Fluorescent polybranched probes for detecting bacteria and/or fungi in vitro and in vivo

Abstract

A probe for detecting bacteria and/or fungi in vitro and in vivo is provided, the probe having a core and a plurality of probe elements; each probe element within the plurality of probe elements extending from the core and having a fluorophore and a binding moiety, wherein the binding moiety is a bacteria binding moiety and selectively binds to bacteria and/or to fungi and not to animal cells. Methods of use of the probe and kits comprising the probe are also provided.

Claims

1. A probe comprising a core and a plurality of probe elements, each probe element within the plurality of probe elements extending from the core and comprising a fluorophore and a binding moiety, wherein the binding moiety is a bacteria binding moiety and selectively binds to at least some bacteria, and does not bind to animal cells, or wherein the binding moiety is a fungi binding moiety, or wherein the binding moiety selectively binds to at least some bacteria and to at least some fungi, and not bind to animal cells, wherein the probe has one of the following structures: ##STR00019## ##STR00020## where L=spacer group F=fluorophore B=binding moiety C=core.

2. The probe according to claim 1, wherein the probe comprises three probe elements connected to a core each probe element comprising NBD-(CH.sub.2).sub.5CO-UBI.sub.Nle.

3. The probe according to claim 1, wherein the probe comprises three probe elements connected to a core each of the three probe elements comprising NBD-(CH.sub.2).sub.2O(CH.sub.2).sub.2OCH.sub.2-PMX.

4. The probe according to claim 1, wherein the probe has the structure: ##STR00021## wherein the binding moiety is SEQ ID NO 3.

5. A method of detecting bacteria and/or fungi in a target area, the method comprising the steps: (1) providing a first probe according to claim 1; (2) delivering the first probe to the target area; (3) illuminating the target area with an appropriate wavelength of light to excite the fluorophores of the first probe; and (4) determining whether the first probe has labelled bacteria and/or fungi within the target area.

6. The method according to claim 5, wherein the method comprises the further steps of: providing a second probe delivering the second probe to the target area; and illuminating the target area with an appropriate wavelength of light to excite the or each fluorophore of the second probe; wherein the second probe comprises a core and a plurality of probe elements, each probe element within the plurality of probe elements extending from the core and comprising a fluorophore and a binding moiety, wherein the binding moiety is a bacteria binding moiety that specifically binds to a subpopulation of bacteria, and does not bind to animal cells, and wherein the probe has one of the following structures: ##STR00022## ##STR00023## where L=spacer group F=fluorophore B=binding moiety C=core.

7. The method according to claim 6, wherein the first probe comprises a ubiquicidin moiety, including full length ubiquicidin, or a fragment thereof as the binding moiety.

8. The method according to claim 6, wherein the second probe comprises a polymyxin moiety, including full length polymyxin or a fragment thereof, as the binding moiety, and the second probe may selectively bind to gram negative bacteria.

9. The method according to claim 8, wherein the first probe comprises a ubiquicidin moiety and the method allows the identification of sterile inflammation, bacterial and/or fungal inflammation, and gram-negative bacterial infection and by inference, gram-positive bacterial infection.

10. A kit of parts comprising a first probe according to claim 1 comprising at least one binding moiety that specifically binds to bacteria and/or fungi, and a second probe according to claim 1 comprising at least one binding moiety that specifically binds to gram-negative or gram-positive bacteria and a suitable buffer within which the probe may be dispersed.

11. The kit of parts according to claim 10, wherein the first probe comprises at least one ubiquicidin moiety, and the second probe comprises at least one polymyxin moiety.

12. The probe according to claim 1, wherein L is C.sub.3-C.sub.10 alkyl, or ((CH.sub.2).sub.2O).sub.x where x=1-6.

13. The probe of claim 1, wherein F is NBD or fluorescein.

14. The probe of claim 1, wherein B is a ubiquicidin moiety or a polymyxin moiety.

15. The probe of claim 1, wherein C is carbon, C(CH.sub.2O(CH.sub.2).sub.3).sub.3, or NH.sub.2CONHC(CH.sub.2O(CH.sub.2).sub.3).sub.3.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1: Fluorophore choice impacts on ability to image labelled bacteria with improved signal to noise for NBD constructs over Fluorescein (FAM). A) Bacteria (Methicillin sensitive Staphylococcus aureus, MSSA), counterstained with FM-464 (shown hi left-hand column), and incubated with FAM-UBI or NBD-UBI demonstrating an improved signal to noise ratio with NBD which avowed the majority of bacteria to be detected (shown in middle column). This was compared with FAM-UBI which detected a minor subset of clumped bacteria. B) NBD-UBI also exhibits specificity for bacteria (Pseudomonas aeruginosa) over mammalian cells (white arrow).

(2) FIG. 2: FAM-PMX does not allow sufficient signal to noise to detect bacteria in vitro. A) Pseudomonas aeruginosa (PA), counterstained with FM-464 (left-hand images) and incubated with A549 cells (human lung epithelial cells highlighted in the middle image) demonstrate bacterial labeling is not seen above background but also there is no labeling of the epithelial calls. B) NBD-PMX (left-hand image) with a nuclear counterstain Syto 82 (middle image) allows bacterial labelling and retains selectivity of labeling over mammalian cells (right-hand image).

(3) FIG. 3: Emission spectra of probes confirms increase in fluorescence in hydrophobic environments. A) Emission spectra shown for NBD-UBI.sub.dend (5 M) and NBD-UBINle (15 M) in the presence of PBS or DMSO (hydrophobic environment) when excited at 488 nm wavelength. B) Emission spectra shown for NBD-PMX (10 M) in the presence of PBS or DMSO (hydrophobic environment) when excited at 488 nm wavelength. The fluorescence of both probes greatly increases under hydrophobic conditions as would be present upon binding to the target bacteria.

(4) FIG. 4: NBD-UBI.sub.dend labels bacteria selectively over mammalian cells in vitro: A) Example images of a clinically relevant bacterial panel with NBD-UBI.sub.dend (5 M; left-hand images) counterstained with the fluorescent generic cellular DNA dye Syto-82 imaged by laser scanning confocal microscopy. B) Quantification of bacterial panel labelling NBD-UBI.sub.dend 5 M (optimal concentration) where every bacteria in the panel is brighter when compared with MSSA and NBD-UBI UBI.sub.Nle 105 M (optimal concentration) C) Flow cytometry data with unstained bacteria (shaded), NBD-UBI.sub.Nle 155 M (white) and NBD-UBI.sub.dend 55 M labelled bacteria (white with grey outline orange). D) MSSA and NBD-UBI.sub.dend (5 M; left-hand image), counterstained with Syto-82 and merge showing human neutrophils (white arrow) and E) A549 cells (white arrow) demonstrating no labelling of mammalian cells. Representative images shown, n3 for all experiments.

(5) FIG. 5: NBD-PMX selectively labels Gram-negative bacteria in vitro, Example images of the bacterial panel with NBD-PMX imaged by laser scanning confocal microscopy. A) Bacterial panel with NBD-PMX 1 M (left-hand column) and counterstain with syto-82. Gram-positive bacteria (bounded by black box) display minimal/no labelling compared with Gram-negative bacteria. B) Quantification of bacterial panel with NBD-PMX 1 M with Gram-positive bacteria in white bars and Gram-negative in black bars, showing high intensity selective labelling of Gram-negative bacteria compared with Gram-positive bacteria. All Gram-negative bacteria showed a statistically significant increase over all Gram-positives. C) Flow cytometric evaluation of bacterial labelling with NBD-PMX showing unstained bacteria (shaded) and NBD-PMX (white) labelled bacteria. No significant labelling of Gram-positive bacteria (bounded by the box), D) MSSA and NBD-PMX 1 M (left-hand images), counterstained with Syto-82 and merge showing human neutrophils (white arrow) and E) A549 cells (white arrow) demonstrating no labelling of mammalian cells (reproduced from FIG. 2), Representative images shown, n3 for all experiments.

(6) FIG. 6: Ex vivo model of bacterial infection. A) Ovine lungs were harvested, ventilated and placed in a neonatal incubator with ambient temperature of 37 C. Pulmonary segments were instilled with PBS (control) or bacteria via bronchoscopy. Probes were then instilled and segments imaged with fibered confocal fluorescence microscopy (FCFM) using 488 nm Laser Scanning Unit (LSU), B) Ex vivo ovine model demonstrates viable bacteria 5 hours following instillation. Bacteria (MSSA) instilled into segments of ovine lung and lavaged at 1, 3 and 5 hours (n=3), plated for CFU/ml and counted following 16 hour incubation.

(7) FIG. 7: NBD-UBINle fails to label bacteria in the ovine lunge There was no consistent labelling of bacteria in the ovine lung confirmed by counterstaining bacterial and imaging on a spectrally distinct imaging system. A: Lung segment with PKH660 labelled MSSA at 1 hour post-instillation, B: Control segment (instilled with PBS and no bacteria) imaged at 660 nm. C) The same segment imaged at 488 nm following 10 M NBD-UBI.sub.Nle instillation. However, the same experiment with counterstained PKH660 labelled K. pneumoniae (Gram negative) shown in panel ID, with NBD-PMX added demonstrating the same punctate signal,

(8) FIG. 8: Representative FCFM images of bacteria in the distal lung demonstrating a distinctive punctate pattern. A) Instillation of PKH labelled S. aureus, Calcein labelled S. aureus or GFP-expressing strain of S. aureus, generate a punctate pattern of fluorescence when segments are imaged using FCFM. Images used to generate positive control for subsequent in situ labelling experiments. Representative images shown, n=3 for all experiments. B) PKH dyes do not stain by-stander cells. Confocal images showing epithelial cells in co-culture with bacteria (S. aureus) that have been counterstained with Syto-82 (bottom left image red) as well as PKH660 (too left image). Bystander mammalian cells (arrows) are labelled with Syto-82 that has leached from the labelled bacteria but there is no transfer of PKH660.

