistant strains with AraN modified lipid A.

[0053] Figure 7. 30-minute incubation sample mass spectra from colistin-susceptible and resistant strains of P. aeruginosa (A.) and K. pneumonia (B.). The lipid A structures associated with colistin-resistant ions are displayed. 10-minute incubation sample mass spectra from colistin-susceptible and resistant strains of P. aeruginosa (C.) and K. pneumonia (D.), and the inherently colistin-resistant strain Morganella morganii (E.).

[0054] Figure 8. Direct bacterial identification from spiked urine samples. Mass spectra (A. - F.) were generated from 10-minute incubation samples produced using the SA method.

Signature lipid A ions were labeled with m/z values for each species: (A.) A. baumanii, (B.) E. coli, (C.) K. pneumoniae, (D.) P. aeruginosa, and (E.) P. mirabilis. Sterile urine control sample spectrum is shown in (F.). Biotyper scores were obtained by searching against a lipid urine library for each species at different incubation time points (10, 20, and 30 minutes; G.). The dashed line in G. indicates the cutoff for a probably positive identification.

[0055] Figure 9. Polymicrobial uncomplicated and complicated UTI models mass spectra from 30-minute incubated mixed samples. (A.) S. aureus, E. coli a d K. pneumoniae were mixed together in sterile urine to mimic uncomplicated polymicrobial UTI. (B.) C. albicans, E. coli and K. pneumoniae were mixed together in sterile urine to mimic complicated polymicrobial UTI. Species-specific ions are assigned to their respective organisms. Corresponding areas of spectra are color highlighted.

[0056] Figure 10. Individual mass spectra of (A.) S. aureus and (B.) C. albicans are shown to facilitate the identification of species in mixed sample spectral profile. DETAILED DESCRIPTION OF THE INVENTION

I Definitions

[0057] Unless otherwise noted, technical terms are used according to conventional usage. Definitions of common terms in molecular biology may be found, for example, in Benjamin Lewin, Genes VII, published by Oxford University Press, 2000 (ISBN 019879276X); Kendrew et al. (eds.); The Encyclopedia of Molecular Biology, published by Blackwell Publishers, 1994 (ISBN 0632021829); and Robert A. Meyers (ed.), Molecular Biology and Biotechnology: a Comprehensive Desk Reference, published by Wiley, John & Sons, Inc., 1995 (ISBN

0471186341); and other similar technical references.

[0058] As used herein, "a" or "an" may mean one or more. As used herein when used in conjunction with the word "comprising," the words "a" or "an" may mean one or more than one. As used herein "another" may mean at least a second or more. Furthermore, unless otherwise required by context, singular terms include pluralities and plural terms include the singular.

[0059] As used herein, "about" refers to a numeric value, including, for example, whole numbers, fractions, and percentages, whether or not explicitly indicated. The term "about" generally refers to a range of numerical values (e.g., +/- 5-10% of the recited value) that one of ordinary skill in the art would consider equivalent to the recited value (e.g., having the same function or result). In some instances, the term "about" may include numerical values that are rounded to the nearest significant figure.

II. The Present Invention

[0060] As briefly described above, the present invention is generally directed to a novel lipid extraction protocol. The protocol enables preparation of MS-ready bacterial and fungal lipid samples within one hour, with concomitant generation of mass spectra data comparable to existing lipid reference libraries. The protocol avoids the need to use noxious chemicals and a lyophilization step, and can be readily performed by a laboratory technician. The general steps include treatment of bacteria or fungi with a sodium acetate (SA) lysis buffer to release lipids from membranes, and a single step purification. The clinically-feasible characteristics of the protocol presented herein elevates the lipid library platform to a competitive approach in diagnostic microbiology. [0061] Current FDA approved MS-based clinical diagnostic systems are Bruker MALDI Biotyper and bioMerieux VITEK MS. Both systems identify microorganisms via matching the protein fingerprinting profile of an unknown spectrum with given organism in the reference database. For identification, pure colonies are needed for both systems, which means these systems cannot be used in the direct analysis of biofluids, such as blood and urine. Other critical limitations include an inability to identify antibiotic resistant strains and an inability to detect multiple pathogens in a single sample.

[0062] A novel lipid-based approach to microbial identification through MS analysis by the inventors [17] successfully solved the limitations of the protein-based existing systems. The present invention extends this work to a new and simple protocol for sample preparation. As reported herein, this lipid extraction protocol is an improvement on existing methodologies, permitting not only the analysis of biofluids, but also the detection of antibiotic-resistance and polymicrobial identification. It can also be used to isolate lipids from bacterial as well as fungal sources, or used with samples containing cellular debris, and also with samples into which lipids were released. The protocol of the present invention also shortens the time-to-results to a range of minutes and in some cases may be instantaneous. Owing to the increasing difficulty in treatment of patients with suspected antibiotic-resistant infections, rapid identification of an antibiotic-resistant agent is crucial to inform physicians in choosing appropriate antimicrobial therapies rather than prescribing antibiotics empirically. The lipid extraction protocol disclosed herein can be used in techniques that considerably strengthens the diagnostic power of physicians and may serve as a tool in reducing overuse of antibiotics.

Lipid Extraction Protocol (SA Method)

[0063] The lipid extraction protocol of the invention, also termed "the sodium acetate method" and "the SA method", can be generally defined as comprising the steps of adjusting the pH of a sample comprising lipids to a pH of about 1.0-5.0, incubating the mixture at a temperature of up to 130°C, and collecting the lipids from the mixture.

[0064] In a first embodiment, the invention is directed to a method for extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-5.0 using a buffer to form a mixture; (b) incubating the mixture of (a) at a temperature of between about 37°C and 130°C for up to 180 minutes; and

(c) collecting the lipids from the mixture of (b), thereby extracting lipids from a sample. Use of this method results in lipids that are ready for mass spectral analysis without a washing step.

[0065] In certain aspects of the first embodiment, the method further comprises (d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution; and (e) collecting the lipids from the mixture of (d). In certain other aspects of the first embodiment, the method further comprises (d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution; (e) collecting the lipids from the mixture of (d); and (f) resuspending the collected lipids of (e) in a solution of containing at least 20% chloroform and/or at least 10% methanol.

[0066] In a second embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-5.0 using a buffer to form a mixture;

(b) incubating the mixture of (a) at a temperature of between about 37°C and 130°C for up to 180 minutes;

(c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is either neat or in an aqueous solution;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution of containing at least 20% chloroform and/or at least 10% methanol, thereby extracting lipids from a sample.

[0067] In a third embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 2.5-4.5 using a buffer to form a mixture;

(b) incubating the mixture of (a) at a temperature of between about 85°C and 105°C for up to two hours; (c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with a washing solution comprising a solvent, wherein the solvent is ethanol having a concentration of about 90-100%;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution containing at least 20%

chloroform and/or at least 10% methanol, thereby extracting lipids from a sample.

[0068] In a fourth embodiment, the invention is directed to a method of extracting lipids from a sample comprising:

(a) adjusting the pH of a sample comprising lipids to a pH of about 3.5-4.0 using a buffer to form a mixture, wherein the buffer is a sodium acetate, potassium acetate, phosphoric acid, or acetic acid solution;

(b) incubating the mixture of (a) at a temperature of about 100°C for up to about 30 minutes;

(c) collecting the lipids from the mixture of (b);

(d) washing the collected lipids of (c) with 95% ethanol;

(e) collecting the lipids from the mixture of (d); and

(f) resuspending the collected lipids of (e) in a solution of containing

chloroform/methanol/water, 3 : 1.5:0.25, v/v/v, thereby extracting lipids from a sample.

