Mol Imaging Biol (2016) DOI: 10.1007/s11307-015-0927-4 * World Molecular Imaging Society, 2016
RESEARCH ARTICLE
Detection of Klebsiella. Pneumoniae Infection with an Antisense Oligomer Against its Ribosomal RNA Ling Chen,1 Dengfeng Cheng,2 Guozheng Liu,1 Shuping Dou,1 Yuzhen Wang,1 Xinrong Liu,3 Yuxia Liu,4 Mary Rusckowski1 1
Division of Nuclear Medicine, Department of Radiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA, 01655, USA 2 Department of Nuclear Medicine, Zhongshan Hospital, Fudan University, Shanghai, 200032, China 3 ImmunoGen Inc, Waltham, MA, 02451, USA 4 Shanghai Institute of Applied Physics, Chinese Academy of Sciences, , Shanghai, 201800, China
Abstract Purpose: Previously, we demonstrated specific accumulation into bacteria of a 12-mer phosphorodiamidate morpholino (MORF) oligomer complementary to a ribosomal RNA (rRNA) segment found in all bacteria using the universal probe called Eub338 (Eub). Here, two MORF oligomers Eco and Kpn with sequences specific to the rRNA of Escherichia coli (Eco) and Klebsiella pneumoniae (Kpn) were investigated along with Eub and control (nonEub). Procedures: To determine bacterial rRNA binding, oligomers were tagged with Alexa Fluor 633 (AF633) for fluorescence in situ hybridization (FISH) and fluorescence microscopy, and radiolabeled with technetium-99m (Tc-99m) for biodistribution and SPECT imaging in infected mice. Results: By both FISH and fluorescence microscopy, Eub showed a positive signal in both E. coli and K. pneumoniae as expected, and Kpn showed significantly higher accumulation in K. pneumoniae with near background in E. coli (p G 0.01). Conversely, Eco was positive in both E. coli and K. pneumoniae, hence nonspecific. As determined by biodistribution, the accumulation of [99mTc]Kpn was higher in the thigh infected with live K. pneumoniae than with live E. coli (p = 0.05), and significantly higher than with heat-killed K. pneumoniae (p = 0.02) in the target thigh. By SPECT imaging, the accumulation of [99mTc]Kpn was obviously higher in its specific target of K. pneumoniae compared to an E. coli infected thigh. Conclusions: Kpn complementary to the rRNA of K. pneumoniae, labeled with Tc-99m or AF633, demonstrated specific binding to fixed and live K. pneumoniae in culture and in infected mice such that Tc-99m-labeled Kpn as the MORF oligomer may be useful for K. pneumoniae infection detection through imaging. Key words: Antisense MORF, Bacterial ribosomal RNA, Bacterial infection imaging
Introduction The emergence of multidrug-resistant bacterial strains and an increase in the immunosuppressed population have increased the Correspondence to: Mary Rusckowski; e-mail: [email protected]
need for a diagnostic agent that can detect by external imaging the presence of a specific bacterial infection at an early stage. To meet, this need oligomer-based imaging probes with specificity to the ribosomal RNA (rRNA) are being developed to locate and assess the presence and extent of bacterial infection. The oligomers are designed to be probes designed to be specific to a species or genus
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
of bacteria. They will be of value to monitor the effectiveness of a patient’s treatment 1) when the infective agent is known, 2) to identify nonresponders at an early time to redirect treatment or 3) perhaps validate the endpoint of treatment. Beyond clinical applications, bacterial specific oligomer probes could be useful in the development of new antibiotics. This is an area where infection establishment and treatment response in animal models such as nonhuman primates are difficult to evaluate, monitor, and quantify. These oligomer-based probes could fill that need. The bacterial rRNA has been a common target for complementary probes in in vitro identification of bacteria [1–5]. Some regions of the rRNA have remained unchanged in most if not all sequenced bacterial species, and these sites could serve as targets for broad-spectrum bacterial probes while other regions that have mutated could serve as sites for species-specific probes. Considered herein are phosphorodiamidate morpholino (MORF) oligomers. These are synthetic DNA analogs that differ from DNA in the linkage between the bases. MORF oligomers are nonionic, stable to nucleases, and maintain proper structure for complementary base pairing [6]. Additionally, MORFs do not bind to serum proteins; they can pass through the cell membrane and are cleared rapidly from circulation [6, 7]. For these reasons, MORF oligomers have been used successfully by this laboratory for the detection of tumors by antisense and pretargeting approaches [7, 8] and for bacterial and fungal infections [9, 10]. An oligomer sequence identified elsewhere [1] and designated Eub338 (Eub) was used by us previously [9]. Eub338 has been shown by others to be complementary to a conserved region of the 16S rRNA found in most if not all bacteria. Previously, using the Eub sequence, we identified the MORF backbone as producing oligomers that maintain high binding and show the greatest accessibility to target RNA in Grampositive and Gram-negative bacteria [9]. We demonstrated accumulation and binding of both fluorophore and technetium99m (Tc-99m)-labeled Eub MORF oligomer to the bacterial RNA in vitro. We also evaluated the biodistribution and targeting of the Tc-99m-labeled Eub MORF in mice with a K pneumoniae infection in one thigh. In this study, we investigated species-specific rRNA probes designated as Eco and Kpn radiolabeled with Tc-99m for accumulation and binding to their specific bacterial strains, E coli and K. pneumoniae, respectively. The specific sequences for E. coli and K. pneumoniae were from the literature [2, 11] that demonstrated specific binding by either in situ hybridization (ISH) or fluorescence in situ hybridization (FISH) methods. This study evaluated these radiolabeled MORF oligomers specific to bacterial strains as agents to detect infection through imaging.