(9) FIG. 9. NBD-UBI.sub.dend detects bacteria in situ in the distal ovine lung Representative FCFM images of NBD-UBI.sub.dend showing that no punctate signal is seen in control segments whereas the distinctive punctate signal described in FIG. 8 above is seen above background fluorescence in segments instilled with bacteria. Furthermore, agarose beads alone show no fluorescence (control beads) but when beads are coated with bacteria a signal is seen when labelled in situ in the ovine lung, Representative images shown, n3 for all experiments.

(10) FIG. 10; NB D-PMX selectively labels Gram-negative bacteria in situ in the ovine lung. A) Control segments and Gram-positive segments show no punctate signal, whereas Gram-negative instilled segments show signal identical to positive control. B) Lavage counts from segments demonstrating no significant difference in counts between segments confirming the presence of bacteria. C) Agarose beads alone show no fluorescence but when beads are coated with bacteria a signal is seen when labelled in situ in the ovine lung. Representative images shown, n3 for all experiments.

(11) FIG. 11: NBD-UBI.sub.dend and NBD-PMX are resistant to wash off allowing detection of bacteria in the ovine lung. Bacteria pre-labelled with either NBD-UBI.sub.dend or NBD-PMX were readily visualised by FCFM when instilled into the ovine lung. However, when bacteria pre-labelled with linear NBD-UBI.sub.Nle are instilled, no punctate signal is seen, Labelling was confirmed by imaging bacterial suspensions (left hand panels) before instillation into the lung and imaging by FCFM (right hand panels). Note the lower signal-to-noise ratio for NBD-UBI.sub.Nle labelled bacteria in suspension. Representative images shown, n=3 for each experiment.

(12) FIG. 12: NBD-UBI.sub.dend retains fluorescence when wash performed. A) The bacterial panel was imaged by laser scanning confocal microscopy in the continued presence of NBD-UBI.sub.dend (black) and following PBS wash (white). Quantification demonstrates fluorescence retention above linear NBD-UBI with MSSA in the continued presence of probe (dotted line). n3, with three random field-of-views assessed for each experiment. B) Bacterial panel imaged with NBD-PMX in the continued presence of probe (black) or following PBS wash (white) demonstrating higher fluorescence retention of all Gram-negative bacteria above NBD-PMX with Gram-positives in the continued presence of probe. n3, with three random field-of-views assessed for each experiment.

(13) FIG. 13: NBD-UBI.sub.dend is stable in Acute Lung Injury Broncholalveolar Lavage Fluid (ALI HALF). Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (MALDI-TOF MS) analysis demonstrated stability of NBD-UBI.sub.Nle, wherein UBI.sub.Nle is SEQ ID NO 3, in the presence of saline (arrow indicates correct peak seen at mass of 1949) but degradation in the presence of ALI lavage (no peak seen at mass 1949 but arrows show predictable degradation compounds, SEQ ID NO 7 and SEQ ID NO 8) when co-incubated for 30 minutes. By contrast NBD-UBI.sub.dend remains stable when assessed by Fourier Transform Mass Spectrometry (FTMS) (data shown represents a theoretical plot and an experimental plot demonstrating the peaks correspond indicating presence of compound). NBD-PMX is stable in ALI BALE MALDI-TOF analysis demonstrated stability of NBD-PMX in the presence of saline or ALI lavage (arrows indicates correct peaks seen at mass of 1291-1293) when co-incubated for 30 minutes.

(14) FIG. 14 NBD-UBI.sub.dend retains higher fluorescence on bacteria in ovine lavage than equimolar equivalent of linear. A) MSSA imaged by laser scanning confocal microscopy in the presence or absence of ovine lavage with NBD-UBI.sub.Nle (15 M) and NBD-UBI.sub.dend (5 M). Both NBD-UBI compounds have reduced fluorescence in ovine lavage but NBD-UBI.sub.dend retains higher fluorescence intensity. B) In contrast, PA3284 imaged in the presence or absence of ovine lavage with NBD-PMX (1 M) demonstrates no significant reduction of fluorescence intensity. Representative images shown, n=3.

(15) FIG. 15: NBD-UBI.sub.dend and NBD-PMX successfully image bacteria in a surfactant-rich environment. Ratio of bacterial fluorescence intensity:surfactant fluorescent intensity quantified from confocal microscopy images. For equimolar NBD concentrations NBD-UBI.sub.dend and NBD-PMX have significantly higher bacterial: surfactant intensity than NBD-UBI.sub.Nle.

(16) FIG. 16A: Elastase dependent bacterial labelling scheme. The fluorophore on the bacterial binding component is quenched by a dark quencher meaning that no fluorescence is seen upon binding to bacteria. Cleavage of the elastase specific sequence AAPV (SEQ ID NO: 9) liberates the bacterial binding component (PV-NBD-UBI) from the dark quencher methyl red (MR).

(17) FIG. 16B: Elastase dependant bacterial labelling in vitro. MSSA, counterstained with FM-464 (left-hand image red), co-incubated with isolated human neutrophils demonstrating bacterial labelling in the presence of neutrophils (middle image), which is inhibited in the presence of an elastase inhibitor.

(18) FIG. 17A: Elastase dependent bacterial labelling scheme. The fluorophore on the bacterial binding component is quenched by a different fluorophore of a longer wavelength meaning that no fluorescence is seen upon binding to bacteria. Cleavage of the elastase specific sequence AAPV (SEQ ID NO: 9) liberates the bacterial binding component (PV-NBD-UBI) from the quencher (TAMRA). TAMRA reports the presence of activated neutrophils by endocytic uptake.

(19) FIG. 17B Elastase dependent bacterial labelling in vitro. Pseudomonas aeruiginosa (PA) counterstained with PKH 660 (third images from left) was incubated with TAMRA-AAPV-NBD-PMX (SEQ ID NO: 9) and demonstrated no labelling in the presence of unactivated neutrophils (top panel), bacterial labelling in an elastase dependent manner and labelling of activated neutrophils in an elastase independent manner.

(20) FIG. 17C: Elastase dependent fluorescence increase of TAMRA following cleavage. Spectrophotometer data to demonstrate the unexpected observation of an increase in fluorescence of TAMRA-AA (cleaved compound) compared to TAMRA-AAPV-NBD-PMX (SEQ ID NO: 9) in an elastase dependent manner, and inhibited with an elastase inhibitor (ex 525 nm, Em 580 nm),

(21) FIG. 18: Image Processing Algorithms delineate positive and negative frames from FCFM data (A) raw data for a frame that is deemed positive or negative. (B) Without pre-processing, spot-detection is ineffective. Following estimation of background noise (C), high-pass filtering (D) and appropriate thresholding (E), the current algorithm detect minimal punctate signal in negative frames (F).

(22) FIG. 19: Automated Image Processing of entire FCFM videos (up to 3500 frames) demonstrates detection of bacteria. A) Analysis of NBD-UBI.sub.dend videos with thresholding at 80 spots per frame and an arbitrary cut-off of 10% enables objective detection of bacteria using a combination of Laplacian of Gaussian (LoG) and Difference of Gaussians (DoG) positive frames. B) Analysis of NBD-PMX videos with thresholding at 80 spots per frame and an arbitrary cut-off of 10% enables objective detection of) Gram status using a combination of LoG and Do G positive frames. n=4 for all analyses, except S. pneumoniae where n=3, Statistical analysis when compared to control.

(23) FIG. 20: Bacteria detected in the presence of human lung tissue autofluorescence. A) FCFM images of human lung tissue with PA3284 demonstrating elastin autofluorescence and no bacterial signal (left) and human lung tissue with PA3284 co-incubated with NBD-PMX (right) demonstrating bacterial signal. B) Live confocal microscopy images (merge) showing merge images of nucleic acid counterstain with no NBD-PMX labelling, Bacteria in right panel can be clearly seen, labelled with NBD PMX and above the background fluorescence. C) The fluorescent lifetime of NBD is significantly longer than that of the autofluorescence of the indigenous cells of the lung, measured by laser excitation at 488 nm and time resolved single photon counting spectrometer, and this can be used to identify the probes in vivo and increase signal-to-noise of sensing and imaging.

(24) FIG. 21: NBD-UBI.sub.dend labels bacteria and variably labels A. fumigatus but not C. albicans imaged by confocal microscopy. Representative images of NBD-UBI.sub.dend 5 M demonstrating A: labeling of bacteria (P aeruginosa) and fungal hyphae of A. fumigatus (white arrow) but not C. albicans (white arrow). Panel B demonstrates the variability of labeling on fungal hyphae.

(25) FIG. 22: NBD-PMX labels bacteria and variably labels A. fumigatus but not C. albicans imaged by confocal microscopy. Representative images of NBD-PMX 5 M demonstrating A: labeling of bacteria (P aeruginosa) and fungal hyphae of A. fumigatus (white arrow) but not C. albicans (white arrow). Panel B demonstrates the variability of labeling on fungal hyphae. Data shown for 5 uM but similar results seen for 1 uM.

