Compositions and methods for stability testing of botulinum toxin

Abstract

Compositions for characterization of botulinum toxin (BoNT) are described that include a genetically modified cell that is transfected with an artificial construct comprising a nucleic acid sequence that encodes for a hybrid protein having (a) a reporter-containing portion chemically coupled to (b) a cleavage site and (c) a control fluorophore. The cleavage site interacts with a BoNT in a manner that cleaves the reporter-containing portion from remainder of the construct. The cleaved portion is destroyed or otherwise degraded by the local environment, and presence of BoNT is evidenced by reduction in signal from the reporter. The cleavage sequence is all or part of a SNARE protein, the cleavable reporter-containing portion is preferably Yellow Fluorescent Protein (YFP), Citrine, Venus, or a YPet protein and the control fluorophore is preferably CFP, mStrawberry, or a mCherry protein.

Claims

1. An in vitro cell base assay method of characterizing the stability of a botulinum neurotoxin (BoNT) comprising: (i) obtaining a botulinum neurotoxin preparation comprising the BoNT; (ii) providing a recombinant cell that includes an artificial construct and an enzyme, wherein the artificial construct has (a) a reporter-containing portion comprising a first fluorophore that is degraded by the enzyme, (b) a cleavage site comprising at a portion of a motif selected from the group consisting of SNAP-25, synaptobrevin, and syntaxin, wherein the motif interacts with a portion of the BoNT in a manner that produces cleavage of the first fluorophore from a remainder of the construct, and (c) a second portion comprising a second fluorophore, wherein the cleavage site is interposed between the reporter containing portion and the second portion, wherein the recombinant cell is from a cultured cell line; (iii) obtaining a baseline signal emission measurement from the artificial construct; (iv) exposing a sample of the BoNT preparation to a temperature that exceeds a storage temperature of the BoNT preparation to generate a heated sample; (v) contacting the recombinant cell with the heated sample; (vi) obtaining a further emission measurement from the reporter fluorophore, wherein the reporter containing portion and the second portion are positioned in the construct such that FRET emission from the first fluorophore does not show a decreasing trend relative to increasing concentration of the BoNT preparation, and wherein a normalization emission measurement is obtained from the second fluorophore; and (vii) characterizing potency of the heated sample relative to that of the BoNT preparation, wherein said in vitro cell-based assay provides a correlation coefficient (r.sup.2) value of at least 0.95 relative to a pharmaceutically acceptable in vivo mouse lethality assay for the BoNT and a coefficient of variation (CV) of less than 10%.

2. The method of claim 1, wherein the BoNT is a serotype A botulinum neurotoxin.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1A and 1B schematically depict an exemplary construct and method of the inventive concept. FIG. 1A depicts a construct of the inventive concept. FIG. 1B depicts a method of the inventive concept, where the construct is cleaved by a botulinum toxin.

(2) FIGS. 2A, 2B, and 2C depict results obtained from testing of different reporting constructs. FIG. 2A shows photomicrographs of transformed cells and transformed cells following exposure to BoNT/A. FIG. 2B graphically depicts results of fluorescence emissions measurements from transformed cells exposed to BoNT/A and control transformed cells. FIG. 2C shows results of SDS-PAGE followed by immunoblotting and probing with anti-SNAP-25 of the products of exposure of various constructs to BoNT/A.

(3) FIG. 3 graphically depicts the results of fluorescence measurements from transformed cells expressing a reporting construct of the inventive concept. Results from the same set of cells are shown for the BioSentinel method (based upon characterization of YFP degradation) on the left, and for the prior art method based on FRET emission (Loss of FRET) from the construct on the right.

(4) FIG. 4 shows a graphical depiction of data comparing the performance of a cell-based assay of the inventive concept to a prior art, regulatory agency approved in vitro mouse time-of-death assay for BoNT/A potency following incubation at elevated temperatures for different periods of time.

DETAILED DESCRIPTION

(5) Various objects, features, aspects and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.