American Type Culture Collection (Rockville, MD) and grown in Nutrient broth. Both were grown at 37 °C. The Alexa Fluor 633 carboxylic acid succinimidyl ester (AF633) and the lipophilic membrane dye FM 1–43 were from Invitrogen (Eugene, OR). The [99mTc]pertechnetate was eluted from a Mo-99/Tc99m generator (Perkin-Elmer, Boston, MA). The S-acetyl NHSmercaptoacetyltriglycine (MAG3) was synthesized in house [12]. The HPLC system was equipped with a 515 pump, an in-line dual UV detector, and an in-line gamma-radioactivity detector under the control of Millennium 32 software (Waters, Milford, MA).
MORF Conjugation The 12-mer MORF sequences used in this study are shown in Table 1. Eub is a universal probe specific for a common sequence in the 16S rRNA present in most if not all bacteria; it served as a positive control [1]. The nonEub is a random sequence and used as a negative control. Eco and Kpn MORFs are species-specific bacterial probes. Eco is complementary to the 16S rRNA of E. coli [11], and Kpn is complementary to the 23S rRNA of K. pneumoniae [2]. The MORF sequences were synthesized by Gene Tools (Philomath, OR) with a primary amine attached via a 6-carbon linker on the 3′ equivalent end for conjugation to the MAG3 chelator for radiolabeling with Tc-99m or to the fluorophore AF633 for fluorescence microscopy. The MORF oligomers were conjugated with MAG3 using methods standard in this laboratory [12]. In brief, a solution of 300 μg oligomer in 200 μl of 0.3 M HEPES buffer, pH 8, was added to 0.7–1.0 mg of MAG3 powder and the suspension immediately mixed on a vortex to form a clear solution. The sample was left undisturbed for 1 h at room temperature before the addition of 50 μl of 1 M ammonium acetate and 120 μl of freshly prepared 20 mg/ml stannous chloride (SnCl2-2H2O) in tartrate buffer (100 mg/ml sodium tartrate in 0.5 M ammonium bicarbonate, 0.25 M ammonium acetate, and 0.18 M ammonium hydroxide, pH 9.2). The sample was mixed and heated to 95 °C for 20 min, then cooled to room temperature. Absolute ethanol was added for a final concentration of 20 % (v/v) and the sample was purified on a 1 × 20 cm Biogel P2 size exclusion column (Bio-Rad, Hercules, CA) using 0.25 M ammonium acetate pH 7 as eluant. The MORF concentration was determined by OD at 265 nm and stored at −80 °C for further use. For fluorescence microscopy, the MORFs were conjugated with the fluorophore AF633. Briefly, MORF (200 μg) in 0.1 M sodium bicarbonate buffer pH 8.4 was mixed with AF633 (10 mg/ml in N-methyl-2-pyrrolidone, Sigma-Aldrich, St. Louis, MO) with an AF633 to a MORF molar ratio of 10:1. After 45 min in the dark, the mixture was purified on a 1 × 20 cm P2 column with eluant 0.25 M ammonium acetate buffer pH 7. The MORF concentration was determined at 265 nm and stored in the dark at −20 °C.