Lipids

[0069] The methods of the present invention, as disclosed and defined herein, can be used in the extraction and analysis of all types and categories of lipids that are produced by microbes such as bacteria and fungi. Categories of lipids that can be extracted from microbes using the lipid extraction protocols of the invention include, but are not limited to, fatty acids,

glycerolipids, glycerophospholipids, sphingolipids, saccharolipids, sterol lipids and prenol lipids. Types of lipids that can be extracted from microbes using the lipid extraction protocols of the invention include, but are not limited to, glycolipids, lipid A, phospholipids, cardiolipins and lipoteichoic acid. Samples

[0070] It will be clear that in each of the steps of the protocol, selected characteristic can be varied without departing from the important attributes of the protocol. In relevant embodiments and aspects of the invention, the samples to which the methods are applied can be any sample that contains or is suspected of containing a lipid produced by a microbe, such as a bacterium or fungus. To name only a few examples, the samples can be obtained from a subject or a patient, such as a human or other mammal, including apes, mice, rats, dogs, cats, horses, cattle, or other animal upon or within which microbes can be found. Such samples can include biological fluids, such as urine, blood, serum, sweat, semen, spinal fluid, etc. The sample can be obtained from a laboratory source, such as liquid cultures or colonies from a cell culture plate suspended in solution, or a medical swab. The sample can also be obtained from locations such as hospitals and GMP manufacturing plants, e.g. a swab of an environmental/device surface to monitor bacterial contamination. The sample can be an environmental source, such as a coating on a medical device, soil or water reservoir.

[0071] The sample can be present in different material states, including lyophilized, pelletized, or an otherwise dry or semi-dry sample, or be in a liquid state. As an example, the sample could be dried fluids, such as dried blood spots.

[0072] The sample to which the methods are applied can include whole microbes, such as a living, dead or mixed populations of such microbes. The sample can also include portions of microbes, such as microbes that have be lysed or otherwise disrupted. The sample can further include lipids that are released from microbes, such as lipids that are collected from an active cell culture of microbes. It will be evident that the sample can also include combinations of one or more of living (vegetative state or dormant) and dead microbes, portions of microbes, and released lipids.

[0073] In relevant embodiments and aspects of the invention, the sample comprises one or more microbes, or portions thereof.

[0074] As indicated above, the samples contain or are suspected of containing a lipid produced by a microbe, such as a bacterium or fungus. The identity of the bacterium or fungus is immaterial to the invention. The only limitation on the bacterium or fungus is that is produce one or more lipids. [0075] In relevant embodiments and aspects of the invention, the sample comprises one or more microbes, or portions thereof.

[0076] In relevant embodiments and aspects of the invention, the one or more microbes are a bacterium or bacteria, a fungus or fungi, or combinations thereof.

[0077] In relevant embodiments and aspects of the invention, the one or more microbes are: one or more species of bacteria; one or more species of fungi; or one or more species of bacteria and one or more fungi.

[0078] It will be evident that the methods of the invention can be practiced directly on a sample that contains or is suspected of containing a lipid produced by a microbe, such as a bacterium or fungus. However, depending on the nature of the sample, lipids in the sample, whether associated with a microbe or portion of a microbe, or free in the sample, can be collected from the sample prior to pH adjustment. While the parameters of the collecting will vary depending on the identity, volume and concentration of the sample, suitable means for collecting the lipids from the sample include one or more of centrifugation, filtration, and precipitation. In relevant embodiments and aspects of the invention, the collecting is via centrifugation, where centrifugation is at a relative centrifugal force (RCF) of about 3000-6000 g for about 10-20 minutes. As a non-limiting example, the collecting is via centrifugation at a relative centrifugal force (RCF) of about 4500 g for about 15 minutes. After centrifugation, the lipid-containing pellet is kept and further processed while the supernatant is discarded. pH Adjustment

[0079] As indicated above, one of the first steps of the methods of the invention is adjusting the pH of a sample containing or suspected of containing a lipid produced by a microbe to a pH of about 1.0-5.0. Various buffers can be used to adjust the pH of the sample, including any weak acid that results in a pH in the range of 1.0-5.0 in combination with the various temperatures and durations of the incubation. Suitable buffers include, but are not limited to, a mixture of sodium acetate and acetic acid, a buffer comprising a mixture of potassium acetate and acetic acid, or a buffer comprising a mixture of sodium phosphate and phosphoric acid. Specific examples include, but are not limited to, a sodium acetate, potassium acetate, phosphoric acid, and acetic acid solution. As an example, the buffer is sodium acetate. The concentration of the buffer will vary based on the identity of the buffer. As an example, when the buffer is sodium acetate, the concentration of the buffer will be between about 5-200 mM. In one example, the buffer is 100 mM sodium acetate.

[0080] The particular pH that will result in the sample may vary depending on such factors as the known or suspect source of the lipids, the temperature at which the sample will be incubated once the pH is adjusted, and the length of time of the incubation, among other factors. In relevant embodiments and aspects of the invention, the pH is adjusted to about 1.0-5.0, 1.5- 5.0, 2.0-5.0, 2.5-5.0, 3.0-5.0, 3.5-5.0, 4.0-5.0, 4.5-5.0, 2.0-4.5, 2.0-4.0, 2.5-4.5, 2.5-4.0, 3.0-4.5, 3.0-4.0, 3.5-4.5, or 3.5-4.0. In relevant embodiments and aspects of the invention, the pH is adjusted to about 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, or 5.0.

Incubation

[0081] As indicated above, one of the next steps of the methods of the invention is incubating the lipids at an elevated temperature for a prior of time. The temperature at which the incubation is conducted may vary depending on such factors as the known or suspect source of the lipids, pH to which the sample has been adjusted, the identity of the buffer use to adjust the pH of the sample, and the length of time of the incubation, among other factors. In relevant embodiments and aspects of the invention, the temperature is about 50-130°C, 50-120°C, 50- 110°C, 50-100°C, 50-90°C, 50-80°C, 60-130°C, 60-120°C, 60-110°C, 60-100°C, 60-90°C, 60- 80°C, 70-130°C, 70-120°C, 70-110°C, 70-100°C, 70-90°C, 70-80°C, 80-130°C, 80-120°C, 80- 110°C, 80-100°C, 80-95°C, 80-90°C, 85-130°C, 85-120°C, 85-110°C, 85-100°C, 80-95°C, 80- 90°C, 90-130°C, 90-120°C, 90-110°C, 90-100°C, or 90-95°C. In relevant embodiments and aspects of the invention, the temperature is about 50°C, 55°C, 60°C, 65°C, 70°C, 75°C, 80°C, 85°C, 90°C, 95°C, or 100°C.

[0082] The length of time for which the incubation is conducted may vary depending on such factors as the known or suspect source of the lipids, pH to which the sample has been adjusted, the identity of the buffer use to adjust the pH of the sample, and the temperature of the incubation, among other factors. In relevant embodiments and aspects of the invention, the length of time is up to about 180, 170, 160, 150, 140, 130, 120, two hours, 110, 100, 95, 90, 85, 80, 75, 70, 65, 60, one hour, 55, 50, 45, 40, 35, 30, 29, 28, 27, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 minute. In relevant embodiments and aspects of the invention, the length of time is about 5-30 minutes, 10-30 minutes, 15-30 minutes, 5-25 minutes, 10-25 minutes, 15-25 minutes, 5-20 minutes, 10-20 minutes, 15-20 minutes, 5-15 minutes, 10-15 minutes, or 5-10 minutes. In relevant embodiments and aspects of the invention, the length of time is about 30, 29, 28, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 minute.

[0083] Optionally, the lipids can be mixed during the incubation, such as via simple hand mixing of the container or vortexing of the container when the sample is present in a microcentrifuge tube or other suitable container. The mixing can take place about every 5-10 minutes during the incubation.