(6) FIGS. 1A and 1B diagram a construct and an exemplary assay in which BoNT/A cell-based reporters are used to detect BoNT/A activity by loss of YFP fluorescence.

(7) FIG. 1A schematically depicts BioSentinel's BoCell™ A BoNT/A construct. The reporter fluorophore, YFP, and the normalization fluorophore, CFP, are coupled by a cleavage sequence, SNAP-25 (green). SNAP-25 palmitoylation localizes the reporter to a plasma membrane.

(8) FIG. 1B shows detection of BoNT/A activity by loss of YFP fluorescence. The YFP moiety is directly excited leading to fluorescence emission in the absence of BoNT/A. Cleavage of the reporter by BoNT/A releases a C-terminal reporter fragment containing the YFP moiety into the cytosol. The fragment is rapidly degraded and, thus, YFP emission is lost. The CFP signal is still used to control for cell-to-cell reporter expression levels and cell density.

(9) Surprisingly, not all fluorescent proteins related to YFP are effective as the reporter fluorophore. For example, FIGS. 2A, 2B, and 2C provide evidence that reporters containing YFP or the closely related derivative Venus can detect BoNT/A activity in cells, but not mCherry or mStrawberry. Here, Neuro2A cells were grown in a 96-well plate to 70% confluency and transiently transfected using Lipofectamine 2000 (Invitrogen™), with reporters containing the indicated N-terminal and C-terminal (N-term/C-term) fluorophore pairs. After 24 h, cells were incubated in the presence or absence of 10 nM BoNT/A at 37° C. for 72 h in 100 μl of phenol red-free MEM medium.

(10) FIG. 2A shows BoNT/A-induced changes in fluorescence responses. Semi-automated YFP and CFP fluorescence measurements were performed using a Nikon™ TE2000-U fluorescent microscope with 20× magnification and Nikon NIS Elements 3.4 software. Shown are randomly selected fields pseudo-colored for the C-terminal/N-terminal fluorescent protein (FP) fluorescence ratio. Ratios were calculated from emissions collected upon direct excitement of each fluorophore.

(11) FIG. 2B represents fluorescence ratios and BoNT/A sensitivities of the cell-based reporters. 30 randomly selected cells per condition were analyzed for fluorescence ratios in the presence or absence of 10 nM BoNT/A. The average signal from the 30 cells from 5 microscopic fields on 3 different wells is shown. Cells exhibiting over-saturated fluorescence were excluded.

(12) FIG. 2C shows a blot showing that BoNT/A was active in cells regardless of the reporter. All reporters show some cleavage in the presence of BoNT/A, and all native SNAP25s are cleaved. Cells were transfected and treated with BoNT/A as described above but scaled up into 6-well plates. After 72 h incubation with BoNT/A, cells were washed 3× with serum-free MEM, collected by scraping, and lysed using M-Per Lysis Buffer (Pierce™). 40 μg of cell lysate was subjected to SDS-PAGE before transfer to nitrocellulose paper and immunoblot analysis using an antibody directed against SNAP-25 (clone 71.2, Synaptic Systems). Arrows indicate the position of the full-length (closed) and cleaved (open) forms of the reporters. Full-length (*) and cleaved (**) native SNAP-25 are indicated.

(13) Viewed from another perspective, the inventive subject matter can be extended beyond cleavable substrates, to any assay having a construct with a reporter that can be de-protected, and then degraded in some manner by the cytosol or other local environment. For example, a susceptible reporter could be modified to include a ‘bait’ domain that is used to screen against a library of recombinant proteins that could possibly bind with the bait domain. Without the bait domain protected by a binding protein, the susceptible reporter will be degraded. In such an assay, cells expressing binding proteins will form a complex to protect the susceptible reporter from degradation, while cells expressing a binding partner to the bait will light up. The bait domain could advantageously be a small peptide, and the binding partners could be members of a library of proteins (or protein mutants). The system could also be reversed such that there is a library of bait domains tested against a single test protein (or test protein library).