MORF Radiolabeling The MORF oligomers were radiolabeled with Tc-99m using methods standard in this laboratory [12]. In brief, the MAG3-
Materials and Methods
Table 1. MORF oligomer sequences and specificity
Bacterial Cultures and Materials
Name
12-mer sequence (5′ → 3′)
Specificity
Eub NonEub Eco Kpn
GCT GCC TCC CGT AGG GCA TCC TCA GCA AAG GTA TTA CAC CAG CGT GCC
Universal Random control E. coli K. pneumoniae
E. coli strain K12 was purchased from the E. coli Genetic Stock Center (Yale University, New Haven, CT) and grown in LuriaBertani (LB) medium. K. pneumoniae was purchased from the
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
conjugated MORF (about 1 μg in 4 μl) was added to a solution of 45 μl 0.25 M ammonium acetate and 15 μl 50 mg/ml tartrate solution followed by 2 μl of freshly prepared 10 mg/ml SnCl22H2O solution in 10 mM HCl with 1 mg/ml ascorbate. After mixing on a vortex, Tc-99m (2–5 μl with 200–500 μCi) was added with agitation, followed by heating at 95 °C for 20 min. Radiochemical purity was determined by size exclusion HPLC Superose 12 column (Amersham Pharmacia Biotech, Piscataway, NJ). The running solution was 20 % acetonitrile in 0.1 M Tris–HCl pH 8 at a flow rate of 0.6 ml/min. Radioactivity recovery was always greater than 95 % and radiochemical purity greater than 90 %. Specific activity was about 200 μCi/μg.
MORF Fluorescence In Situ Hybridization In preparation for FISH, K. pneumoniae and E. coli log-phase cultures were diluted 1:3 (v/v) with 4 % formaldehyde prepared in Dulbecco’s phosphate-buffered saline (D-PBS). The sample was mixed on a vortex and left at room temperature for at least 3 h, then spun at 12,000 rpm for 2 min at 4 °C. The sample was washed with D-PBS to remove residual formaldehyde, spun again, and the pellet resuspended in D-PBS with 108 to 109 cells per milliliter. Finally, the fixed cell suspension was mixed with an equal volume of cold absolute ethanol and stored at −20 °C overnight. For hybridization, 3 μl of the fixed bacterial cell suspension was deposited onto an eight-chambered glass slide (Lab-Tek, Rochester, NY) and air dried. The AF633-conjugated MORFs were added at 5 ng/μl in 150-μl buffer containing 750 mM NaCl, 100 mM Tris-Cl pH 7.8, 5 mM EDTA, 0.2 % bovine serum albumin, 0.01 % polyadenylic acid, 0.1 % sodium dodecyl sulfate (all from SigmaAldrich), and 10 % dextran sulfate (dextran MW 500 kD, Calbiochem, Gibbstown, NJ), as described by Ouverney [3], and incubated at 43 °C for 2 h. The chambers were washed with distilled water at 43 °C, and then washed with buffer containing 30 mM NaCl, 4 mM Tris-Cl pH 7.8, and 0.2 mM EDTA for 30 min at 43 °C with two changes of wash solution. The cell membrane dye FM1-43 (0.2 μl at 5 μg/μl) was added about 10 min before viewing the cells under oil immersion with ×100 objective on an Olympus IX-70 inverted fluorescence microscope (Olympus America, Inc., Center Valley, PA). The quantification of fluorescence in cells was measured by ImageJ software (NIH). At least five areas were drawn about the cells of interest and background, and the corrected total cell fluorescence was calculated.
Accumulation of Fluorescent and Radiolabeled MORFs in Live Bacteria An overnight culture of K. pneumoniae and E. coli were diluted 1:50 with their respective media, and 200 μl of the diluted cultures were mixed with 15 μl of AF633-MORF at 15 ng/μl. The samples were incubated for 2 h at 37 °C with rocking in the dark then washed with 0.85 % NaCl and resuspended in 200 μl 0.85 % NaCl. About 3 μl was placed into a single chamber of an eight-chambered glass slide followed by 0.2 μl (5 μg/μl) of FM1-43 membrane stain. The samples were then air dried and mounted with fluorescence mounting medium (Dako, Carpinteria, CA) and viewed with an oil immersion ×100 objective. The incorporation and binding of the Tc-99m-labeled MORFs to RNA were also evaluated in live cells. Overnight cultures of K.
pneumoniae and E. coli were diluted 1:50 with media and 5 ml were mixed with 50 nmol of Tc-99m-labeled Eub, Kpn, or Eco MORFs and incubated at 37 °C on a lab rocker for 2 h. Thereafter, the samples were spun and washed three times with 0.85 % NaCl. Total RNA was isolated using TRIzol® Max™ bacterial RNA isolation kit from Invitrogen (Eugene, OR) following the manufacturer’s instructions. The RNA concentration was determined by OD at 260 nm using 25 μl/ μg/cm extinction coefficient. The RNA fraction was carefully transferred to fresh tubes and measured for radioactivity in a gamma well counter, and results were reported as nanomoles bound per 107 cells. To determine the number of bacteria in the incubation mixture, 100 μl of the sample was serially diluted and each dilution was grown overnight on LB agar plates. The bacterial cell count was determined from the colony number on each plate and dilution factor.