Collecting after Incubation

[0084] After the incubation, the lipids are collected from the sample. The means used to collect the lipids from the incubated sample may vary and be any of those known to one having ordinary skill in the art for collecting lipids from a sample. Exemplary methods include, but are not limited to, one or more of centrifugation, filtration, adsorption, and precipitation.

[0085] When centrifugation is used to collect the lipids after incubation, a force and time sufficient to pellet the lipids but not so great as to make resuspension difficult or to damage the lipids is used. An acceptable RCF is between about 7000-9000 g. The duration of the

centrifugation will generally span about 2-15 minutes. As an example, the centrifugation may be at a RCF of about 8000 g for about 5 minutes. After centrifugation, the lipid-containing pellet is kept and further processed while the supernatant is discarded.

Cooling

[0086] After the incubation, the sample may optionally be cooled. When the sample is cooled, the temperature to which the sample is cooled with generally be less than about 50°C. As non-limiting examples, the sample may be cooled to less than about 49°C, 48°C, 47°C, 46°C, 45°C, 44°C, 43°C, 42°C, 41°C, 40°C, 39°C, 38°C, 37°C, 36°C, 35°C, 34°C, 33°C, 32°C, 31°C, 30°C, 29°C, 28°C, 27°C, 26°C, 25°C, 24°C, 23°C, 22°C, 21°C, 20°C, 19°C, 18°C, 17°C, 16°C, 15°C, 14°C, 13°C, 12°C, 11°C, or 10°C, or cooler. Is a further example, the sample may be cooled to about room temperature. The cooling may take place before, during, and/or after the collecting of the lipids from the incubated sample. Optional Washing

[0087] After the collecting of the lipids from the incubated sample, and the optional cooling step, the lipids are optionally washed using a washing solution comprising a solvent. Washing serves to remove non-lipid materials. In addition to the solvent, the washing solution may also comprise an acid to lower the pH of the solution. Exemplary solvents include, but are not limited to, 90-100% ethanol, 90-100% methanol, and a chloroform/methanol mixture (1 :4, 1 :3, 1 :2, and 1 : 1, v/v). The solvent may be used need, i.e. in an undiluted state, or as an aqueous solution.

[0088] When ethanol is used as the solvent, for example, the ethanol may have a

concentration of about 90-100%). As a specific example, the washing may be conducted using 95% ethanol.

Collecting after Washing

[0089] If the optional washing step is performed, the lipids are again collected. The means used to collect the lipids from the washed sample may vary and be any of those known to one having ordinary skill in the art for collecting lipids from a sample. Exemplary methods include, but are not limited to, one or more of centrifugation, filtration, adsorption, and precipitation.

[0090] When centrifugation is used to collect the lipids after washing, an acceptable RCF is between about 4000-6000 g. The duration of the centrifugation will generally span about 2-15 minutes. As an example, the centrifugation may be at a RCF of about 5000 g for about 5 minutes. After centrifugation, the lipid-containing pellet is kept and further processed while the supernatant is discarded.

Resuspending

[0091] After the lipids are collected, whether after incubation, the optional cooling, or the optional washing, the lipids are resuspended in a solution comprising at least 20%> chloroform, or a solution comprising at least 10%> methanol, or a solution comprising at least 20%> chloroform and at least 10%> methanol. In relevant embodiments and aspects of the invention, the washed and collected lipids are resuspended in a solution containing chloroform/methanol/water, wherein the concentration of chloroform in the solution ranges from about 2.5-3.5, wherein the concentration of methanol in the solution ranges from about 1.0-2.0, and wherein the concentration of water in the solution ranges from about 0-0.5 (v/v/v). In relevant embodiments and aspects of the invention, the solution is chloroform/methanol/water, 3 : 1.5:0.25, v/v/v.

[0092] Depending on the manner in which the resuspended lipids are used, the solution in which the lipids are resuspended may further comprise an ion exchanger such as, but not limited to, Dowex 50WX8.

[0093] Depending on the manner in which the resuspended lipids are used, the resuspended lipids may be concentrating in the solution in which they were resuspended. Acceptable means for concentrating a solution comprising lipids and one or more of chloroform and methanol will be known to one of ordinary skill. Acceptable means include, but are not limited to one or more of evaporation, induced evaporation, dialysis, and precipitation.

Optional Initial Washing and Collecting

[0094] Depending on such factors as the volume and concentration of the sample to which the methods of the invention are performed, and the known or suspect source of the lipids, among other factors, an optional initial washing and collecting of the lipids can be performed. The optional initial washing and collecting takes place after the lipids are collected from the incubated sample and before the washing with a washing solution comprising a solvent.

[0095] When performed, the lipids are washed with 50-70% ethanol in the optional initial washing. As a specific example, the optional initial washing may be conducted using 60% ethanol.

[0096] After the optional initial washing, the lipids are again collected. The means used to collect the lipids from the washed sample may vary and be any of those known to one having ordinary skill in the art for collecting lipids from a sample. Exemplary methods include, but are not limited to, one or more of centrifugation, filtration, adsorption, and precipitation.

[0097] When centrifugation is used to collect the lipids after the optional initial washing, an acceptable RCF is between about 4000-6000 g. The duration of the centrifugation will generally span about 2-15 minutes. As an example, the centrifugation may be at a RCF of about 5000 g for about 5 minutes. After centrifugation, the lipid-containing pellet is kept and further processed while the supernatant is discarded. Optional Secondary Washing and Collecting

[0098] Depending on such factors as the volume and concentration of the sample to which the methods of the invention are performed, and the known or suspect source of the lipids, among other factors, an optional secondary washing and collecting of the lipids can be performed. The optional secondary washing and collecting takes place after the washing with a washing solution comprising a solvent and the subsequent collecting.

[0099] When performed, the lipids are washed with 90-100% ethanol in the optional secondary washing. As a specific example, the optional secondary washing may be conducted using 100% ethanol.

[00100] After the optional secondary washing, the lipids are again collected. The means used to collect the lipids from the washed sample may vary and be any of those known to one having ordinary skill in the art for collecting lipids from a sample. Exemplary methods include, but are not limited to, one or more of centrifugation, filtration, adsorption, and precipitation.

[00101] When centrifugation is used to collect the lipids after the optional secondary washing, an acceptable RCF is between about 4000-6000 g. The duration of the centrifugation will generally span about 2-15 minutes. As an example, the centrifugation may be at a RCF of about 5000 g for about 5 minutes. After centrifugation, the lipid-containing pellet is kept and further processed while the supernatant is discarded.

[00102] After the optional secondary washing and collection of the lipids, the lipids are resuspended in the manner discussed and defined above.

Methods for Obtaining Mass Spectral Data on Extracted Lipids

[00103] As indicated above, the lipids that are extracted from a sample using the lipid extraction protocols of the invention can be assayed using further techniques. One such technique is mass spectrometry.

[00104] The invention thus includes methods for obtaining mass spectral data on lipids. Such methods comprise the following steps:

(a) extracting lipids from a sample via one of the lipid extraction protocols defined herein; and

(b) performing mass spectral analysis to obtain mass spectral data from the extracted lipids of (a). [00105] The mass spectral analysis can be conducted via any of the means commonly known for performing mass spectral analysis on a substance, such as a lipid. Acceptable means include, but are not limited to, spectrometry, mass spectrometry combined with liquid or gas phase separation, ion mobility spectrometry, and mass spectrometry combined with ion mobility spectrometry. Acceptable means of mass spectrometry include, but are not limited to, MALDI mass spectrometry, electrospray mass spectrometry, ambient ionization mass spectrometry, liquid chromatography mass spectrometry, capillary electrophoresis mass spectrometry, gas chromatography mass spectrometry, or ion mobility mass spectrometry.