(26) FIG. 23. Suspensions of C. albicans do not demonstrate increased fluorescence when imaged with FCFM. Figure shows positive control of C. albicans labeled with Syto green and imaged in suspension but no labeling without the probe, with NBD-UBIdend or NBD-PMX with FCFM.

(27) FIG. 24: A. fumigatus demonstrates fluorescence in a distinctive and characteristic manner when imaged with FCFM. Top panel is without the probe demonstrating no intrinsic autofluorescence and bottom panels show distinct pattern with NBD-UBI.sub.dend and NBD-PMX.

(28) FIG. 25: A. fumigatus demonstrates linear pattern of fluorescence in the presence of ovine lung tissue, which is distinct from a bacterial signal. Top panel shows the pattern of fluorescence with a bacterial solution with the probe and bottom panels show A. fumigatus filaments with NBD-UBI.sub.dend and NBD-PMX on ovine lung.

(29) FIG. 26: No haemolysis seen with NBD-UBI.sub.dend or NED-PMX up to 100 M. Haemolysis assay (n=3) demonstrating no haemolysis on concentrations up to 100 M, Positive control was 0.2% Triton-X and values corrected to represent 100% haemolysis for Triton-X.

(30) FIG. 27. A: NBD-PMX shows no organ toxicity in a 2 week single instillation model in mice. Representative histology images (100) for NBD-PMX compared to PBS control animals at 48 hours and 14 days. No differences from control were seen in any group (blindly scored by a consultant histopathologist). B: NBD-UBI.sub.dend shows no organ toxicity in a 2 week single instillation model in mice. Representative histology images (100) for NBD-UBI.sub.dend compared to PBS control animals at 48 hours and 14 days. No differences from control were seen in any group (blindly scored by a consultant histopathologist).

(31) FIG. 28. Fluorescence Life Time imaging easily distinguishes lung tissue from bacteria. Left panel shows confocal at 488 nm excitation of lung and bacteria (dots) in lung tissue. FLIM (right panel) shows dear differences in imaging of lung tissue and bacteria. Imaging performed using NBD-UBI.sub.dend and S. aureus on human lung tissue

SPECIFIC DESCRIPTION OF EMBODIMENTS OF THE INVENTION

(32) While the making and using of various embodiments of the present invention are discussed in detail below, it should be appreciated that the present invention provides many applicable inventive concepts that can be embodied in a wide variety of specific contexts. The specific embodiments discussed herein are merely illustrative of specific ways to make and use the invention and do not delimit the scope of the invention.

(33) To facilitate the understanding of this invention, a number of terms are defined below. Terms defined herein have meanings as commonly understood by a person of ordinary skill in the areas relevant to the present invention. Terms such as a, an and the are not intended to refer to only a singular entity, but include the general class of which a specific example may be used for illustration. The terminology herein is used to describe specific embodiments of the invention, but their usage does not delimit the invention, except as outlined in the claims.

(34) In the following description of example embodiments of the invention, binding moieties comprising a polymyxin moiety are given the code PMX, and binding moieties comprising a ubiquicidin moiety are given the code UBI. For example, embodiments of the invention that comprise a plurality of probe elements comprising the NBD and the modified ubiquicidin fragment UBI.sub.Nle (Dendron Probes) is referred to as NBD-UBI.sub.dend.

Choice of Probe Element Fluorophore

(35) For the reporting of bacteria we synthesised a probe comprising the probe element only, and substituting the methionine of the ubiquicidin fragment UBI.sub.29-41 for a norleucine, NBD-UBI.sub.Nle. This probe was compared to the same bacterial detecting moiety with another always on fluorophore, fluorescein (FAM), FAM-UBI.sub.Nle, and showed an improved signal-to-noise on live benchtop confocal microscopy for the NBD reported (FIG. 1). Furthermore, it was confirmed that the labelling is specific to bacteria and not cell membranes in general. For example, FIG. 1 shows that isolated human neutrophils were not labelled by the NBD-UBI.sub.Nle probe, whilst the bacteria were labelled by the NBD-UBI.sub.Nle probe.

(36) To confirm the same would be observed for the PMX bacterial detecting moiety we constructed NBD-PMX and FAM-PMX demonstrating an improved signal-to-noise with NBD-PMX over FAM-PMX and confirm this construct is also specific to mammalian cells (FIG. 2).

Branched/Dendron or Multivalent Probes

(37) A probe comprising a core and three probe elements connected to the core (a three branch probe) was prepared (NBD-UBI.sub.dend), each probe element comprising a NBD-UBI.sub.Nle moiety.

(38) To confirm the fluorescent reporter NBD retains its characteristics when coupled with our peptide moieties we measured the fluorescence of the compounds in conditions to mimic a hydrophobic environment (DMSO). Linear NBD-UBI.sub.Nle, NBD-UBI.sub.dend and NBD-PMX were excited at 488 nm (Biotek fluorescent plate reader) and demonstrated significant increase in fluorescence when the probes were in the presence of dimethyl sulfoxide (DMSO) (hydrophobic environment) when compared to phosphate buffered saline (PBS) (FIG. 3) confirming environmentally sensitive fluorescent reporting. NBD-UBI.sub.dend demonstrated the same fluorescence increase as linear when using eqimolar concentrations of NBD.

(39) A panel of bacteria which represent >70% of VAP causing pathogens (Chastre J et al. Am J Respir Crit Care Med. 2002 Apr. 1; 165(7):867-903) (Gram-negative: P. aeruginosa (two strains), A. baumannii, S. maltophilia, K. pneumoniae, E. coli and H. influenzae. Gram-positive: Methicillin Resistant S. aureus (MRSA), Methicillin Sensitive S. aureus (MSSA) and S. pneumoniae) (strain list in Table 3 below) were labelled with NBD-UBI.sub.dend. Labelling was observed with variable intensity (FIGS. 4A and B). Nevertheless all bacteria were brighter than MSSA labelled with linear NBD-UBI.sub.Nle (FIG. 4B) and on flow cytometry NBD-UBI.sub.dend demonstrated increased labelling over linear NBD-UBI.sub.Nle (FIG. 4C). Furthermore, NBD-UBI.sub.dend did not label human neutrophils or A549 cells supporting prokaryotic selectivity. (FIGS. 4D and E).

(40) NBD-PMX incubated with bacteria, demonstrated significantly higher fluorescence on Gram-negative bacteria (P. aeruginosa, A. baumannii, S. maltophilia, K. pneumoniae, E. coli and H. influenzae) than Gram-positive bacteria (MRSA, MSSA and S. pneumoniae) (p<0.05) on confocal analysis (FIGS. 5A and B), which was confirmed by flow cytometry (FIG. 5C). Furthermore, there was no labelling of human neutrophils or A549 cells. (FIGS. 5D and E).

(41) NBD-UBI.sub.dend and NBD-PMX were assessed for in situ specificity and sensitivity in an ex vivo ovine model of bacterial infection (FIG. 6). In this model, the instillation of PKH fluorescent dye-labelled bacteria into the distal ovine lung and imaged with FCFM reveals a characteristic and distinctive pattern of punctate fluorescence in each field of view (FIG. 8). We observed identical patterns with Calcein-labelled bacteria and Green Fluorescent Protein-expressing S. aureus which were instilled into the lung and imaged with FCFM (FIG. 8). This pattern is reproduced identically by instilling bacteria pre-labelled with NBD-UBI.sub.dend and NBD-PMX into lung segments.

(42) Following this thorough characterisation of the model and positive controls, we instilled PBS or VAP-relevant bacteria into distinct segments within the ex vivo lung model, followed by the microdosed delivery of probes according to the invention. We demonstrated that the linear NBD-UBI.sub.Nle could not label bacteria in situ despite the ability in vitro (FIG. 7). For these experiments the assays were repeated with counterstained gram-negative bacteria (K. pneumoniae) which demonstrated labelling with NBD-PMX (FIG. 7). However, In segments instilled with bacteria we demonstrated the same signal as seen with the positive controls when NBD-UBI.sub.dend is instilled (P. aeruginosa, S. aureus, E. coli and K. pneumonia) but minimal/no signal with PBS control, confirming the ability to label bacteria in situ and image using the FCFM system (FIG. 9).

(43) In segments instilled with Gram-negative bacteria, P. aeruginosa (laboratory strain PA01 and clinical VAP isolate J3284), K. pneumoniae and E. coli, we have demonstrated the same signal as in the positive controls when NBD-PMX is instilled but no signal in segments with PBS or Gram-positive bacteria MSSA, MRSA and S. pneumoniae (FIG. 10). In these experiments we confirmed the equal density of Gram-positive and Gram-negative bacteria in all of the segments imaged with NBD-UBI.sub.dend and NBD-PMX by bronchoalveolar lavage (demonstrating no difference in CFU/ml between segments) and through counterstained bacteria imaged on a LSU at 660 nm.

(44) To further demonstrate in situ bacterial detection and to assess the ability of the probes to image bacterial aggregation, we embedded bacteria in agarose beads which were then instilled into the lung. Microdosed probe instillation and FCFM imaging demonstrated that bacterial beads are clearly and exclusively detected whereas control beads (beads without bacteria) are not (FIGS. 9 and 10).