Animal Studies The studies in mice were with the approval of the Institutional Animal Care and Use Committee. The biodistribution of the Tc99m-labeled Kpn was measured in CD-1 mice (Charles River Laboratories International, Inc., Wilmington, MA) with live or heat-killed K. pneumoniae or E. coli in one thigh. An overnight culture of K. pneumoniae or E. coli was diluted with culture medium to an OD of 0.6 at 600 nm (about 5 × 108 cells/ml). Half of the culture was used as the live preparation while the remaining half was heated in a boiling water bath for 30 min to kill the bacteria and to provide a sample for injection of bacterial debris and possibly intact rRNA [13]. Either the live or heat-killed preparation of K. pneumoniae or E. coli was injected subcutaneously (0.1 ml with about 5 × 107 cells) into the one thigh of mice (n = 4). After 2 h, about 1 μg of the [99mTc]Kpn in 0.1 ml saline was injected through a tail vein. Animals were euthanized at 60 min and the organs of interest and blood were removed, weighed, and counted in a gamma well counter. A similar set of mice (n = 3) were prepared as above for imaging on a NanoSPECT/CT (Bioscan, Washington, DC) small animal camera. Animals received live K. pneumoniae or E. coli subcutaneously in the right thigh and after 2 h [99mTc]Kpn, about 3 μg with 600 μCi, was delivered through a tail vein in 0.1 ml saline. Immediately after injection, the mice were anesthetized with 1–2 % isoflurane carried in oxygen, and whole-body scans were taken at 30, 60, and 90 min with 24 projections and 60 s per projection. The regions of interest (ROI) were drawn about the infected and noninfected thighs for the quantification of accumulated radioactivity for analysis using VivoQuant 2.1 software (inviCRO, Boston, MA).
Statistical Analysis The Student t test was used to test for significance where indicated.
Results MORF Fluorescence In Situ Hybridization The binding of AF633-MORFs to RNA of K. pneumoniae or E. coli was evaluated in fixed cells. Fig. 1 presents images by
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 1 Fluorescence in situ hybridization images of AF633-conjugated MORFs Eub, nonEub, Eco, and Kpn with fixed bacteria a E. coli and b K. pneumoniae. Top row (green) shows cells stained with the membrane stain FM1-43, and bottom row (red) shows cells with AF633-conjugated MORFs Eub, nonEub, Kpn, and Eco. (Magnification ×1000). c Quantification of cell fluorescence calculated for panels a E. coli and b K. pneumoniae (*p G 0.01).
FISH showing the binding of Eub, nonEub, Eco, and Kpn to E. coli and K. pneumoniae ( bottom row, panels a and b, respectively). The bacterial membrane stained with FM 1–43 is shown across the top row. Fig. 1c shows the the corrected quantification of total AF633-MORF cell fluorescence for Fig. 1a and b. A signal for Eub, the universal probe, is evident in both E. coli and K. pneumoniae indicating hybridization of the Eub sequence to RNA of both bacteria, while lower staining is evident for the nonEub (nonspecific probe) in both bacteria. Eco provides a positive signal in its specific bacteria E. coli, but also in nontarget K. pneumoniae. On the other hand, Kpn
shows near background signal to the nontarget bacteria E. coli, but a positive signal, significantly higher (* p G 0.01), to its specific bacteria K. pneumoniae, indicating specific binding.
Accumulation of Fluorescent and Radiolabeled MORFs in Live Bacteria Fluorescence microscopy was used to evaluate the accumulation of AF633-MORFs in live E. coli (Fig. 2a) and K. pneumoniae (Fig. 2b). Fig. 2c shows the corrected
Fig. 2 Fluorescence microscopy of AF633-conjugated Eub, Kpn, and Eco MORFs to live a E. coli and b K. pneumoniae. Top row (green) shows the cell membrane stained with FM1-43, bottom row (red) shows accumulations of AF633-conjugated MORFs in cells. (Magnification ×1000). c Quantification of cell fluorescence calculated for panels a E. coli and b K. pneumoniae (*p G 0.01).