[00106] In one embodiment of the invention, resuspended lipids prepared using one of the lipid extraction protocols defined herein are mixed with a matrix and spotted on to a MALDI target. Suitable matrices including, but are not limited to, norharmane, dihydroxybenzoic acid (DHB), 2,4,6-trihydroxyacetophenone (THAP), 5-chloro-2-mercaptobenzothiazole (CMBT), and 6-aza-2-thiothymine (ATT).

[00107] The mixture of the resuspended lipids and the matrices can vary in ratio, from about 2: 1 to about 1 :2 (v/v). In certain aspects of the invention, the ratio is about 2: 1, about 1.5: 1, about 1 : 1, about 1 : 1.5, or about 1 :2.

[00108] When norharmane is used as the matrix, suitable concentrations include a solution of about 10 mg/mL to a saturated solution.

[00109] In a non-limiting example, the instrument used to perform the mass spectral analysis is a Bruker Microflex LRF MALDI-TOF MS (Bruker Daltonics Inc., Billerica, MA, USA). In certain embodiments of the method, the instrument is operated in the reflectron mode or linear mode. In certain embodiments of the method, the laser energy ranges from 50% to 100%, and will vary depending on the amount lipids on a MALDI spot. In certain embodiments of the method, sufficient shots are summed to generate a spectrum where lipids ions are observed with signal to noise (S/N) greater than 3. As a non-limiting example, mass spectral analysis is performed using a Bruker Microflex LRF MALDI-TOF MS, operated in the reflectron mode, with mass spectra acquired at 72% laser power with 900 shots summed. Methods for Identifying a Microbe in a Sample

[00110] The data resulting from the mass spectral analysis of the lipids extracted from microbes can be used to identify the microbes in a sample, for example by comparing the data obtained to libraries of mass spectral analyses on known microbial lipids.

[00111] The invention thus includes methods for identifying one or more microbes in a sample. Such methods comprise the following steps:

(a) obtaining mass spectral data from lipids via one of the methods for obtaining mass spectral data on lipids defined herein; and

(b) comparing the mass spectral data obtained in (a) to a library of lipid mass spectral data or one or more mathematical models not comprising a library, corresponding to known microbes;

wherein when the mass spectral data obtained in (a) substantially matches one or mathematical models or one or more lipid mass spectra in the library of (b), the microbe is identified.

[00112] In the methods for identifying one or more microbes in a sample, the identifying may be at one or more of the level of species, strain, or genus. The method may also be used to determine the presence of bacteria in a sample, or determine the presence of fungi in a sample.

[00113] In the methods for identifying one or more microbes in a sample, the identifying may also comprise determining that that sample is polymicrobial. Such determining can occur whether or not the microbes in the sample are specifically identified.

[00114] In the methods for identifying one or more microbes in a sample, the microbe may be a bacterium or a fungus.

Methods for Characterizing a Microbe in a Sample

[00115] The data resulting from the mass spectral analysis of the lipids extracted from microbes can also be used to detect a marker for antimicrobial resistance, for example by comparing the data obtained to libraries of mass spectral analyses on known microbial lipids.

[00116] The invention thus includes methods for detecting a marker for antimicrobial resistance from one or more microbes in a sample. Such methods comprise the following steps (a) obtaining mass spectral data from lipids via one of the methods for obtaining mass spectral data on lipids defined herein; and (b) comparing the mass spectral data obtained in (a) to a library of lipid mass spectral data or one or more mathematical models not comprising a library, corresponding to known markers of antimicrobial resistance;

wherein when the mass spectral data obtained in (a) substantially matches one or mathematical models or one or more lipid mass spectra in the library of (b), the instance of antimicrobial resistance is identified.

[00117] In the methods for detecting a marker for antimicrobial resistance from one or more microbes in a sample, the single comparison in (b) can both identify a microbe as well as determine whether the microbes exhibit antimicrobial resistance.

[00118] In the methods for detecting a marker for antimicrobial resistance from one or more microbes in a sample, the microbe may be a bacterium or a fungus.

III. Examples

[00119] As reported in the following paragraphs, the lipid extraction protocol of the present invention, also term the sodium acetate (SA) method as a short hand, was analyzed using 18 different microorganisms, including clinical important ESKAPE pathogens and counterpart colistin-resistant mcr-1 strains, naturally colistin-resistant species and certain fungal species. The selection of these species was due to their prevalence in nosocomial infections and rapid antibiotic resistance acquisition [28]. The experiments described herein also examined the robustness of the SA method on direct analysis of biofluids such as urine.

A. Materials and Methods

1. Bacterial strains

[00120] Clinical relevant bacterial strains were used, including ESKAPE pathogens, corresponding mcr-1 mutant strains, and inherently colistin-resistant species A complete list is provided in Table 1. Strain identities were confirmed by microbiological culture or Bruker Biotyper system. All bacterial species were grown in liquid Luria-Bertani (LB) medium overnight with shaking after inoculated from solid agar plates, mcr-1 mutant species were cultured in LB media where 50 μg/mL gentamicin was added. Table 1 - Species tested by the SA method

2. Reagents and Materials

[00121] All solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Commercially available monophosphoryl lipid A (MPL; Avanti Polar Lipids, Inc., Alabaster, AL, USA) was used as an internal standard for Design of Experiment (DOE) study. A stock solution of MPL was prepared in a 1 :2, v/v mixture of methanol/chloroform at a concentration of 1 mg/mL. Sodium acetate (Sigma-Aldrich, Cat # S8750, lot # SLBL-4426V) buffer was prepared by dissolving in endotoxin free water (Quality Biological Inc., MD, USA, Cat # 1 18-162-101) to give 100 mM. The pH was adjusted by adding acetic acid and checked by a pH meter (Fisher Scientific Accumet AP50, Hampton, H, USA).

3, SA Method for Lipid A Microextraction

[00122] 1 mL overnight cell culture was pelleted and resuspended in 400μΙ. of lOOmM sodium acetate (SA) buffer (pH 4.0) for 10 - 30 minutes with incubation at 100°C. Samples were vortexed every 5 - 10 minutes. After incubation, samples were cooled on ice to room temperature and centrifuged at 8,000 x g for 5 minutes. Supernatants were discarded and pellets were washed with 1 mL 95% ethanol. The insoluble lipid A was extracted and reconstituted in 50-100 μΐ _^ chloroform/methanol/water (3 : 1.5 : 0.25, v/v/v) mixtures and centrifuged at 5,000 x g for 5 minutes. 0.75 μΐ _^ of the supernatant was spotted on a MALDI plate well with 0.75 μΐ _^ pre- spotted 10 mg/mL norharmane (Sigma-Aldrich, St. Louis, MO, USA) matrix for MS analyses. In DOE study, after chloroform/methanol/water mixture extraction, MPL was spiked as an internal standard at the final concentration of 20 μg/mL.

4. AI Method for Lipid A Microextraction

[00123] Overnight cultures (1 -5 mL) in LB medium were harvested, processed and converted to lipid A/LTA by the optimized hot ammonium isobutyrate-based protocol [17,33]. The original protocol was described in detail by EI Hamidi et al. [21]. Briefly, bacterial pellets were treated with a 400 μΕ of 5 :3 mixture of 70% (v/v) isobutyric acid/1 M ammonium hydroxide and incubated at 100°C for 1 hour. Reaction mixtures were spun down at 2,000 x g for 15 minutes and supernatants were transferred to clean tubes, combined with 1 : 1 ratio of distilled water, and lyophilized overnight. The lyophilized residues containing lipid A molecules were washed twice with 1 mL methanol and then resuspended in 50-100 μΕ of a 3 : 1.5 :0.25

chloroform/methanol/water solvent mixture. Aliquots of 1 μΐ, each were spotted directly onto a MALDI stainless steel target plate with 1 μΕ 10 mg/mL norharmane matrix for MALDI-TOF MS analyses. 5. MALDI-TOF MS Data Acquisition

[00124] Mass spectra were recorded in negative ion mode using a Bruker Microflex LRF MALDI-TOF MS (Bruker Daltonics Inc., Billerica, MA, USA) operated in the reflectron mode. For SA method, each mass spectrum was acquired at 72% laser power with 900 shots summed per spectrum. Each sample was acquired triplicate spectra for further statistical analyses.