(45) In the distal lung, there is likely to be significant and rapid dissipation of the probes immediately after delivery. Therefore, it is imperative that probe-bacterial labelling remains persistent under these conditions. NBD-UBI.sub.dend-labelled bacteria retain labelling upon probe wash-off, as is seen for NBD-PMX. When instilled into the ovine lung, bacteria pre-labelled with NBD-UBI.sub.Nle are undetectable by FCFM whereas bacteria pre-labelled with NBD-UBI.sub.dend or NBD-PMX are readily visualised (FIG. 11). As such, this resistance to wash-off represents a surrogate indicator of probe-bacterial affinity which appears to be an absolute requirement for distal lung in situ labelling. Currently, both NBD-UBI.sub.dend and NBD-PMX are resistant to wash-off whilst most of the other structural variants do not. Hence, bacteria labelled with both NBD-UBI.sub.dend and NBD-PMX retain sufficient intensity of fluorescence upon probe dissipation that occurs in the distal lung (FIG. 12).

(46) Secondly we assessed stability of the probes in BALF from patients with ALI by FTMS and MALDI-TOF MS analysis. NBD-UBI.sub.Nle, NBD-UBI.sub.dend and NBD-PMX were incubated for 30 minutes with BALF (FIG. 13). NBD-UBI.sub.dend and NBD-PMX demonstrate stability with peaks corresponding exactly to the predicted theoretical spectra of the intact probes readily identified in both saline and BALF incubated samples. Predicted spectra of breakdown products were also not seen. By contrast breakdown products of the NBD-UBI.sub.Nle were seen indicating instability in lavage fluid. Thus, probe stability and retention of function in the context of the inflamed highly enzymatically-active distal lung environment is a key determinant of in vivo utility. This was confirmed by imagining in the presence of ovine lavage, which demonstrated higher fluorescence intensity of NBD-UBI.sub.dend over the linear compound, and no reduction in intensity for NBD-PMX when imaged in ovine lavage (FIG. 14).

(47) In vitro experiments were conducted using lung surfactant constituents to investigate the ability of probes to preferentially detect bacteria in the presence of large amounts of surfactant. The nature of the fluorescent reporter (NBD) incorporated in NBD-UBI.sub.dend and NBD-PMX, suggested the possibility of fluorescent activation in the hydrophobic rich surfactant environment. A suspension of surfactant constituents in buffered saline was prepared (20 mg/ml, 65% dipalmitoylphosphatidylcholine, 30% phosphatidylglycerol, 5% palmitic acid with 1 mg/ml tyloxapol as a spreading agent) and incubated with and without A549 epithelial cell monolayer. Particles of surfactant constituents seen to coat the epithelial cell-surface were fluorescent suggesting the NBD fluorescence increases in this hydrophobic solution. We clearly demonstrate that NBD-UBI.sub.dend and NBD-PMX both possess selectivity for bacterial labelling over lung surfactant constituents. At equivalent molarity, the NBD-UBI.sub.dend has significantly improved bacterial selectivity over linear NBD-UBI.sub.Nle in the presence of surfactant constituents (FIG. 15). Despite a mechanism relying on fluorescence increase in hydrophobic environments by NBD, both NBD-UBI.sub.dend and NBD-PMX preferentially increase in fluorescence on bacteria over purified surfactant constituents, in keeping with their ability to image bacteria in the distal lung. Although the comparative distribution and relative abundance of surfactant constituents in the lung is unknown it is entirely likely that this contributes to the background fluorescence that is observed upon probe instillation and that retention of labelling in vitro in the presence of surfactant is required for the probe to image successfully, providing it is sufficiently resistant to degradation and retains labelling upon wash-off.

Elastase (HNE) Sensitive Probes

Example 1

(48) A first example of elastase sensitive probes (shown schematically in FIG. 16A) comprises methyl red (acting as a dark quencher) connected via the peptide sequence AAPV (SEQ. ID NO: 9). (acting as the enzyme cleavable peptide sequence of the cleavable linker) to 7-nitrobenz-2-oxa-1,3-diazole (NBD) and the ubiquicidin fragment UBI.sub.29-41 (together acting as the probe element). Accordingly, the NBD fluorescence is quenched by the methyl red and no fluorescent signal will be observed (i.e. the methyl red and NBD are acting as a FRET pair) and therefore, whether or not the probe is bound to any bacteria that may be present in the target area via UBI.sub.29-41, no signal is observed using confocal microscopy.

(49) When the probe is in the presence of human neutrophil elastase (HNE), such as in inflamed tissue, HNE cleaves the peptide sequence of the linker, to thereby free the probe element from the methyl red quencher. Accordingly, the NBD fluorophore is no longer quenched and produces a fluorescent signal. In addition, due to the environmental sensitivity of the NBD, the signal produced is greatly amplified if the NBD is in a hydrophobic environment, such as within a cell membrane.

(50) FIG. 16B shows confocal images for a sample comprising the probe incubated with a co-culture of bacteria and neutrophils. These images confirm the ability of the probe to be cleaved by neutrophil derived HNE and to label bacteria. When the elastase inhibitor sivelestat was added to the medium, no labelling was observed as the probe remains uncleaved (and therefore the NBD is quenched by the methyl red).

Example 2

(51) In a second example, the probe comprises carboxytetramethylrhodamine (TAMRA, acting as a fluorescent quencher) connected to NBD and polymyxin (acting as the probe element) via the peptide sequence AAPV (SEQ ID NO: 9) (acting as the enzyme cleavable peptide sequence of the cleavable linker). Accordingly, the NBD fluorescence is quenched by the TAMRA to give rise to a fluorescent signal from TAMRA (i.e. the TAMRA is accepting the energy absorbed by the NBD and is itself fluorescing, TAMRA and NBD are acting as a FRET pair). Accordingly, whether or not the polymyxin has bound to any bacteria that may be present, only a signal from the TAMRA is observed.

(52) Once the probe has been cleaved by elastase, the bacteria are labelled by the probe element due to the fluorescence of NBD (FIG. 17A), whereas the neutrophils, once activated, are labelled by the TAMRA moiety (FIG. 17B). Furthermore, we demonstrate the fluorescence of the cleaved TAMRA compound increases in an elastase dependent manner and are inhibited with an elastase inhibitor such as sivelestat (S, FIG. 17C).

Data Analysis

(53) We have begun to develop bespoke image analysis/processing strategies to perform rapid real-time objective analysis of the large datasets generated by the probe/FCFM platform. Unequivocal detection of bacteria and the delineation of their Gram status will be achieved by employing these image processing algorithms in real-time. A processing algorithm based upon single frame analysis has been applied to entire video sequences (up to 3500 frames), and even at this early stage we are able to unequivocally delineate bacterial presence and Gram status. These signal and image processing algorithms will be rapidly iterated in readiness for clinical application, and we expect significant components of machine-learning to be incorporated into further optimisation of NBD-UBI.sub.dend and NBD-PMX datasets.

(54) FIG. 18 shows the stepwise processing for single positive and negative frames that were initially chosen to develop the model, based on two image processing algorithms (The Laplacian of Gaussian (LoG) and the Difference of Gaussians (DoG)) commonly used to detect bright spots in fluorescence microscopy images in an objective computational manner. Spots are enhanced in an image by convolving with a LoG or DoG filter. Pre-processing is necessary to improve the accuracy by removing the influence of noise in the image. First, background noise is estimated by dividing an image into small windows (100 per frame) and then calculating the standard deviation of the pixel intensities. High-pass filtering is used to decrease low frequency noise. Thresholding the resulting image at 3 standard deviation removes background noise and keeps only the higher pixel intensity values. Finally, spots are detected with LoG or DoG.

(55) We have employed the method described above to analyse entire FCFM image videos (up to 3500 frames). With an initial thresholding limit of 80 spots per frame to indicate a positive frame, we determined the percentage positive frames per video and set an arbitrary cut-off of 20% as the threshold for a binary outcome of YES/NO. This shows unequivocal detection of bacteria and Gram status using NBD-UBI.sub.dend and NBD-PMX respectively (FIG. 19).

(56) To ensure the compounds can label bacteria above the normal human lung autofluorescence, bacteria were incubated with human lung tissue and imaged using FCFM and on benchtop confocal (FIG. 20) demonstrating that bacteria can be detected, once labelled, above the level of autofluorescence.

(57) We have identified a number of determinants of distal lung in situ labelling with probes. These explain why the promising in vitro data for the linear NBD-UBI.sub.Nle probe did not translate to reproducible in situ labelling in the ovine lung. Structural variants were synthesised and assessed, exploring the structure-activity-relationship. The initial aims were to improve resistance to degradation and improve signal-to-noise. Certain modifications which greatly improved a single functional aspect such as stability (such as insertion of D-amino acids/N-methyl or variants including exclusive D-amino acid variants) or labelling intensity did not permit reproducible visualisation of bacteria in the distal lung. We comprehensively assessed the UBI analogues for stability in ALI BALF, in vitro labelling of bacteria using live benchtop confocal imaging at 37 C. and in situ labelling in the ovine lung.

(58) A number of structural modifications have been undertaken which are broadly divided into two groups:

(59) (a) Increase the signal-to-noise ratio by examining different environmentally-sensitive fluorophores or increasing the NBD-UBI.sub.Nle payload.

(60) We have assessed the utility of including two NBD fluorophores for each UBI fragment and have also assessed this with an N-methylated amino acid variant designed to enhance stability. We have assessed a series of alternative environmentally-sensitive fluorophores, including malachite green and styryl-based dyes, with the aim of producing a higher signal-to-noise ratio, which may enable lower levels of bacterial detection or detection at lower effective probe concentrations.

(61) (b) Improve the resistance to proteolytic degradation.