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
quantification of total AF633-MORF cell fluorescence for Fig. 2a and b. Accumulation for the universal Eub is observed in both E. coli and K. pneumoniae. In comparison, Kpn shows significantly higher accumulation in its specific bacteria K. pneumoniae and only background signal in nontarget E. coli (*p G 0.01), indicating specificity for Kpn. In contrast, Eco shows only minimal accumulation in its specific bacteria E. coli, which is similar to its accumulation observed in K. pneumoniae. That the Eco shows binding to the nonspecific target and binding that is at a level similar to that obtained for the other probes suggests loss of RNA binding specificity. To further establish binding to bacterial rRNA, the Tc99m-labeled Eub, Kpn, and Eco MORFs were incubated with E. coli or K. pneumoniae and the RNA was isolated and counted. As shown in Fig. 3, Kpn shows a higher percentage of radioactivity bound to RNA from its specific bacteria K. pneumoniae compared to Eco (*p = 0.04) and Eub for the same target. Not surprising is that the difference was not significant in the latter case, which is expected as Eub is a universal probe. In addition, Kpn shows significantly higher binding to its specific bacteria K. pneumoniae compared to nontarget E. coli (**p = 0.03), strongly suggesting the specificity of Kpn MORF. On the other hand, Eco shows a higher percentage of radioactivity bound to RNA from its specific bacteria E. coli in comparison to Kpn and Eub, but then similar binding to RNA from both E. coli and K. pneumoniae, indicating lack of specificity.
Animal Studies The biodistribution of [99mTc]Kpn (1 μg, 200 μCi) was measured at 60 min in mice with either live or heat-killed K. pneumoniae or E. coli in one thigh (Fig. 4). The time of 60 min post administration was based on the experience of
this laboratory with other radiolabeled MORFs for pretargeting applications. It is known that similar oligomers clear from circulation within 30 min [7]. The delay to 60 min was to minimize background. Fig. 4 shows the percent injected dose per gram of [99mTc]Kpn in tissues for the four study groups. Each organ shows similar accumulation of probes regardless of study group. The kidneys show the highest accumulation and are presented on a separate scale in the inset. High accumulation was also found in the small intestine. The [99mTc]Kpn shows the highest accumulation in the target thigh with live K. pneumoniae, which is significantly higher than for mice with the nontarget live E. coli (*p = 0.05) and significantly higher than for mice with heat-killed K. pneumoniae (**p = 0.02). However, when comparing both heat-killed bacterial preparations, [99mTc]Kpn accumulation in the thigh with K. pneumoniae was not significantly different from the thigh with E. coli (#p = 0.16). Fig. 5 shows the SPECT/CT images of [99mTc]Kpn (3 μg, 600 μCi) obtained at 30, 60, and 90 min in mice with K. pneumoniae (panels a, b, c) or E. coli (panels d, e, f) in one thigh (left in images). As in the biodistribution study, high accumulations are observed in the kidneys and small intestines. The bladder is the primary route of excretion and is evident in all images. [99mTc]Kpn shows obvious accumulation in the thigh with its target bacteria K. pneumoniae (panels a, b, c) that does not decrease over the course of 90 min. In comparison, the animals with nontarget E. coli in the thigh (panels d, e, f) show much lower infected thigh accumulations at all times and a gradual decrease (washout) through 90 min. The corresponding radioactivities in the target thighs and normal thighs from the ROI of a similar area are shown in Fig. 6a. The resultant target to nontarget thigh ratios (T:NT), shown in Fig. 6b, for K. pneumoniae, increase gradually reaching 5.7 at 90 min, whereas with E. coli as target, the ratios are lower at all time points and decrease at the latest time. A study to block
Fig. 3 Evaluation of the incorporation and binding of Tc-99m-labeled MORFs into live bacteria. Tc-99m-labeled MORFs Eub, Eco, and Kpn were incubated with cultures of E. coli and K. pneumoniae, and the percent of radioactivity bound to RNA per 107 cells is shown. Mean of n = 3 with SD (*p = 0.04, **p = 0.03).
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 4 Biodistribution 60 min post administration of [99mTc]Kpn MORF (1 μg, 200 μCi) to mice (n = 4) receiving either live or heat-killed K. pneumoniae or E. coli in one thigh 2 h earlier (*p = 0.05, **p = 0.02, # not significant).
specific binding by delivering excess unlabeled material was not performed. It was reasoned that since bacteria grow at an exponential rate, together with the associated increase in target RNA, it would be too unpredictable to determine the mass of material for blocking.