6. DOE Data Analysis

[00125] Mutant E. coli strains expressing mcr-1, a gene encoding a phosphoethanolamine (PEtN) transferase that confers resistance to colistin, were used for method optimization. Two factors with multiple levels: incubation time (10, 20, 30, 60, and 120 minutes) and pH (3.0, 3.5, 4.0, and 4.5) were considered as key parameters for optimization, which was carried out using a DOE approach. The ratios were calculated by referencing intensities of lipid A signature ions to the spiked MPL ion intensity, and then were used as y output for DOE analysis. Minitab 17 (Minitab Inc., State College, PA, USA) DOE function was used to quantitatively compare ratios across combinatorial tested conditions. The optimal condition was determined by the DOE results.

7. Bacterial Spiked Urinary Sample Analysis

[00126] Sterile urine was purchased from LEE Biosolutions Inc. (Maryland Heights, MO, USA, cat # 991-03-P, lot # R94503). Commonly identified UTI bacterial species were selected and cultured in 5 mL liquid LB medium. After overnight culture, bacterial cells were measured as colony forming units (CFU) using a standard plating method. 1 mL bacterial liquid culture was seeded into 9 mL pre-warmed (37°C) sterile urine. Seeded urine samples were sat in a warm room for few minutes. Then urine samples were spun down at 4,000 x g for 5 minutes to obtain bacterial pellets. SA method was used to process bacterial pellets as described.

8. Statistical Analysis

[00127] MALDI mass spectra were imported and processed using the MALDIquant (vl .16.4) and MALDIquantForeign (vO.11) R packages [34] in R (v3.4.2). Each mass spectrum Intensities were square root transformed and smoothed with the Savitzky-Golay filter [8]. Mass spectra were baseline removed using the SNIP method [35] over 100 iterations, and then aligned. Lipid ions were detected by selecting m/z with a signal-to-noise ratio greater than 8.0, with noise calculated by the median absolute deviation and binning matching mass peaks between spectra using a 0.5 m/z tolerance. For heat map comparison and correlation analysis, a Pearson's correlation coefficient was calculated from pairwise vectors created from a mass list of ions from each mass spectrum. All scripts used for this analysis are freely available and can be found in the following GitHub repository: https://github.com/Tao-Liang-UMB/Rapid-Extraction-Project.gi t.

B. Results

[00128] The initial focus of these studies, using the materials and methodologies described above, was on protocol development for bacterial lipid A extraction, which was designed to also extract simultaneously lipids from Gram -positive bacteria and fungi. Lipid A can be liberated and extracted either from LPS by SDS promoted mild-acid hydrolysis procedure [36] or directly from bacterial cells [21]. However, using LPS as a starting material is time-consuming and requires the use phenol. Furthermore, SDS causes strong interference to the mass signal, and methods using SDS require multiple steps of purification and cleaning. Direct bacterial cell extraction is a good option. However, the overnight lyophilization step required in the ammonium isobutyrate (Al)-based protocol [17,33] is a time-consuming process, which hinders reducing the required time to a clinically-practical range (ideally less than an hour). Given the advantages of extracting lipid A directly from bacterial cells, the rapid, sodium acetate (SA)- based method for lipid A isolation presented herein was developed. This method was systemically optimized using a DOE approach and evaluated by comparing with the most commonly used ΑΙ-based method. Additional studies discussed below demonstrated that Gram- positive bacterial lipids as well as fungal membrane lipids can be extracted by the same SA method. Proof of concept was further established using urinary samples spiked with common UTI pathogens.

1.1. Rapid SA Method Optimization: Determination of Optimal Extraction Conditions using DOE

[00129] The extraction method was divided into two modules for optimization: (i) incubation module focused on sodium acetate buffer pH and incubation time length, and (ii) post- purification module focused on solvent selection to remove cell debris and unwanted molecules. The goal was to have a protocol that has minimum experimental steps while maintain the quality of lipid extracts for good MS analysis. [00130] For the incubation module optimization, two key parameters (sodium acetate buffer pH, incubation time) were selected for optimization based on pilot experiment results. According to results from a pilot study, initial levels for pH (3.0, 3.5, 4.0, and 4.5) and incubation time (10, 20, 30, 60, and 120 minutes) were determined. Then a 4 x 5 general full factorial design was created in Minitab 17 DOE function. With triplicate biological replicates, a total of 60 discrete experiments were carried. Each biological replicate was blocked in the general full factorial design for statistical analysis. E. coli lipid A ions were normalized to MPL and ratios of each pH and time combination were imported into DOE function for further analysis. Contour plot (Fig. 1A) shows that extraction conditions (pH 4.0, 30 min; pH 4.5, 60 min; and pH 4.0, 4.5, 120min) fall in the darkest grey region, indicating lipid A ions ratios in the range of 3.0 to 4.0. By analyzing from mass spectra, ions at m/z 1716 and 1839 were assigned as fragment ions due to a loss of phosphate group (Δ m/z = 80) from precursor ions m/z 1796 and 1919 respectively [33]. Regarding lipid A ions ratio calculation, each lipid A ion intensity including fragments ions was normalized to MPL intensity. Lipid A ions ratio was calculated as sum of ratios of Intensities (1796-1717+1919-1839)/1744. This model allows determination of a condition that would extract the highest amount of intact lipid A molecules while preserve the labile moieties such as phosphate, PEtN or aminoarabinose (AraN). As expected, when strong acidic buffer was used (pH 3.0 and 3.5), more lipid A fragments were generated due to the loss of labile moiety from lipid A, such as terminal phosphate groups. The pH 4.0 sodium acetate buffer falls in the range of mild acid [36] but efficiently liberates lipid A from bacterial membrane. Longer incubation time increases the yield of lipid A extraction. Although other data points (pH 4.5, 60 minutes; and pH 4.0, 4.5, 120 minutes) also extracted relative high amount of lipid A, with the interest of extraction time, the condition pH 4.0, 30 minutes was selected as the optimal extraction condition for further experiments.

[00131] It was also found that with only a 10-minute incubation, pH 4.0, good quality mass spectra were obtained where lipid A signature ions were observed including antibiotic-resistant associated ions (Fig. 2). Therefore, the total protocol time can be reduced to less than 30 minutes taking into consideration of purification, lipid A recovery and data acquisition.

[00132] For the post-purification module optimization, Escherichia coli (E. coli) mcr-1 was used as the model strain for optimization. In solvent screening experiment, eight commonly used solvent/mixtures for lipids isolation were selected. After 30 minute incubation, 1 mL of each solvent was used to wash the pellet and followed by lipid A extraction step. MPL was spiked as an internal standard for ion intensities normalization to statistically compare different solvent purification efficiency. Lipid A ions intensities were examined by the MALDI-TOF MS. 95% ethanol as the purification solvent gave rise to the highest ratios of lipid A ions compared to 'no wash' conditions (Fig. IB). 95% methanol and mixtures of chloroform: methanol 3 : 1 v/v also show statistical significant yield improvements, but not high as 95% ethanol. Thus, 1 mL of 95% ethanol was chosen as the purification solvent after chemical reaction.