(62) We have assessed a number of compounds including variants incorporating D-amino acids and N-methylated amino acids at selected positions identified by MALDI analysis of the parent NBD-UBI compound as sites susceptible to proteolytic degradation. To reduce degradation without altering the amino acid constituents we have synthesised variants including PEG units at the amino and/or carboxy termini in order to block degradation from the ends of the peptide sequence. We have also synthesised and assessed variants consisting entirely of D-amino acids and D-amino adds with inversion of sequence with or without blocking PEG units. Furthermore, a cyclic variant of NBD UBI.sub.Nle has also been assessed as well as further variants of this compound incorporating IN-methylated amino adds at selected positions identified by MALDI analysis of the cyclic variant of NBD-UBI.sub.Nle as remaining susceptible to proteolytic cleavage

(63) ##STR00007##

(64) All compounds have undergone biological assessment. For stability we have assessed each compound in the presence of 0.9% NaCl (Saline) or pooled lavage fluid from patients with acute lung injury and analysed by matrix-assisted laser desorption/ionization (MALDI) or fourier transform mass spectrometry (FTMS). In vitro labelling was assessed on benchtop confocal in the presence of compound with bacteria and labelling was compared to NBD-UBI.sub.Nle which served as a reference control for bacterial labelling. Where appropriate we have also assessed labelling of the compound on isolated human neutrophils and primary human cell lines (A549 human lung adenocarcinoma cell line). For the ex vivo and in vivo ovine lung experiments each compound was assessed in a control lung segment (instilled with 2 ml PBS) or a bacterial segment (instilled with 2 ml of 2 optical density of bacteria). Following bacterial instillation, the compound was administered to the segment of interest and this was imaged by probe based FCFM.

(65) Assessment of the different NBD-UBI.sub.Nle variants has given us a clearer understanding of different mechanistic factors affecting the function of the probe in the lung environment, and specifically clarified the reasons why, out of all the UBI variants, only the NBD-UBI.sub.dend is able to image bacteria in the lung.

(66) The alternative fluorophores were inferior to NBD for this application. A Malachite Green variant gave a much lower labelling intensity on the bacteria and although the Styryl-dye compounds exhibited an increased intensity of labelling on bacteria these compounds had a decrease in selectivity over mammalian cells, most likely due to the propensity of the dyes themselves to enter lipophilic membranes overcoming the targeting of the ubiquicidin moiety. Consequently, the styryl-dye variants exhibited greatly increased off-target labelling in the ex vivo lung and no bacterial signal was observed.

(67) The D-amino acid variants as well as the variants with PEG blocking groups at the ends of the peptide sequence exhibited reduced labelling in vitro and no labelling ex vivo. Despite improved stability, of the linear UBI variants which retained function in vitro none of these were able to image bacteria in the lung. We have obtained evidence that the wash-off properties of the probes, most likely related to affinity and/or the nature of subsequent insertion into the bacterial membrane, impact on whether or not they can be used to successfully image in the lung. The retention of labelling upon removal of probe solution was investigated in vitro by confocal. With all of the linear UBI variants labelling was lost completely upon wash-off. This suggests that, in the lung, labelling would be rapidly lost once the probe concentration around the bacteria decreased as a consequence of fluid dissipation. Bacteria were pre-labelled with these linear NBD-UBI.sub.Nle compounds and successful labelling was confirmed by imaging the bacterial suspension pre-instillation. These suspensions were instilled into the ex vivo lung and when the segment was subsequently imaged by FCFM no bacterial signal was detected (there is an inherent time-delay in change-over from the instillation catheter to passage of the FCFM fibre into the same segment). However when bacteria pre-labelled with NBD-PMX, which didn't lose labelling upon wash-off when assessed by in vitro confocal, were instilled a bacterial signal was detected. The NBD-UBI.sub.dend construct, as well as giving an increased bacterial signal at equimolar concentrations, retained labelling upon wash-off. As predicted, when bacteria pre-labelled with these compounds were instilled into the ex vivo lung the bacteria were successfully imaged.

(68) We tested the compounds ability to label non-bacterial pulmonary pathogens, including fungi. Within immunocompetent patients the development of VAP secondary to eukaryotic fungi is uncommon, and accounts for <2% of VAP cases. The most commonly isolated pathogens are Candida albicans (C. albicans) and Aspergillus fumigatus (A. fumigatus). C. albicans colonises up to 50% patients in the ICU and colonisation increases with antibiotic use. However, the development of invasive candidiasis as a cause of VAP is rare and remains contentious. Indeed some large clinical trials have excluded candida isolation in their diagnostic algorithms for VAP and post-mortem studies have demonstrated growth of candida from BAL is not a reliable marker of candida infection in the immunocompetent patient. A fumigatus is also occasionally isolated in 1-2% of respiratory samples in critically ill patients. However, of this 1-2% only 20% are believed to be pathogenic and the remaining 80% are believed to colonisation.

(69) NBD-UBI.sub.dend demonstrated no labelling of C. albicans, however, there was variable labelling of A. fumigatus hyphae upon imaging (FIG. 21). Furthermore, imaging demonstrated the fungal hyphae are significantly larger than bacteria and show a distinct morphological appearance. NBD-PMX demonstrated an identical pattern to NBD-UBI.sub.dend when imaged on confocal with the retention of labelling of Gram-negative bacteria, variable labelling of A. fumigatus hyphae and no labelling of C. albicans (FIG. 22).

(70) Finally, to detect if this pattern would be identified in lung tissue, A. fumigatus was co-cultured with ovine lung and the probes and imaged with FCFM. This again demonstrated the size and pattern of the fungal hyphae to be distinct from a bacterial signal (FIG. 24). Similar experiments were conducted with suspensions of A. fumigatus and demonstrate a distinct characteristic, linear pattern of fluorescence which differs to a bacterial signal (FIG. 25).

(71) NBD-PMX and NBD-UBI.sub.dend were assessed for direct red cell toxicity by a haemolysis assay and demonstrated no red cell haemolysis up to 100 M (FIG. 26). The compounds were assessed for toxicity in a rodent single dose, intratracheal toxicity assessment. Mice received a single intratracheal administration of 100 g/25 g mouse (3000 fold human dose assuming 100 g delivered to a 75 kg man) of NBD-UBI.sub.dend or NBD-PMX, or PBS control. Mice were then euthanized at 48 hours and 14 days (n=3 per group). The data shows no toxicity for NBD-PMX (Table 2 and FIG. 27A) or NBD-UBI.sub.dend (Table 1 and FIG. 27B).

(72) TABLE-US-00001 TABLE 1 NBD-UBI.sub.dend 48 hours 14 days PBS Control NBD-UBI.sub.dend PBS Control NBD-UBI.sub.dend Cytospin (% of 96.5/0.7/0 96.4/3.6/0 ns 100/0/0 100/0/0 ns Mononuclear 0.7/0.67/0 2.0/2.0/0 0/0/0 0/0/0 cells/ Neutrophils/Red Blood Cells) BALF (Cells/ul) 354.9 66.6 477.1 150.6 ns 361.1 38.6 342.7 6.5 ns PenH value 0.5 0.04 0.5 0.06 ns 0.38 0.03 0.5 0.04 ns Creatinine (u/l) 8.0 0.6 8.3 0.7 ns 12.7 1.7 9.0 1.5 ns Bilirubin (u/l) 2 0 2 0 n/a* 2 0 2 0 n/a* ALT (u/l) 33.3 3.9 24.0 1.2 ns 39.3 10.8 28.3 1.9 ns ALP (u/l) 120.0 6.2 222.3 80.3 ns 178.3 27.4 227.0 27.1 ns Albumin (u/l) 21.0 1.0 22.3 1.8 ns 24.3 0.7 25.0 1.2 ns ns = not significant. *Bilirubin for these time points was below 2 for all animals.

(73) TABLE-US-00002 TABLE 2 NBD-PMX 48 hours 14 days PBS Control NBD-PMX PBS Control NBD-PMX Cytospin (% of 97.1/0.3/2.6 96.1/0/3.9 ns 100/0/0 93.0/0/2.3 ns Mononuclear 2.9/0.3/2.6 3.9/0/3.9 0/0/0 2.3/0/2.3 cells/ Neutrophils/Red Blood Cells) BALF (Cells/ul) 243.1 80.5 509.0 270.2 ns 478.4 88.4 542.0 84.7 ns PenH value 0.7 0.1 0.5 0.05 ns 0.4 0.03 0.5 0.07 ns Creatinine (u/l) 35.3 28.9 16.3 8.8 ns 8.7 1.8 9.7 0.3 ns Bilirubin (u/l) 2.4 0.4 1.2 0.4 ns 2.4 0.3 1.2 0.4 ns ALT (u/l) 43.0 14.0 36.0 9.0 ns 32.3 7.9 40.3 4.8 ns ALP (u/l) 100.0 18.7 107.7 17.3 ns 89.0 8.1 126.3 8.5 ns Albumin (u/l) 24.7 0.3 23.7 0.3 ns 24.0 0.6 25.3 0.7 ns ns = not significant.

Methods of Synthesis of Probes

(74) Synthesis of Ubiquicidin Based Elastase Probes (Methyl Red (MR)-AAPV-NBD-UBI.sub.29-41 and TAMRA-AAPV-NBD-UBI.sub.29-41) (SEQ ID NO: 9)

(75) MR-AAPV-K(NBD)-PEG-OH (SEQ ID NO: 20) (AL3-74) fragment was synthesised on solid-phase employing Fmoc-strategy, with standard amino acid coupling cycles (230 min at rt) with DIC and oxyma in peptide grade DMF at 0.1 mM reagent concentration. Fmoc deprotection steps were done in 20% piperidine in DMF (230 min). Between each step, the resin was washed with DMF, DCM and MeOH.