Discussion The bacterial rRNA has regions that have remained unchanged over time while other regions have undergone random mutations leading to bacterial diversity. These conserved regions have been used as targets for oligomer probes in the identification of bacterial pathogens by in vitro methods such as ISH and FISH [4, 5]. The Eub sequence is one such probe and has been demonstrated by others to bind to a conserved region of the bacterial rRNA and thus useful to detect bacterial contamination [1]. However, it cannot distinguish between bacterial species [1]. Previously, we demonstrated that the Tc99m-labeled Eub probe may be useful to detect bacterial infection in vivo [9]. Alternatively, the rRNA variant region is an ideal site for species-specific probes. The objective of this study was to evaluate Tc-99m-labeled species-specific probes with targets in the variant region by in vitro methods and in mouse infection models. Probes against E. coli and K. pneumoniae were chosen for the study since both are the source of common, clinically relevant infections each with increasing multidrug resistance strains [14, 15]. The MORF sequences from which Eub, nonEub, Eco, and Kpn were derived were reported previously [1, 2, 11] for in vitro bacterial identification. Based on findings by Deere et al. [16], the optimum length for oligomers to cross the bacterial cell wall was determined to be 9 to 12-mer. Therefore, we reduced the sequences to 12-mer by removing 6 bases from the 3′ equivalent end for Eub and Eco and from 5′ equivalent end for nonEub and Kpn. The results in fixed bacteria by FISH presented in Fig. 1 demonstrate that Kpn shows specific hybridization to the RNA of K. pneumoniae. However, this was not the case for Eco as a positive signal was obtained in both E. coli and K. pneumoniae.
Fig. 5 SPECT/CT images after administration of [99mTc]Kpn MORF (3 μg, 600 μCi) in mice (n = 3) with either K. pneumoniae in the left thigh obtained at a 30, b 60, or c 90 min, or E. coli in the left thigh obtained at d 30, e 60, or f 90 min.
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 6 a Radioactivity (μCi) in the target thighs and normal thighs. b Target/nontarget (T:NT) ratios when K. pneumoniae or E. coli infected thigh is the target organism at 30, 60, and 90 min.
The accumulation of MORFs into live bacteria was confirmed by fluorescence microscopy. Fig. 2 shows the positive accumulation of the Eub MORF in both live K. pneumoniae and E. coli. Specific accumulation was observed for Kpn in K. pneumoniae. However, for Eco, the fluorescent signal in its target bacteria E. coli was equivalent to that observed with Kpn and lower than with Eub. Also, Eco is bound to nontarget K. pneumoniae to a level greater than the Eub. Both observations in Figs. 1 and 2 suggest the lack of specificity for this probe. The absence of specificity is not likely a problem associated with membrane penetration. First of all, Deere J et al. [16] showed that MORFs 9 to 12 bases in length were effective in inhibiting gene expression in live bacterial cultures. Thus, the oligomers were able to pass through the membrane. Secondly, in our previous study [9], the 12-mer Eub showed accumulation in both E. coli K12 and a membrane permeable E. coli SM101, indicating that the 12-mer Eub oligomer has no problem penetrating the cell membrane. The results obtained with the radiolabeled MORFs in live E. coli or K. pneumoniae, presented in Fig. 3, are similar. [99mTc]Kpn shows specific binding only to the RNA from K. pneumoniae while [99mTc]Eco shows binding to RNA from both E. coli and K. pneumoniae. Again, these negative results with Eco suggest the lack of specificity. As stated before, the loss of specificity may be due to shortening the sequence or perhaps the state of bacterial growth, or possibly hindrance due to the secondary structure of the RNA complementary site. Since Kpn shows specific accumulation in live cells with the both fluorescent tag and radiolabel, it was studied in animal models. The biodistribution at 60 min post administration of the [99mTc]Kpn in mice with either live or heatkilled K. pneumoniae or E. coli, presented in Fig. 4, shows the significantly higher accumulation in the thigh of mice receiving live K. pneumoniae, in comparison to mice with nontarget E. coli (*p = 0.05), and significantly higher than mice receiving heat-killed K. pneumoniae (**p = 0.02). It has been reported that the portions of the bacterial rRNA can remain stable even at 100 °C [13, 17, 18]. Therefore, the
higher accumulation of Kpn in the thigh with either heatkilled K. pneumoniae or E. coli compared to the normal thigh may be in part the reason for this observation. Similar to that seen by necropsy, imaging shows obvious accumulation of [99mTc]Kpn in the thigh infected with its specific bacteria K. pneumoniae compared to the thigh with nontarget E. coli. The target thigh to nontarget thigh ratios for K. pneumoniae-infected mice increase over time reaching 5.7 at 90 min, whereas for [99mTc]Kpn in E. coli-infected mice, the ratios remained lower and decreased to 2.8 at 90 min, suggesting washout of the probe with non-target bacteria.