1.2. SA Method Optimization: Limit of Detection (LOP)

[00133] In order to determine the sensitivity of the method, the limit of detection (LOD) was defined using two quantitative measurements: lyophilized cell mass and colony forming unit (CFU). The sensitivity of the SA method was compared with the AI method. Identical amounts of lyophilized cells were weighed and aliquoted into two portions for each method for lipid A extraction. Bacterial cell pellets were processed and serial diluted for LOD testing. Fig. 3 shows mass spectra from serially diluted samples.

[00134] SA and AI methods LOD was found to be around 0.82 μg/spot, which corresponding to about 50 μg lyophilized cells as starting materials. This LOD is consistent with El Hamidi et al. [21], who reported sensitivity in the range of 50 - 100 μg.

[00135] As clinicians are often interested in how many cells are needed to generate high quality spectra, the CFU was further used as a quantitative measurement to estimate the number of cells in samples. After enumeration, the same volume of liquid culture was spun down to get bacterial pellets. The SA method was used to process the samples, which were serially diluted after extraction. Each dilution was tested by MALDI-TOF MS and the LOD was determined by signal-to-noise (S/N) ratio > 3 of lipid A ions of interest {m/z 1796). The LOD was found to be 2.3 x 10 ⁶ CFU deposited on a MALDI spot for both methods (data not shown). The 10-minute incubation LOD is higher than 30-minute incubation LOD, which is attributed to fewer lipid A molecules released in the shorter reaction period. SA Method Evaluation

[00136] Once the SA method was optimized and optimal extraction conditions were determined, the SA method was incorporated into a lipids platform for bacterial identification. A lipids library was constructed as described in detail elsewhere [17]. The mass spectra included in the library were acquired from samples produced using the AI method. Although Leung et. al.

[17] demonstrated that the lipid approach is much faster than traditional bacterial identification methods, the whole workflow takes about 18 hours to obtain final identification results. The long sample preparation process becomes a bottleneck in clinical implementation of this platform. Here, samples produced using the SA method of the present invention were evaluated to ensure that: i) mass spectra generated from samples produced using the SA method are similar to those generated from samples produced using the AI method, which can be cross-searched in the lipid library; and ii) mass spectra generated from samples produced using the SA method for different bacterial species can be differentiated from each other.

2.1. SA Method Mass Spectra Show High Similarity to AI Method Mass Spectra

[00137] The lipids library has been successfully built in the Biotyper platform with AI method sample mass spectra and demonstrated the ability of identifying ESKAPE pathogens. To evaluate the robustness of SA method samples, the spectral similarity between the SA and AI method samples was objectively calculated using Pearson' s correlation, as used in SEQUEST [37]. Although several statistical methods are available for mass spectra similarity comparison, Liu et al. [38] reported that Pearson' s correlation is the most robust method when compared with other popular methods such as dot product and counting number of shared peaks in two spectra. ESKAPE pathogens were also selected for comparison as these are clinical relevant species and the major component of lipid library. A threshold of signal-to-noise ratio of 8.0 was set to select m/z channels from each spectrum after baseline subtraction and intensities normalization. The list containing selected ions was used to calculate Pearson's correlation as a measure of spectrum similarity. Fig. 4 shows the Pearson's correlation between the two methods at the optimal extraction condition (pH 4.0, 30 minutes). All ESKAPE pathogens can be distinguished from each other as shown in the correlation plot. In addition, 8 out of 10 strains' Pearson's correlation coefficients are higher than 0.6, a value usually considered as 'strong correlation' [39]. In particular, antibiotic-resistance strains of K. pneumoniae, A. baumanii and P. aeruginosa are differentiated from their counterpart susceptible strains. As it was observed that 10-minute incubations could generate high-quality mass spectra for sample produced by the SA method, 10- minute and 20-minute correlation plots (Fig. 5A, 5B) were included to further demonstrate the rapidness of SA method. Interestingly, two strains (A. baumanii mcr-1 and K. pneumoniae) showed moderate positive correlation to AI spectra, while five strains show strong correlation from 10-minute incubation ESKAPE pathogens correlation plot.

2.2. SA method mass spectra can be used for bacterial identification

[00138] As the AI method sample mass spectra have been successfully used to construct the lipid library, it was further assessed whether SA method sample mass spectra from different bacterial species were distinct from each other. Although 30-minute incubation is the optimal condition, mass spectra was examined for samples from 10- and 20-incubation times as well to validate the hypothesis that shorter incubation times can generate enough species-specific lipid ions to allow microbial identification.

[00139] In this case, each strain ion list obtained by processing SA method sample mass spectra in R was compared with each other and itself. A heat map of selected representative species was represented in Fig. 6 including 11 Gram-negative species with corresponding antibiotic-resistant strains, three Gram-positive and two fungal species. Fig. 6 shows that SA method sample mass spectra are able to distinguish all tested species from one another, in addition to distinguish antibiotic susceptible and resistant strains, including PEtN lipid A modification of A. baumanii mcr-1, E. coli mcr-1, K. pneumoniae mcr-1, and P. aeruginosa mcr- 1 as well as aminoarabinose (AraN) type of modification C. sakazakii, K. pneumoniae colistin R**, M. morganii, P mirabilis, and S. marcescens. Different growth temperatures at 25°C and 37°C were investigated to mimic insect and mammalian growth conditions, respectively, resulting lipid A structure alternation [40,41]. Fig. 6 shows a clear distinction for different culture temperatures for species F. novicida and Y. pestis.

[00140] Notably, mass spectra from 10-minute group are clearly differentiated from each other not only in species level but also between bacteria and eukaryotic fungi (Fig. 6). For 20- minute and 30-minute groups (data not shown), the same trend holds for the selected strains where clearer differences are shown. This may result from more species-specific lipid ions released during longer incubation.

3, Detection of Antimicrobial Resistance

[00141] Mass spectra from Fig. 6 was analyzed to further evaluate the ability of the SA method samples in generating distinguishable mass spectra between colistin-susceptible and - resistant strains. Antibiotic-resistant strains were selected and determined by a minimum inhibitory concentration method that was described in detail elsewhere [17,33]. Four Gram- negative species were examined and their antibiotic-resistant marker ions were observed in spectra. For example, P. aeruginosa mcr-1 strains clearly show unique ions at m/z 1569 which are not found in susceptible strains (Fig. 7A). The mass shift from m/z 1446 to 1919 was caused by a PEtN (Am/z 123) addition to lipid A. Monophosphorylated modified lipid A ion is detected at m/z 1489, resulting from a loss of phosphate moiety from PEtN modified lipid A (m/z 1569). This ion is not found in a susceptible strain, which is also considered as a diagnostic marker for microbial resistance [17,33]. In addition, another type of colistin-resistant mechanism that is related to AraN addition to lipid A structure was examined. Tested strains included K.

pneumoniae TBE 805, and inherently colistin-resistant strains S. marcescens, C. sakazakii, M. morganii, and P. mirabilis. Fig. 7B shows an AraN (Am/z 131) attached to the terminal phosphate group of K. pneumoniae lipid A (m/z 1824), resulting in the modified lipid A structure (m/z 1955). Similar results are found when samples from 10-minute incubations were used (Fig. 7C - P. aeruginosa; Fig. 7D - K. pneumoniae). Inherently colistin-resistant strains were also observed pairs of ions with the mass shift of Am/z 131 that are considered as AraN modification. For example, a representative mass spectrum from Morganella morganii showed the AraN- modified lipid A at m/z 1927 (Fig. 7E).