(76) 2 g of chlorotrityl polystyrene resin (loading 0.3 mmol/g) was treated with Fmoc PEG-OH (3 eq) and DIPEA (6 eq) in anhyd. DCM (2 mL) for 3 h. After washing and Fmoc deprotection, the Fmoc-AAPVK(Dde) sequence (SEQ ID NO: 20) was synthesised as described above using, Fmoc-Lys(Dde)-OH, Fmoc-Val-OH, Fmoc-Pro-OH, and Fmoc-Ala-OH, After the sequence was completed, the synthesis was continued with half of the resin (0.3 mmol scale) and Ode protecting group was orthogonally removed with NH.sub.2OH/imidazole in NMP/DCM (290 min). The resin was treated with NOB-Cl (3 eq) and DIPEA (6 eq) in DMF (245 min). After Fmoc deprotection, the synthesis was continued in 0.15 mmol scale and Methyl Red was coupled to the N-terminus as described above, After washing, the fragment was cleaved off the resin with TFA-TIS-H.sub.2O (95:2.5:2.5) (30 min) and precipitated with cold ether to give AL3-74 (ESI-MS 1044.4 and 1066.4).

(77) ##STR00008##

(78) UBI.sub.29-41 sequence was synthesised on Rink-amide ChemMatrix resin (loading 1 mmol/g) as described below. Next, AL3-74 (0.055 mmol) in anhyd. DMF (0.6 mL) was added to Ubi.sub.29-41 on a ChemMatrix resin AL3-68 (0.03 mmol), followed by addition of HBTU (0.055 mmol) and DIPEA (0.22 mmol). The reaction mixture was shaken overnight covered from light. After filtration, the resin was washed with DMF, DCM and MeOH. The resin was swollen with DCM and the probe was deprotected and cleaved off the resin with TFA/thioanisole/EDT/anisole (90:5:3:2) (3 h). The crude precipitated with cold ether and collected by centrifugation. The product AL3-79 was purified by preparative HPLC with detection at 490 nm and gradient of H.sub.2O-ACN with 0.1% formic acid as an eluent. MALDI-TOF MS 2719.4, >95% HPLC purity.

(79) TAMRA-AAPV-NBD-UBI29-41 (SEQ ID NO: 9) AL3-88 (Maldi-TOF MS 281.5, >95% HPLC purity) was synthesised in similar manner expect fragment TAMRA-AAPV K(NBD)-PEG-OH (SEQ ID NO: 20) (AL3-75) was coupled to the N-terminus of UBI-based peptide on resin AL3-68.

(80) ##STR00009##

Synthesis of Polymyxin-Based NeBac-Probe

(81) ##STR00010##

(82) TAMRA-AAPV-K(NBD)-PEG-OH (SEQ ID NO: 20) fragment 3A was synthesised on solid-phase employing Fmoc-strategy, with standard amino acid coupling cycles (230 min at rt) with DC and oxyma in peptide grade DMF at 0.1 mM reagent concentration. Fmoc deprotection steps were done in 20% piperidine in DMF (230 min). Between each step, the resin was washed with DMF, DC M and MeOH.

(83) 500 mg of chlorotrityl polystyrene resin (loading 0.3 mmol/g) was treated with Fmoc-PEG-OH (3 eq) and DIPEA (6 eq) in anhyd. DCM (2 mL) for 3 h, After washing and Fmoc deprotection, the Fmoc-AAPVK(Dde)-(SEQ ID NO: 20) sequence was synthesised as described above using, Fmoc-Lys(Dde)-OH, Fmoc-Val-OH, Fmoc-Pro-OH, and Fmoc-Ala-OH. After the sequence was completed, Dde protecting group was orthogonally removed with NH.sub.2OH/imidazole in NMP/DCM (290 min). The resin was treated with NDB-Cl (3 eq) and DIPEA (6 eq) in DMF (245 min). After Fmoc deprotection, 5(6) carboxyTAMRA was coupled to the N-terminus as described above. After washing, the fragment was cleaved off the resin with TFA-TIS-H.sub.2O (96:2.5:2.5) (30 min) and precipitated with ether. 3A ESI-MS1044.4, and 1066.4.

(84) Next, to 3A (0.011 mmol) in anhyd. DMF (0.5 mL), HSPyU (0.011 mmol) and DIPEA (0.033 mmol) were added, and the reaction was stirred at rt for 1 h. Boc-protected Polymyxin (15 mg, 0.012 mmol in 0.5 mL DMF) and DIPEA (0.033 mmol) were added, and the reaction mixture was stirred overnight covered from light. DMF was evaporated, the crude dissolved into 1 mL TFA-DCM (1:1), and stirred for 90 min. TFA-DCM was evaporated, the crude precipitated with cold ether, and collected by centrifugation. The product was purified by preparative HPLC with detection at 490 nm and gradient of H.sub.2O-ACN with 0.1% formic acid as an eluent. Maldi-TOF MS 2151.6 and 2173.6, 100% HPLC purity.

Synthesis of NBD-UBIdend

Synthesis of Monomer (5)

(85) Synthesis required the preparation of the monomer (5) which was synthesised in six steps.sup.1 as shown in Scheme 1. Monomer (5) was prepared by the 1,4 addition of the hydroxy groups of 1,1,1-tris(hydroxymethyl)amino-methane onto acrylonitrile, followed by amino group protection (Boc). Hydrogenolysis of the nitrile groups with PtO.sub.2/H.sub.2 gave (3) which was treated with DdeOH to give the tris-Dde protected amine (4). Following removal of the Boc protecting group, the isocyanate (5) was prepared following the procedure of Knlker..sup.2

(86) ##STR00011##

Fmoc-Rink Amide ChemMatrix Resin (6)

(87) 4-[(2,4-Dimethoxyphenyl)-(Fmoc-amino)methyl]phenoxyacetic acid (Rink amide linker) was attached to ChemMatrix resin (LV=1 mmol/g). Thus the Fmoc-Rink-amide linker (0.2 mmol, 1 eq) was dissolved in DMF (4 mL) and ethyl oximinocyanoacetate (Oxyma) (0.2 mmol, 1 eq) was added and the mixture was stirred for 5 min. N,N-Diisopropylcarbodiimide (DIC) (0.2 mmol 1 eq) was then added and the resulting mixture was stirred for a further 2 min. The solution was added to ChemMatrix resin (0.1 mmol, 1.0 mmol/g, 1 eq) and shaken for 0.5 hour. The resulting resin was washed with DMF (35 mL), DCM (35 mL) and MeOH (35 mL). The coupling reaction was monitored by a quantitative ninhydrin test.sup.3.

(88) ##STR00012##

(89) The probe was synthesised on a ChemMatrix resin derivatized with an Fmoc-Rink Amide type linker (Scheme 2). The linker (6) was loaded with monomer (5) to give the tri-branched scaffold (7). Following the removal of the Dde groups (2% hydrazine in DMF) the appropriate Fmoc-Amino acids were coupled sequentially followed by the attachment of 4-PEG-7-nitrobenzofurazan N-hydroxysuccinimide ester (NBD-PEG-NHS) and cleaved from the resin using TFA/TIS/DCM (90/5/5).

General Procedure for the Fmoc Deprotection

(90) To the resin (pre-swollen in DCM) was added 20% piperidine in DMF (5 mL) and the reaction mixture was shaken for 10 min. The solution was drained and the resin was washed with DMF (310 mL), DCM (310 mL) and MeOH (310 mL). This procedure was repeated twice. The coupling reaction was monitored by a quantitative ninhydrin test.sup.3.

Isocyanate Coupling to Give (7)

(91) To resin (0.30 mmol), pre-swollen in DCM (10 mL), was added a solution of isocyanate (6) (920 g, 0.93 mmol), DIPEA (0.2 mL, 0.93 mmol) and DMAP (22 mg, 0.17 mmol) in a mixture of DCM/DMF (1:1, 5 mL) and the mixture was shaken overnight and the reaction monitored by a quantitative ninhydrin test. The solution was drained and the resin was washed with DMF (320 mL), DCM (320 mL) and MeOH (320 mL) and ether (320 mL). (320 mL). The coupling reaction was monitored by a quantitative ninhydrin test.sup.3.

PEG Coupling8-(9-Fluorenylmethyloxycarbonyl-amino)-3,6-dioxaoctanoic Acid (Fmoc-PEG-OH) Coupling

(92) A solution of Fmoc-PEG-OH (3.0 mmol, 10 eq) in DMF (3 mL) and Oxyma (3.0 mmol, 10 eq) was added and the mixture was stirred for 5 min. DIC (3.0 mmol, 10 eq) was then added and the resulting mixture was stirred for a further 2 min. The solution was added to pre-swollen resin (7) in DCM and the reaction mixture was shaken for 0.5 h. The solution was drained and the resin was washed with DMF (310 mL), DCM (310 mL) and MeOH (310 mL). The coupling reaction was monitored by a quantitative ninhydrin test.sup.3.