Conclusions The [99mTc]Kpn MORF oligomer with a sequence complementary to the variant region of the K. pneumoniae rRNA demonstrated specific accumulation in live K. pneumoniae in vitro and in vivo. The [99mTc]Kpn MORF showed obvious accumulation in its specific infection target in mice and may be useful for K. pneumoniae infection detection through imaging. Acknowledgments. Funding was provided by a grant from NIH no. AI070857. Compliance with Ethical Standards Conflict of Interest The authors declare that they have no conflict of interest. Ethics Approval The studies in mice were with the approval of the Institutional Animal Care and Use Committee.
References 1. Amann RI, Krumholz L, Stahl DA (1990) Fluorescent-oligonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology. J Bacteriol 172(2):762–770 2. Kempf VA, Trebesius K, Autenrieth IB (2000) Fluorescent in situ hybridization allows rapid identification of microorganisms in blood cultures. J Clin Microbiol 38(2):830–838 3. Ouverney CC, Fuhrman JA (1997) Increase in fluorescence intensity of 16S rRNA in situ hybridization in natural samples treated with chloramphenicol. Appl Environ Microbiol 63(7):2735–2740
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
4. Moter A, Gobel UB (2000) Fluorescence in situ hybridization (FISH) for direct visualization of microorganisms. J Microbio Meth 41:85–112 5. Giovannoni SJ, DeLong EF, Olsen GJ et al (1988) Phylogenetic groupspecific oligodeoxynucleotide probes for identification of single microbial cells. Bacteriol 170(2):720–726 6. Summerton JE (2007) Morpholino, siRNA, and S-DNA compared: impact of structure and mechanism of action on off-target effects and sequence specificity. Curr Top Med Chem 7:651–660 7. Liu G, Mang’era K, Liu N et al (2002) Tumor pretargeting in mice using (99m)Tc-labeled morpholino, a DNA analog. J Nucl Med 43(3):384–391 8. Liu G, He J, Dou S et al (2004) Pretargeting in tumored mice with radiolabeled morpholino oligomer showing low kidney uptake. Eur J Nucl Med Mol Imaging 31(3):417–424 9. Chen L, Wang Y, Cheng D et al (2013) 99mTc-MORF oligomers specific for bacterial ribosomal RNA as potential specific infection imaging agents. Bioorg Med Chem 21(21):6523–6530 10. Wang Y, Chen L, Liu X et al (2013) Detection of Aspergillus fumigatus pulmonary fungal infections in mice with (99m)Tc-labeled MORF oligomers targeting ribosomal RNA. Nucl Med Biol 40(1):89–96 11. Geller BL, Deere JD, Stein DA et al (2003) Inhibition of gene expression in Escherichia coli by antisense phosphorodiamidate morpholino oligomers. Antimicrob Agents Chemother 47(10):3233–3239
12. Wang Y, Liu G, Hnatowich DJ (2006) Methods for MAG3 conjugation and 99mTc radiolabeling of biomolecules. Nat Protocols 1:1477–1480 13. Sheridan GEC, Masters CI, Shallcross JA et al (1998) Detection of mRNA by reverse transcription-PCR as an indicator of viability in Escherichia coli cells. Appl Environ Microb 64(4):1313–1318 14. Kallen AJ, Hidron AI, Patel J et al (2010) Multidrug resistance among Gram-negative pathogens that caused healthcare-associated infections reported to the national healthcare safety network, 2006–2008. Infect Control Hosp Epidemiol 31(5):528–531 15. Sahm DF, Thornsberry C, Mayfield DC et al (2001) Multidrug-resistant urinary tract isolates of Escherichia coli: prevalence and patient demographics in the United States in 2000. Antimicrob Agents Chemother 45(5):1402–1406 16. Deere J, Iversen P, Geller BL (2005) Antisense phosphorodiamidate morpholino oligomer length and target position effects on gene-specific inhibition in Escherichia coli. Antimicrob Agents Chemother 49:249–255 17. Uyttendaele M, Bastiaansen A, Debevere J (1997) Evaluation of the NASBA nucleic acid amplification system for assessment of the viability of Campylobacter jejuni. Int J Food Microbiol 37:13–20 18. Aellen S, Que YA, Guignard B et al (2006) Detection of live and antibiotic-killed bacteria by quantitative real-time PCR of specific fragments of rRNA. Antimicrob Agents Chemother 50(6):1913–1920
RESEARCH ARTICLE
Detection of Klebsiella. Pneumoniae Infection with an Antisense Oligomer Against its Ribosomal RNA Ling Chen,1 Dengfeng Cheng,2 Guozheng Liu,1 Shuping Dou,1 Yuzhen Wang,1 Xinrong Liu,3 Yuxia Liu,4 Mary Rusckowski1 1
Division of Nuclear Medicine, Department of Radiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA, 01655, USA 2 Department of Nuclear Medicine, Zhongshan Hospital, Fudan University, Shanghai, 200032, China 3 ImmunoGen Inc, Waltham, MA, 02451, USA 4 Shanghai Institute of Applied Physics, Chinese Academy of Sciences, , Shanghai, 201800, China