4. Direct Bacterial Identification from Urinary Samples

[00142] In order to extend the application of the approach presented herein to biofluids, clinical relevant pathogens (A. baumanii, E. coli, K. pneumoniae, P. aeruginosa, P. mirabilis) were selected based on their prevalence in urinary tract infections (UTIs). 1 mL overnight bacterial culture was spun down to keep the pellet, and then was suspended in 1 mL pre-warmed sterile urine. After thoroughly vortexing, the suspension was transferred into a pre-warmed (37°C) tube containing 9 mL sterile urine to mimic mammalian infection condition. Spiked urinary samples were directly processed using the SA method without further culture. Figs. 8A- E show mass spectra of five selected gram-negative spectra with samples from a 10-minute incubation. Lipid A signature ions were observed from each tested species, for example, A. baumanii (Fig. 8A) with unique ions of hepta-acylated lipid A show at m/z 1910. Series of ions at m/z 1446, 1462, and 1616 are assigned to P. aeruginosa in Fig. 8D. A sterile urinary control sample spectrum (Fig. 8F) demonstrates that there are minimal background contaminants within the m/z range of these lipid A molecules of interests.

[00143] In order to highlight the adaptability of the SA method to the existing lipid library in Biotyper, mass spectra from these Gram-negative species including two additional time points, 20-minute and 30-minute, were searched against the urine lipid library. This reference library was built containing 25 clinically relevant microbial entries that were stored as 25 Main Spectra (MSP) that were generated based on multiple replicates of a single strain. The same criteria were used to interpret the Biotyper confidence log score. The higher the Biotyper log score, the higher the degree of similarity to a given strain in the database. The best match organism with a confidence log score in 2.00 to 3.00 is considered as high confidence identification. A log score in the range of 1.70 to 1.99 is determined as probable identification. Any score < 1.69 is considered as unreliable identification. Mass spectra from A. baumanii, E. coli, P. aeruginosa and P. mirra were correctly identified as A. baumanii, E. coli, P. aeruginosa and P. mirra with log scores > 2.00 (high confidence identification) respectively (Fig. 8G). Although 10-minute spectra of K. pneumoniae were misidentified as P. mirabilis (top match) with average log scores of 2.33, scores for K. pneumoniae were 2.00 in the third place of identification list. 20-minute K. pneumoniae mass spectra were identified correctly as K. pneumoniae from one replicate spectrum with score of 2.07 while other two replicate spectra were positively identified as P. mirabilis. Table 2 provides the full list of Biotyper log score of these bacterial species processed by different incubation time.

Table 2 - Biotyper log score of urine spiked samples. Values listed in the table are log score of each species from their classification ranking list.

Top Top Top

10 minute 20 minute 30 minute

Match Match Match

A. baumanii 2.31 ± 0.01 Yes 2.32 ± 0.00 Yes 2.30 ± 0.00 Yes

E. coli 2.66 ± 0.00 Yes 2.74 ± 0.00 Yes 2.75 ± 0.00 Yes

No Partial

K. pneumoniae 2.00 ± 0.00 1.86 ± 0.19 2.07 ± 0.00 Yes

(P. mirabilis) (P. mirabilis)

P. aeruginosa 2.57 ± 0.05 Yes 2.44 ± 0.00 Yes 2.57 ± 0.05 Yes

P. mirabilis 2.16 ± 0.00 Yes 2.22 ± 0.12 Yes 2.14 ± 0.00 Yes Polymicrobial Infection Analysis from UTI Model

[00144] Some UTIs in nature are polymicrobial infections. However, little is known about the most effective treatment for polymicrobial infections. Many studies [42] reveal the clinically relevant bacterial -fungal interactions, for example in Candida-E. coli interactions where E.coli facilitates the establishment of C. albicans infection in urinary tract [43]. When multiple microbial species present in the infection simultaneously, a small ecosystem is formed and stabilized by these species, which increases the resistance to antibiotics [44]. It was therefore asked if sample produced using the SA method can extract species-specific lipids in the presence of multiple organisms. Clinically frequently identified causative agents were selected for two types of UTI: uncomplicated UTI and complicated UTI [30]. UPEC, K. pneumoniae and S.

aureus were selected for uncomplicated UTI model while UPEC, K. pneumoniae and C. albicans were the uropathogens used to test the applicability on a complicated UTI model.

[00145] Each pathogen was cultured individually and spiked together into the sterile urine. The urine samples were co-extracted following the protocol for lipid A microextraction from cells using the SA method as defined above and a 30-minute incubation. Fig. 9A shows the mass spectrum of the mixed urinary sample containing three species: S. aureus, E. coli and K.

pneumoniae. Signature ions labeled in the spectrum were determined by comparing with each individual mass spectrum as well as lipid library reference spectra. For S. aureus, m/z at 1352 and 1366 shown in Fig. 9A were also detected in Leung et al. [17], reporting on mixed sample mass spectrum. Notably, a new ion was found at m/z 1046 in Fig. 9A that was not present in AI method sample produced spectrum [17]. This ion was assigned to S. aureus based on the evidence that it was consistently found in S. aureus individual mass spectra (Fig. 10A). Although ions assigned to E. coli at low mass range of m/z 1300 to 1400 may interfere the identification of S. aureus, the new ion at m/z 1046 combined with confirmed ions at m/z 1352 and 1366 can be used to differentiate from E. coli.

[00146] Signature ions of E. coli and K. pneumoniae were successfully detected in both uncomplicated and complicated UTI models (Fig. 9A and 9B). In complicated UTI mass spectrum, C. albicans unique ions were detected at m/z 1087, 1103 and 1115 as well as at higher mass range of m/z 1424 and 1449 (Fig. 9B). The monomicrobial mass spectrum of C. albicans is shown in (Fig. 10B). Uncomplicated and complicated mixed sample mass spectra from 10- minute incubations exhibited similar results (data not shown).

[00147] The results presented here demonstrate that the SA method can be readily incorporated into the lipid library platform to rapidly identify microorganisms within an hour from complex biological matrix samples to final results. The SA method has been systemic optimized by determining the optimal condition and evaluating its performance in terms of rapidness, sensitivity, and handling of biological matrix and polymicrobial detection. Samples prepared using the SA method show the capability to produce mass spectra that are

distinguishable from each corresponding species with only 10-minute incubation. This indicates that by incorporating the SA method into a lipid database platform, the diagnostic time can be reduced to < 30 minutes from the obtaining the sample to the result.

[00148] Although the SA method was originally designed for lipid A extraction from Gram- negative bacteria, as shown herein the same protocol can be extended to Gram-positive and fungal species with successful extraction of their lipid markers from cell membranes. Thus, the diverse range of extractable microorganisms of the SA method allows it be easily adapted to a lipid approach.

* * * *

[00149] While the invention has been described with reference to certain particular embodiments thereof, those skilled in the art will appreciate that various modifications may be made without departing from the spirit and scope of the invention. The scope of the appended claims is not to be limited to the specific embodiments described.

REFERENCES

[00150] All patents and publications mentioned in this specification are indicative of the level of skill of those skilled in the art to which the invention pertains. Each cited patent and publication is incorporated herein by reference in its entirety. All of the following references have been cited in this application:

1. Fauci, A. S. & Morens, D. M. The Perpetual Challenge of Infectious Diseases. N. Engl. J.

Med. 366, 454-461 (2012).

2. Clark, A. E., Kaleta, E. J., Arora, A. & Wolk, M. Matrix- Assisted Laser Desorption

Ionization - Time of Flight Mass Spectrometry: a Fundamental Shift in the Routine Practice of Clinical. 26, 547-603 (2013).

3. Belkum, A. Van et al. Rapid Clinical Bacteriology and Its Future Impact. 14-27 (2013).

4. Bertelli, C. & Greub, G. Rapid bacterial genome sequencing: methods and applications in clinical microbiology. Clin. Microbiol. Infect. 19, 803-13 (2013).