Peptide Synthesis

(93) Peptide Sequence: Thr-Gly-Arg-Ala-Lys-Arg-Arg-Nle-Gln-Tyr-Asn-Arg-Arg (SEQ ID NO: 3) A solution of the appropriate Fmoc-amino acid (3.0 mmol, 10 eq) (Fmoc-Arg(Boc)-OH, Fmoc-Asn(Trt)-OH, Fmoc-Tyr(tBu)-OH, Fmoc-Gln(Trt)-OH, Fmoc-Nle-OH, Fmoc-Lys(Boc)-OH, Fmoc-Ala-OH, Fmoc-Gly-OH, Fmoc-Thr(tBu)-OH) and Oxyma (3.0 mmol, 10eq) was added and the mixture was stirred for 5 min, DIC (3.0 mmol, 10 eq) was then added and the resulting mixture was stirred for a further 2 min. The solution was added to pre-swollen resin in DCM and the reaction mixture was shaken for 0.5 h. The solution was drained and the resin washed DMF (320 L), DCM (320 mL) and MeOH (320 mL). The coupling reactions were monitored by a quantitative ninhydrin test.sup.3.

7-Nitrobenzofurazan (NBD) Coupling

(94) To a solution of NBD-PEG-NHS (3.0 mmol, 10 eq) in DMF (3 mL) was added DIPEA (3.0 mmol, 10 eq). The resulting solution was added to resin (1 eq), pre-swollen in DCM, and the reaction mixture was shaken for 0.5 h. The solution was drained and the resin washed with DMF (3), DCM (3) and MeOH (3). The coupling reaction was monitored by a quantitative ninhydrin test.sup.3.

TFA Cleavage and Purification of Reporter NBD-UBIdend

(95) The resin (45 mg), pre-swollen in DCM, was treated with a cleavage cocktail of TFA/TIS/DCM (90/5/5, 300 L) for 2.5 h. The solution was drained and the resin was washed with the cleavage cocktail and the solution was removed in vacuo. The crude material was dissolved in a minimum amount of cleavage cocktail (50 L) and added to ice-cold ether (7.5 mL). The precipitated solid (22 mg) was collected by centrifugation and the solvent removed by decantation and the precipitate was washed with cold ether (35 mL). The precipitate was then purified by preparative reverse phase HPLC and the required fractions were pooled and lyophilized to afford NBD-UBI.sub.dend.

(96) ##STR00013##

Synthesis of NBD-PMX

(97) The NBD-PMX probe was synthesised from its precursor Polymyxin B sulfate in four steps (Scheme 6). The probe and its intermediates were synthesised using reported methods.sup.1 with moderate modifications. The fluorophore is incorporated as an amide coupling between the NHS ester of the NBD-PEG and the tetra-Boc polymyxin C. NBD-PMX probe is obtained after the TFA cleavage and HPLC purification.

(98) ##STR00014## ##STR00015##

Preparation of Compound B

(99) Polymyxin B sulfate (10 g, 7.7 mmol, 1 eq) was dissolved in deionized water (200 mL) at a pH of 6.5 (use HCl aq solution to adjust the pH). Papain (1.5 g) was dissolved in water (25 mL) (same pH). The solutions were combined and toluene (0.5 mL) was added, and the mixture was gently stirred at 65 C. overnight. The mixture was then stirred in boiling water for 5 min and the precipitate formed (denatured papain) was removed by centrifugation and filtration. The filtrate was concentrated in vacuo and freeze dried to give the crude product B in quantitative yield. This step was carried forward to the next step without any further purification. MS m/z 963.2 (100%, [M+H].sup.+).

Preparation of Compound C

(100) Crude B (5.5 g, 5.7 mmol, 1 eq) was dissolved in a mixture of H.sub.2O:Dioxane:Et.sub.3N (150 mL, 1:1:1) and Boc-ON (4.52 g, 17.1 mmol, 3 eq) was added. The solution was stirred for 20 min at room temperature and then quenched with methanolic ammonia (20 mL, 2M ammonia in MeOH). The reaction was followed up by ELSD. Solvents were evaporated and the resulting mixture was subjected to silica gel chromatography column (MeOH:DCM, 15:85) to afford white solid B (1.7 g, 22%). MS m/z 1363.7 (100%, [M+H].sup.+).

N-(4-Nitrobenz-2-oxa-1,3-diazol-7-yl)amino-3,6-dioxaoctanoic Acid (NBD-PEG-OH):2

(101) DIEA (850 l, 5.00 mmol) and solid 8-Amino-3,6-dioxaoctanoic acid (NH.sub.2-PEG-OH) (392 mg, 2.40 mmol, 1 eq) were added slowly, over an hour, to a solution of NBD-Cl (401 mg, 2.01 mmol) in methanol (20 mL) at 0 C. The reaction mixture was stirred overnight at room temperature. The solvent was evaporated and the remaining material was purified by chromatography on silica with DCM/MeOH (8:2) as the eluent to give NBD-PEG-OH (400 mg, 1.23 mmol, 51%) as dark red oil. .sup.1H NMR (500 MHz, DMSO): 10.9 (s, 1H; COOH), 8.49 (d, J=8.5 Hz, 1H; CH NBD), 7.1 (s, 1H, NH), 6.23 (d, J=8.5 Hz, 1H; CH NBD), 4.25 (s, 2H), 3.93 (t, J=5.3 Hz, 2H; CH.sub.2), 3.80 (s, 4H), 3.72 (t, J=6.8 Hz, 2H; CH.sub.2) ppm; MS (ESI): m/z calcd for C.sub.12H.sub.14N4O.sub.7 [MH]: 325.1; found: 325.2.

N-(4-Nitrobenz-2-oxa-1,3-diazol-7-yl)amino-3,6-dioxaoctanoic Acid, Succinimidyl Ester (NBD-PEG-NHS)

(102) To a solution of NBD-PEG-OH (2.4 g, 7.4 mmol, 1 eq) in anhydrous DCM (500 mL) was added EDC.HCl (1.56 g, 8.18 mmol, 1.1 eq) and DIPEA (1.36 mL, 10 mmol). After stirring the mixture for 10 min, N-hydroxysuccinimide (0.94 g, 8.18 mmol) was added and allowed to stir for 16 h. The reaction mixture was diluted with DCM (250 mL) and treated with 5% aqueous citric acid (2200 mL), sat. aqueous NaHCO.sub.3 and brine. The organic layer was dried over Na.sub.2SO.sub.4, filtered and reduced in vacuo to afford product as dark brown solid (1.0 g, quantitative). The crude was used for next step without further purification.

Preparation of F

(103) A solution of NBD-PEG-NHS (466 mg, 1.1 mmol, 1 eq), DIPEA (384 L, 2.2 mmol, 2 eq) and amine C (1.5 g, 1.1 mmol, 1 eq) in DMF (150 mL) was stirred at room temperature for 1 h and protected from light. After completion of the reaction (TLC), volatiles are removed under vacuum. The crude mixture was purified by flash chromatography (DCM:MeOH, 90:10) to afford dark orange/brown solid (1.2 g, 65%). HPLC (254 nm & 495 nm) Rt=7.80 min; m/z 1671.7 (25%, [M+H].sup.+); 1693.9 (65%, [M+Na].sup.+).

Preparation of NBD-PMX Probe

(104) A solution of Boc-protected polymyxin F (150 mg, 0.09 mmol) in 20% TFA in DCM (2 mL) was vigorously stirred for 45 min at room temperature and protected from light. The reaction mixture was evaporated in vacuo and the resultant was dissolved in ether. Ether layer was decanted after centrifugation (32 mL). The resultant yellow/brown solid (40 mg, quantitative) was dried under vacuum. The crude product was purified by preparative HPLC in MeOH/H.sub.2O as gradient solvent system with 0.1% formic acid as an additive. The fractions collected from prep-HPLC were freeze dried to afford red/orange solid (30 mg, 26% recovery from HPLC).

(105) Characterisation:

(106) For analytical HPLC, a Poroshell 120 SB-C18, 2.7 m, 4.650 mm column was used with a diode array detector. For prep-HPLC method: Discovery C18 reverse-phase column (5 cm4.6 mm, 5 m) with a flow rate of 1 mL/min and eluting with H2O/MeOH/HCOOH (95/5/0.05) to H2O/MeOH/HCOOH (5/95/0.05), over 6 min, holding at 95% MeOH for 4 min, with detection at 254 and 495 nm and by ELSD. HPLC (495 nm): Rt=4.1 min; MS m/z 1271.7 (95%, [M+H].sup.+); 1293.7 (100%, [M+Na].sup.+); FTMS calc. 636.3282 ([M+2H]/2).sup.+, found 636.3344.

(107) Absorption/Emission: 467 nm/539 nm.

(108) Solubility: Fully soluble in water.

(109) Stability: stable at room temperature for > than 1 week.

(110) Storage: Stored at 20 C. under inert atmosphere. Protect from light.

Biological Methods

(111) Bacterial Growth:

(112) Bacteria used in assays include Pseudomonas aeruginosa (PA01-reference strain and J3284-clinical isolate from VAP patient), Acinetobacter baumannii, Stenotrophomonas maltophilia, Staphylococcus aureus (Inc. methicillin-resistant S. aureus (MRSA), methicillin-sensitive S. aureus (MSSA)), Klebsiella pneumoniae, Escherichia coli, Haemophilus influenzae and Streptococcus pneumoniae.

(113) TABLE-US-00003 TABLE 3 Bacteria, strain reference and original source used in experiments. Bacteria Strain Original Source Gram- P. aeruginosa ATCC 47085 ATCC negative (PA01) bacteria P. aeruginosa J3284 Clinical Isolate* A. baumannii J3433 Clinical Isolate* S. maltophilia J3270 Clinical Isolate* K. pneumoniae ATCC BAA1706 ATCC E. coli ATCC 25922 ATCC H. influenzae Clinical Isolate Clinical Isolate* Gram- Methicillin ATCC25923 ATCC positive Resistant S. aureus bacteria (MRSA) Methicillin Sensitive ATCC 252 ATCC S. aureus (MSSA) S. pneumoniae D39 NCTC 7466 Health protection agency culture collection GFP fluorescent S RN6390-Gfp- Nottingham aureus EryR University (Gift from Professor Phil Hill) *Gifts from Professor John Govan, University of Edinburgh.