(14) In each of these instances it is considered advantageous to include a second reporter that is not degraded post-exposure by the cytosol or other local environment, or is at least degraded much more slowly post-exposure than the first reporter.

(15) Still further, whereas the reporter can conveniently be selected from suitable fluorophores, it is contemplated that the reporter could be replaced or augmented by any other protein or other component with a defined function that is known to (a) have a relatively fast turnover in the cell without protection, and (b) that can be protected by interaction with a binding partner. Defined functions include transcription activators for reporter gene, repressors for lethal genes, etc. (anything that can be easily identified or selected against).

(16) FIG. 3 shows an embodiment in which data was collected for both YFP degradation and loss of FRET according to the state of the art practice from the exact same plates of cells. For YFP degradation, directly and singularly excited YFP emissions (top, Ex500, Em526) and CFP emissions (middle, Ex434, Em470) are collected. Those emission are then background subtracted and the YFP emission is divided by CFP emission to control for cell density and reporter expression in the individual wells. That emission ratio (YFP/CFP, bottom) is then used for the essay report.

(17) For loss of FRET, FRET emissions (top, Ex434, Em526) and CFP emissions (middle, Ex434, Em470) are collected. Those emissions are then background subtracted, and the FRET emission is divided by CFP emission to control for cell density and reporter expression in the individual wells. That emission ratio (FRET/CFP, bottom) is shown here to compare to the normal method.

(18) The key comparison is the loss of directly excited YFP versus the loss of FRET emission. From the comparison between the measurements and the corresponding curves, it becomes immediately apparent that the overall dynamic range for YFP degradation is much larger than the dynamic range of loss of FRET emissions. In some cases, there is no difference, statistically, between cells treated with no BoNT versus sells treated with saturating concentrations of BoNT when looking solely at the raw FRET emissions. For the loss of FRET method, the BoNT dose response only becomes clear after dividing the FRET emission by the CFP (donor) emission. The CFP (donor) emission shows a small increase emission due to de-quenching in response to reporter cleavage.

(19) In summary, the loss of FRET method reports BoNT-induced changes in the reporter very poorly, or not at all, and therefore cannot be therefore used for a correct qualitative and quantitative determination. In contrast, preferred methods contemplated herein have a high degree of specificity and reproducibility, which allow one to rely on the data for both the qualitative and quantitative analysis.

(20) Genetic Construction of Alternative Reporters

(21) Twenty alternate fluorophore constructs were generated in four plasmid backgrounds. Below is the internal name of each construct with a brief description of the background and cloned fragment. A more detailed summary of the construction methods follows.

(22) TABLE-US-00001 Alternative Construct Plasmid Reporter Name Background Description Mechanism pMD0076a pcDNA4/TO mRaspberry-SNAP-YFP 1 pMD0076b pcDNA4/TO mCherry-SNAP-YFP 1 pMD0077 pIRES SNAP-YFP, CFP 3 pMD0078 pECFP-C1 Synapsin promoter, 2 (Modified) SNAP-YFP pMD0079 pcDNA4/TO SNAP-YFP 2 pMD0080 pECFP-C1 SNAP-YFP 2 (Modified) pMD0081 pECFP-C1 mRaspberry-SNAP-YFP 1 (Modified) pMD0082 pECFP-C1 mCherry-SNAP-YFP 1 (Modified) pMD0090 pIRES SNAP-Venus, CFP 3 pMD0091 pcDNA4/TO SNAP-Venus 2 pMD0092 pECFP-C1 SNAP-Venus 2 (Modified) pMD0097 pcDNA4/TO mKate2-SNAP-YFP 1 pMD0098 pcDNA4/TO TagRFP-SNAP-YFP 1 pMD0099 pECFP-C1 mKate2-SNAP-YFP 1 (Modified) pMD0100 pECFP-C1 TagRFP-SNAP-YFP 1 (Modified) pMD0103 pBudCE4.1 SNAP-YFP, CFP 3 pMD0104 pBudCE4.1 SNAP-YFP, mRaspberry 3 pMD0105 pBudCE4.1 SNAP-YFP, mCherry 3 pMD0106 pBudCE4.1 SNAP-Venus, CFP 3 pMD0107 pBudCE4.1 SNAP-Venus, mRaspberry 3 pMD0108 pBudCE4.1 SNAP-Venus, mCherry 3
pMD0076a, pMD0076b, pMD0097, and pMD0098