5. Espy, M. J. et al. Real-Time PCR in Clinical Microbiology: Applications for Routine Laboratory Testing. Clin. Microbiol. Rev. 19, 165-256 (2006).

6. Glish, G. L. & Vachet, R. W. The basics of mass spectrometry in the twenty-first century.

Nat. Rev. Drug Discov. 2, 140-150 (2003).

7. Sauer, S. & Kliem, M. Mass spectrometry tools for the classification and identification of bacteria. Nat. Rev. Microbiol. 8, 74-82 (2010).

8. Savitzky, A. & Golay, M. J. E. Smoothing and Differentiation of Data by Simplified

Least Squares Procedures. Anal. Chem. 36, 1627-1639 (1964).

9. Seng, P. et al. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clin. Infect. Dis. 49, 543-51 (2009).

10. Cheng, K., Chui, H., Domish, L., Hernandez, D. & Wang, G. Recent development of mass spectrometry and proteomics applications in identification and typing of bacteria. Proteomics - Clin. Appl. 10, 346-357 (2016).

11. Ho, Y.-P. & Reddy, P. M. ADVANCES IN MASS SPECTROMETRY FOR THE

IDENTIFICATION OF PATHOGENS. Mass Spectrom. Rev. 30, 1203-1224 (2011).

12. Biswas, S. & Rolain, J. M. Use of MALDI-TOF mass spectrometry for identification of bacteria that are difficult to culture. J. Microbiol. Methods 92, 14-24 (2013). Croxatto, A., Prod'hom, G. & Greub, G. Applications of MALDI-TOF mass

spectrometry in clinical diagnostic microbiology. FEMS Microbiol. Rev. 36, 380-407 (2012).

Levesque, S. et al. A Side by Side Comparison of Bruker Biotyper and VITEK MS: Utility of MALDI-TOF MS Technology for Microorganism Identification in a Public Health Reference Laboratory. PLoS One 10, 1-21 (2015).

Maier, T., Klepel, S., Renner, U. & Kostrzewa, M. Fast and reliable MALDI-TOF MS- based microorganism identification. Nat. Methods 3, 1-2 (2006).

Opota, O., Prod'hom, G. & Greub, G. in MALDI-TOF and Tandem MS for Clinical Microbiology, First Edition 3, 55-91 (2017).

Leung, L. M. et al. Identification of the ESKAPE pathogens by mass spectrometric analysis of microbial membrane glycolipids. Sci. Rep. 7, 6403-6413 (2017).

Percy, M. G. & Griindling, A. Lipoteichoic Acid Synthesis and Function in Gram- Positive Bacteria. Annu. Rev. Microbiol. 68, 81-100 (2014).

Brown, L., Wolf, J. M., Prados-Rosales, R. & Casadevall, A. Through the wall:

extracellular vesicles in Gram-positive bacteria, mycobacteria and fungi. Nat. Rev.

Microbiol. 13, 620-630 (2015).

Miller, S. I, Ernst, R. K. & Bader, M. W. LPS, TLR4 and infectious disease diversity. Nat. Rev. Microbiol. (2005). doi: 10.1038/nrmicrol068

El Hamidi, A. Microextraction of bacterial lipid A: easy and rapid method for mass spectrometric characterization. J. Lipid Res. 46, 1773-1778 (2005).

Li, Y., Wang, X. & Ernst, R. K. A rapid one-step method for the characterization of membrane lipid remodeling in Francisella using matrix-assisted laser desorption ionization time-of-flight tandem mass spectrometry. Rapid Commun. Mass Spectrom. 25, 2641-2648 (2011).

Zhou, P. et al. Microwave-assisted sample preparation for rapid and sensitive analysis of H. pylori lipid A applicable to a single colony. J. Lipid Res. 50, 1936-1944 (2009).

Larrouy-Maumus, G., Clements, A., Filloux, A., McCarthy, R. R. & Mostowy, S. Direct detection of lipid A on intact Gram-negative bacteria by MALDI-TOF mass

spectrometry. J. Microbiol. Methods 120, 68-71 (2016). W.J.Dyer, E. G. B. and. Canadian Journal of. Can. J. Biochem. Physiol. 37, 911-917 (1959).

Kawasaki, K. Alternative procedures for analysis of lipid A modification with phosphoethanolamine or aminoarabinose. J. Microbiol. Methods 76, 313-315 (2009). Chafchaouni-Moussaoui, I. et al. A new rapid and micro-scale hydrolysis, using triethylamine citrate, for lipopolysaccharide characterization by mass spectrometry. Rapid Commun. Mass Spectrom. 25, 2043-2048 (2011).

Santajit, S. & Indrawattana, N. Mechanisms of Antimicrobial Resistance in ESKAPE Pathogens. Biomed Res. Int. 2016, (2016).

Liu, Y. Y. et al. Emergence of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in China: A microbiological and molecular biological study. Lancet Infect. Dis. 16, 161-168 (2016).

Flores-Mireles, A. L., Walker, J. N., Caparon, M. & Hultgren, S. J. Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat. Rev. Microbiol. 13, 269-84 (2015).

Barber, A. E., Norton, J. P., Spivak, A. M. & Mulvey, M. A. Urinary tract infections: Current and emerging management strategies. Clin. Infect. Dis. 57, 719-724 (2013). Dielubanza, E. J. & Schaeffer, A. J. Urinary Tract Infections in Women. Med. Clin. North Am. 95, 27-41 (2011).

Liu, Y. Y. et al. Structural modification of lipopolysaccharide conferred by mcr-1 in gram-negative ESKAPE pathogens. Antimicrob. Agents Chemother. 61, 580-17 (2017). Gibb, S. & Strimmer, K. MALDIquant: a versatile R package for the analysis of mass spectrometry data. Bioinformatics 28, 2270-2271 (2012).

SNIP, a statistics-sensitive background treatment for the quantitative analysis of PIXE spectra in geoscience applications. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact, with Mater. Atoms 34, 396-402 (1988).

Caroff, M., Tacken, A. & Szabo, L. Detergent-accelerated hydrolysis of bacterial endotoxins and determination of the anomeric configuration of the glycosyl phosphate present in the 'Isolated lipid A' fragment of the Bordetella pertussis endotoxin.

Carbohydr. Res. 175, 273-282 (1988). Eng, J. K., Mccormack, A. L. & Yates, J. R. An Approach to Correlate Tandem Mass Spectral Data of Peptides with Amino Acid Sequences in a Protein Database. Am. Soc. Mass Spectrom. 5, 976-989 (1994).

Liu, J. et al. Methods for peptide identification by spectral comparison. Proteome Sci. 5, 3 (2007).

Evans, J. D. Straightforward statistics for the behavioral sciences. (Brooks/Cole Pub. Co, 1996).

Matsuura, M. Structural modifications of bacterial lipopolysaccharide that facilitate gram-negative bacteria evasion of host innate immunity. Front. Immunol. 4, 1-9 (2013). Gunn, J. S. & Ernst, R. K. The structure and function of Francisella lipopolysaccharide. Ann. N. Y. Acad. Sci. 1105, 202-218 (2007).

Dhamgaye, S., Qu, Y. & Peleg, A. Y. Polymicrobial infections involving clinically relevant Gram-negative bacteria and fungi. Cell. Microbiol. 18, 1716-1722 (2016). Levison, M. E. & Pitsakis, P. G. Susceptibility to Experimental Candida albicans Urinary Tract Infection in the Rat. J. Infect. Dis. 155, 841-846 (1987).

de Vos, M. G. J., Zagorski, M., McNally, A. & Bollenbach, T. Interaction networks, ecological stability, and collective antibiotic tolerance in polymicrobial infections. Proc. Natl. Acad. Sci. 114, 10666-10671 (2017).