(114) All bacteria were grown on agar broth, chocolate agar or blood agar plates, stored at 4 C. For assays, a single colony of bacteria was taken using an inoculating loop and added to 10 ml liquid broth in a 50 ml Falcon Tube. This was transferred to an incubator at 37 C. for 16 hours (for Streptococcus pneumoniae supplemented with 5% CO.sub.2). Cultures were either used as overnight cultures (stationary phase) or from these cultures a sub-culture was taken (1:100) and the sample was grown until they entered mid log phase (reads of 0.5-0.6 optical density (OD) on spectrophotometer at 595 nm). The culture was then centrifuged at 4000 rpm for 5 minutes and pellet resuspended in phosphate buffered saline (PBS). Following three washes this was reconstituted to 0.5 OD.sub.595nm for confocal assays, 0.1 OD.sub.595 nm for flow cytometry or 2 OD.sub.595nm for ovine ex vivo lung experiments (unless otherwise stated).

(115) Bacterial Counting:

(116) Samples (prepared bacteria or lavage from ovine lung segments) were vortexed briefly then serial dilutions (1:10) were performed to dilutions to the 8th dilution. The broth/blood agar plate was divided into quadrants with 520 ul drops in each quadrant. These were incubated at 37 C. for 16 hours (for Streptococcus pneumoniae supplemented with 5% CO.sub.2) and plates were counted with data reported as colony forming units per millilitre (CFU/ml).

(117) Surfactant Constituent Synthesis:

(118) Surfactant 5 g 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and 2.5 g L--Phosphatidyl-DL-glycerol sodium salt (from egg yolk lecithin; PG) were dissolved in 500 l chloroform and evaporated under nitrogen to a thin lipid film in a round bottom flask. The lipid film was rehydrated with PBS at 48 C. for 1 hour with agitation (750 rpm) to generate multilammelar vesicles (MLV). These were diluted 1:4 for use in confocal experiments.

(119) Agarose Bacterial Beads:

(120) Bacteria were grown to midlog phase in 400 ml TSB, pelleted by centrifugation and resuspended in 2 ml PBS. This was mixed with 18 ml molten tryptic-soy agar (50 C.) and injected rapidly into vortexing mineral oil+0.01% Span 80, pre-warmed to 50 C. This was then rapidly cooled to 4 C. whilst continuing to vortex to allow the beads to set. Bacterial agar beads were pelleted by centrifugation (20 minutes, 3000 g) and washed in 0.5% sodium deoxycholate (SDC) in PBS (20 minutes, 3000 g), followed by 0.25% SDC (20 minutes, 3000 g) in PBS, washed in PBS (10 minutes, 3000 g) and 3PBS (5 minutes, 200 g). Beads were resuspended at 50% v/v in PBS for instillation.

(121) Neutrophil Extraction:

(122) Neutrophils were isolated from the peripheral blood of healthy human volunteers by dextran sedimentation followed by centrifugation through discontinuous plasma-Percoll gradients.

(123) A549 Cultures:

(124) A549 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 units/mL penicillin G and 100 g/mL streptomycin to 80% confluence. Cells were dispersed with Trypsin-EDTA and seeded onto glass coverslips or 8 well confocal imaging chambers and grown to confluence in the presence of DMEM.

(125) Confocal Analysis:

(126) Bacteria were prepared and counterstained with Syto 82 nucleic acid stain (Invitrogen, Calif., USA) in a shaking heat block at 37 C. and 350 rpm for 20 minutes. They were co-incubated with the probe at required concentration in a sealed POC mini chamber, or 8 well confocal chamber. When required the glass coverslip for the POC chamber was coated in fibronectin (for neutrophil experiments) or poly-d-lysine (for bacteria and cell lines) and incubated with cells with one hour at 37 C. to allow adherence prior to bacterial innoculation. Analysis was with ImageJ. Briefly, the Syto channel was automatically thresholded (Huang) and an ROI generated from this. The mean fluorescence intensity on the probe channel within this ROI was quantified. Data presented represents the mean of three separate fields of view.

(127) Flow Cytometry:

(128) Bacteria were prepared and counterstained with Syto 82 nucleic acid stain (Invitrogen, Calif., USA) in a shaking heat block at 37 C. and 350 rpm for 20 minutes. Bacteria were washed in PBS3, and probe (50 l) added in 50 ul OD.sub.595 1 od bacteria. This was diluted to 500 uL and analysed using BD FACS Calibur using FL-1 and FL-2 channels, with 10,000 events. Analysis was with FlowJo software following gating on the FL-2 channel.

(129) Lung Harvesting and pCLE procedure:

(130) From a cohort of surplus stock ewes which were destined for cull, one ewe was identified and terminally euthanized with an overdose of anaesthetic. Death was confirmed and the trachea was identified and clamped in situ. The thoracic cavity was then accessed and the lungs were freed from surrounding tissues and organs and the heart/lung was removed en block. The right pulmonary artery was identified, cannulated and perfused with 1000 ml 0.9% NaCl. Once filling of the left ventricle was confirmed an incision was made to allow free drainage and perfusion continued until the drainage from the left atrium was clear. The trachea was intubated with an 8.0 endotracheal tube immediately following clamp release. The lungs were placed in a neonatal incubator with an ambient temperature of 37 C. and humidity of 65% and ventilated using a Pressure Controlled Ventilator (Breas Vivo PV 403). Ventilator setting was adjusted to aid maximal parenchymal recruitment and aiming to achieve tidal volume>1 litre. Following 1 hour of optimal ventilation, bronchoscopy was undertaken and individual segments were identified and instilled with 2 ml of bacteria or PBS control. Following instillation a separate sheath (ERBE) was introduced and the probe was instilled. Then the probe-based Confocal Laser Endomicroscopy (pCLE) fibre was passed down the working channel and the segment was imaged. For BALF, the bronchoscope was wedged and 20 ml of 0.9% NaCl instilled and carefully withdrawn with lavage yields of 40-50%. Control segments were anatomically distinct and/or in the contralateral lung. The bronchoscope was decontaminated between each segment imaged.

(131) Haemolysis Assay:

(132) Erythrocytes were isolated from freshly drawn, anticoagulated human blood and resuspended to 20 vol % in PBS (pH 7.4). In a 96-well microtiter plate, 100 l of erythrocyte suspension was added to 100 l of NLLP solution in PBS (prepared by 1:2 serial dilutions) or 100 l of PBS in the case of negative controls. One-hundred percent haemolysis wells contained 100 l of red cell suspension with 100 l of 0.2 vol % Triton X-100. The plate was incubated for 1 h at 37 C., and then each well was diluted with 150 l of PBS. The plate was then centrifuged at 1,200 g for 15 min, 100 l of the supernatant from each well was transferred to a fresh microtiter plate, and A350 was measured. Percentage of haemolysis was determined as (AA0)/(AtotalA0)100, where A is the absorbance of the test well, A0 the absorbance of the negative controls, and Atotal the absorbance of 100% haemolysis wells, all at 350 nm on a Biotek plate reader.

(133) MALDI-TOF:

(134) Probe was added to saline or pooled BALF from patients with ALI incubated for 30 minutes. A ZipTip (C-18, 0.2 L) with 5 L MeCN (with 0.1% TFA as an additive) followed by 20 L of H2O was washed. The ZipTip was loaded with the sample, washed and eluted into 5 L of 80% aq. MeCN (with 0.1% TFA as an additive). The sample was analysed by MALDI-TOF (PerSeptive Biosystems Voyager DESTR MALDI-TOF mass spectrometer (Applied Biosystems, Foster City, Calif.)).

(135) Statistical Analysis:

(136) All experiments were performed at least three times unless otherwise stated and results expressed as meanSEM. Data was analysed by unpaired t-test or ANOVA, significance was determined as p<0.05 (GraphPad Prism).

REFERENCES

(137) 1. M. Ternon, J. J. Diaz-Mochon, A. Belsom, M. Bradley, Tetrahedron, 2004, 60, 8721 2. H. J. Knlker, T. Braxmeier, G. Schlechtingen, Angew. Chem. Int. Ed., 1995, 34, 2497 3. E. Kaiser, R. L. Colescott, C. D. Bossinger and P. I. Cook, Analytical Biochemistry, 1970, 34, 595-598.

LISTING OF RELEVANT SEQUENCES

(138) TABLE-US-00004 Ubiquicidin(full) SEQIDNO.1 KVHGSLARAGKVRGQTPKVAKQEKKKKKTGRAKRRMQYNRRFVNVVPTFGKKKGPNANS Ubiquicidin(UBl.sub.29-41) SEQIDNO.2 TGRAKRRMQYNRR Ubiquicidin(UBl.sub.Nle) SEQIDNO.3 TGRAKRRNleQYNRR Polymyxin SEQIDNO.4 Dab= L---Diaminobutyricacid PolymyxinB1 SEQIDNO.5 PolymyxinB2 SEQIDNO.6
SEQ ID NO 7
Mass spectrometry fragment of (UBI.sub.29-41)
YNRR
SEQ ID NO 8
Mass spectrometry fragment of (UBI.sub.29-41)
Nle-QYNRR