(23) Constructs were generated by amplifying the mRaspberry, mCherry, mKate2, and TagRFP fluorophores with engineered KpnI and XhoI restriction sites. The amplified fragments and the previously used pcDNA4/TO BoCell vector generated were then digested with KpnI/XhoI. The vector DNA, minus the excised CFP, was then ligated with the mRaspberry, mCherry, mKate2, and TagRFP fragments to create the final vectors.

(24) pMD0079 and pMD0091

(25) For pMD0079, SNAP YFP was amplified with engineered BamHI and XhoI restriction sites. The amplified fragment and pcDNA4/TO vector DNA were then digested with BamHI/XhoI and then ligated together. pMD0091 was then generated by amplifying the Venus fluorophore with engineered EcoRI and XbaI restriction sites. The amplified Venus fragment and pMD0079 were then digested with EcoRI/XbaI. The pMD0079 vector DNA, minus the excised YFP, was then ligated with the Venus fragment to create pMD0091.

(26) pMD0077 and pMD0090

(27) For pMD0077, SNAP YFP was amplified with engineered NheI and XhoI restriction sites. The amplified fragment and ORES vector DNA were then digested with NheI/XhoI and then ligated together. The CFP fluorophore was then amplified with engineered XbaI/NotI restriction sites. The amplified fragment and previously generated SNAP YFP-containing pIRES vector were then digested with XbaI/NotI and ligated together to create pMD0077. pMD0090 was then generated by amplifying the Venus fluorophore with engineered EcoRI and MluI restriction sites. The amplified Venus fragment and pMD0077 were then digested with EcoRI/MluI. The pMD0077 vector DNA, minus the excised YFP, was then ligated with the Venus fragment to create pMD0090.

(28) pMD0078 and pMD0080

(29) For pMD0080, SNAP YFP was amplified with engineered NheI and XhoI restriction sites. Then amplified fragment and pECFP-C1 were then digested with NheI/XhoI. The vector DNA, minus the excised CFP, was then ligated with the SNAP YFP to create pMD0080. pMD0078 was then generated by amplifying the synapsin promoter with engineered AseI and NheI restriction sites. The amplified fragment and pMD0080 were then digested with AseI/NheI. The pMD0080 vector DNA, minus the excised CMV promoter, was then ligated with the synapsin promoter to create pMD0078.

(30) pMD0081, pMD0082, pMD0099, pMD0100

(31) Constructs were generated by amplifying the mRaspberry, mCherry, mKate2, and TagRFP fluorophores with engineered NheI and XhoI restriction sites. The amplified fragments and original BoCell construct from Min (pECFP-C1 background) were then digested with NheI/XhoI. The BoCell construct, minus the excised CFP fragment, was then ligated with the mRaspberry, mCherry, mKate2, and TagRFP fragments to create pMD0081, pMD0082, pMD0099, and pMD0100.

(32) pMD0092

(33) For pMD0092, the Venus fluorophore was amplified with engineered EcoRI and XbaI restriction sites. The amplified fragment and pMD0080 were then digested with EcoRI/XbaI. The pMD0080, minus the excised YFP fragment, was then ligated with the Venus fragment to generate pMD0092.

(34) pMD0103, pMD0104, and pMD0105

(35) The SNAP YFP construct was amplified with engineered XbaI and BamHI restriction sites. The amplified fragment and pBudCE4.1 vector were then digested with XbaI/BamHI and ligated together. The CFP, mRaspberry, and mCherry fluorophores were then amplified with engineered KpnI and BglII restriction sites. The amplified fragments and previously generated pBudCE4.1 vector containing SNAP YFP were then digested with KpnI/BglII and ligated together to generate pMD0103, pMD0104, and pMD0105.

(36) pMD0106, pMD0107, and pMD0108

(37) The SNAP Venus construct was amplified with engineered XbaI and BamHI restriction sites. The amplified fragment and pBudCE4.1 vector were then digested with XbaI/BamHI and ligated together. The CFP, mRaspberry, and mCherry fluorophores were then amplified with engineered KpnI and BglII restriction sites. The amplified fragments and previously generated pBudCE4.1 vector containing SNAP Venus were then digested with KpnI/BglII and ligated together to generate pMD0106, pMD0107, and pMD0108.

(38) Primary Screening of the Alternative Reporters

(39) Neuro2A cells were seeded into 96-well plates and allowed to expand 24-48 h before transiently transfecting the cells using the above genetic constructs and Lipofectamine 2000™ according to the manufacturer's instructions. Transfected cells were allowed to recover for 24 hr before applying 0 or 30 nM BoNT/A holotoxin and incubating the cells an additional 24 hr at 37° C., 5% CO.sub.2. Cells were then imaged using a Nikon-TE2000U fluorescence microscope taking a minimum of three images per condition. Fluorescence emissions were collected using filters appropriate for the listed fluorophores. Total fluorescence emissions were also collected using a Varioskan™ fluorescence microplate reader using appropriate excitation and emission wavelength settings.

(40) Fluorescence microscopy data was processed to gate out over expressing (saturated) cells based on pixel intensities for a given channel. Total emissions from each channel were then collected and, when indicated, the BoNT/A-responsive YFP or Venus emissions were divided by the BoNT/A-unresponsive CFP, RFP (mKate2, mRaspberry, or mCherry), or exogenously added membrane dye (Alternative Reporter 2) emissions.

(41) Each reporter construct, using the data collected above, was analyzed for the following: Cellular targeting of each reporter was judged by the presence of uniform fluorescence expression on the plasma membrane. Poor plasma membrane targeting was associated with the presence of bright, punctate spots within the cell. Reporters lacking plasma membrane targeting were eliminated from further consideration. Total fluorescence emissions and, thus total reporter expression, were judged by the emissions of a given fluorescent probe relative to background emissions. Probes that did not give a signal >2 times that of background were eliminated from further consideration.

(42) Secondary Screening of the Alternative Reporters: BoNT/A-Dose Responses

(43) Genetic constructs that passed primary screening were transiently transfected into cells as described above but using varying DNA concentrations. Varying the DNA concentration generated cells with varying levels of reporter expression. After transfection and a 24 hr recovery period, the transfected cells were titrated with 10 pM to 30 nM BoNT/A allowed to further incubate. After incubation, fluorescence emissions were collected using a Varioskan fluorescence microplate reader using appropriate excitation and emission wavelength settings. For all experiments, test reporter responses were directly compared to the current BoCell reporter CFP-SNAP25-YFP that was transiently transfected in parallel.

(44) Fluorescence emissions for the BoNT/A-response YFP were divided by the BoNT/A-unresponsive CFP, RFP (mKate2, mRaspberry, or mCherry), or exogenously added membrane dye (Alternative Reporter 2) emissions generating an emission ratio. The emission ratio was then plotted as a function of BoNT/A concentration. Data was compared to the BoCell reporter. Test reporter suitability was qualitatively assessed by comparison to the BoCell reporter: Does BoNT/A elicit a similar response with the test reporter compared to the current BoCell reporter?

(45) Secondary Screening of the Alternative Reporters: FRET Emissions

(46) Reporter genetic constructs were transfected into cells plated on glass cover slips using Lipofectamine according to the manufacturer's protocol. For each construct, single fluorophore controls were also transfected. After a 24 hr recovery period, cells were treated with or without 30 nM BoNT. Using the controls for each reporter, images were captured by fluorescence microscopy and the images used to calibrate FRET determinations by a three-filter set method. Reporters were then evaluated for FRET efficiency in the presence and absence of BoNT/A. For all experiments, test reporter responses were directly compared to the current BoCell reporter CFP-SNAP25-YFP that was transiently transfected in parallel. Each test reporter was evaluated for the presence of FRET and whether FRET emissions were responsive to BoNT/A. The conclusion was that FRET emissions do not represent a reliable screening method, in line with the conditions already observed in example 3.

(47) Stability Testing and Performance Relative to In Vivo Mouse Time-of-Death Assay

(48) As noted above, conventional in vitro mouse time-of-death assays for BoNT potency (as are typically used for stability studies in pharmaceutical preparations) have numerous disadvantages, including expense and increasing objections to such testing on ethical grounds. Surprisingly, Inventors have found that composition and methods of the inventive concept provide results that are essentially statistically equivalent (i.e. having a r.sup.2 value that exceeds 0.95) to those provided by in vivo studies approved for pharmaceutical industry use in BoNT characterization. In addition, Inventors have found that the reproducibility of cell-based assays of the inventive concept that are directed to BoNT toxin potency demonstrate greatly improved reproducibility relative to such in vivo mouse time-of-death assays. The CV of repeated measurements from a sample made using a cell-based assay of the inventive concept can be less than 50%, less than 40%, less than 30%, less than 25%, less than 20%, less than 15%, and/or less than 10% of a conventional in vivo mouse time-of-death assay that is approved for pharmaceutical use and performed on the same sample.

(49) An example of this is shown in FIG. 4. FIG. 4 shows the results of characterization of the potency of BoNT/A neurotoxin incubated at elevated temperature (40° C.) for different periods of time. The same samples were characterized using an in vitro cell-based assay of the inventive concept and by an in vivo mouse time-of-death study performed by a pharmaceutical company using established and regulatory agency-approved testing protocols. Results are shown as a percentage of the value determined for a control sample that was not exposed to elevated temperatures. The coefficients of variation are shown as error bars. In this example the correlation coefficient indicate an extremely high degree of correlation between the results of the two methods (r.sup.2=0.97). In addition it is evident that the cell-based method is more reproducible than an approved in vitro assay, with CVs that are typically less than 25% of that of an in vitro assay performed on the same sample.

(50) Such potency assays are particularly useful for characterizing stability. In such stability studies samples of a pharmaceutical preparation (for example, a preparation that includes BoNT/A) are stored at a temperature that is elevated relative to the conventional storage temperature. For example, for a preparation normally stored at 2° C. to 8° C. such stability testing can be performed by exposing samples of a pharmaceutical preparation to temperatures of 15° C., 20° C., 25° C., 30° C., 35° C., 37° C., 40° C., 45° C., 50° C. or higher for various periods of time. Following such treatment the activity or potency of the heated samples can be characterized and compared to that of the pharmaceutical preparation that has not been subjected to heat treatment. The loss of potency over time can be used to estimate a shelf life for the pharmaceutical preparation when stored at recommended temperatures, for example by comparison to a correlation between accelerated and real-time aging at the recommended temperature or through application of the Arrhenius equation. It should be appreciated that the higher degree of precision provided by cell-based assays of the inventive concept improve the accuracy of such estimations, with the high degree of correlation to the in vivo mouse lethality assay providing assurance that the observed loss of potency would also be observed in actual use.

(51) It should be apparent to those skilled in the art that many more modifications besides those already described are possible without departing from the inventive concepts herein. The inventive subject matter, therefore, is not to be restricted except in the scope of the appended claims. Moreover, in interpreting both the specification and the claims, all terms should be interpreted in the broadest possible manner consistent with the context. In particular, the terms “comprises” and “comprising” should be interpreted as referring to elements, components, or steps in a non-exclusive manner, indicating that the referenced elements, components, or steps may be present, or utilized, or combined with other elements, components, or steps that are not expressly referenced. Where the specification claims refers to at least one of something selected from the group consisting of A, B, C . . . and N, the text should be interpreted as requiring only one element from the group, not A plus N, or B plus N, etc.