Abstract Purpose: Previously, we demonstrated specific accumulation into bacteria of a 12-mer phosphorodiamidate morpholino (MORF) oligomer complementary to a ribosomal RNA (rRNA) segment found in all bacteria using the universal probe called Eub338 (Eub). Here, two MORF oligomers Eco and Kpn with sequences specific to the rRNA of Escherichia coli (Eco) and Klebsiella pneumoniae (Kpn) were investigated along with Eub and control (nonEub). Procedures: To determine bacterial rRNA binding, oligomers were tagged with Alexa Fluor 633 (AF633) for fluorescence in situ hybridization (FISH) and fluorescence microscopy, and radiolabeled with technetium-99m (Tc-99m) for biodistribution and SPECT imaging in infected mice. Results: By both FISH and fluorescence microscopy, Eub showed a positive signal in both E. coli and K. pneumoniae as expected, and Kpn showed significantly higher accumulation in K. pneumoniae with near background in E. coli (p G 0.01). Conversely, Eco was positive in both E. coli and K. pneumoniae, hence nonspecific. As determined by biodistribution, the accumulation of [99mTc]Kpn was higher in the thigh infected with live K. pneumoniae than with live E. coli (p = 0.05), and significantly higher than with heat-killed K. pneumoniae (p = 0.02) in the target thigh. By SPECT imaging, the accumulation of [99mTc]Kpn was obviously higher in its specific target of K. pneumoniae compared to an E. coli infected thigh. Conclusions: Kpn complementary to the rRNA of K. pneumoniae, labeled with Tc-99m or AF633, demonstrated specific binding to fixed and live K. pneumoniae in culture and in infected mice such that Tc-99m-labeled Kpn as the MORF oligomer may be useful for K. pneumoniae infection detection through imaging. Key words: Antisense MORF, Bacterial ribosomal RNA, Bacterial infection imaging
Introduction The emergence of multidrug-resistant bacterial strains and an increase in the immunosuppressed population have increased the Correspondence to: Mary Rusckowski; e-mail: [email protected]
need for a diagnostic agent that can detect by external imaging the presence of a specific bacterial infection at an early stage. To meet, this need oligomer-based imaging probes with specificity to the ribosomal RNA (rRNA) are being developed to locate and assess the presence and extent of bacterial infection. The oligomers are designed to be probes designed to be specific to a species or genus
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
of bacteria. They will be of value to monitor the effectiveness of a patient’s treatment 1) when the infective agent is known, 2) to identify nonresponders at an early time to redirect treatment or 3) perhaps validate the endpoint of treatment. Beyond clinical applications, bacterial specific oligomer probes could be useful in the development of new antibiotics. This is an area where infection establishment and treatment response in animal models such as nonhuman primates are difficult to evaluate, monitor, and quantify. These oligomer-based probes could fill that need. The bacterial rRNA has been a common target for complementary probes in in vitro identification of bacteria [1–5]. Some regions of the rRNA have remained unchanged in most if not all sequenced bacterial species, and these sites could serve as targets for broad-spectrum bacterial probes while other regions that have mutated could serve as sites for species-specific probes. Considered herein are phosphorodiamidate morpholino (MORF) oligomers. These are synthetic DNA analogs that differ from DNA in the linkage between the bases. MORF oligomers are nonionic, stable to nucleases, and maintain proper structure for complementary base pairing [6]. Additionally, MORFs do not bind to serum proteins; they can pass through the cell membrane and are cleared rapidly from circulation [6, 7]. For these reasons, MORF oligomers have been used successfully by this laboratory for the detection of tumors by antisense and pretargeting approaches [7, 8] and for bacterial and fungal infections [9, 10]. An oligomer sequence identified elsewhere [1] and designated Eub338 (Eub) was used by us previously [9]. Eub338 has been shown by others to be complementary to a conserved region of the 16S rRNA found in most if not all bacteria. Previously, using the Eub sequence, we identified the MORF backbone as producing oligomers that maintain high binding and show the greatest accessibility to target RNA in Grampositive and Gram-negative bacteria [9]. We demonstrated accumulation and binding of both fluorophore and technetium99m (Tc-99m)-labeled Eub MORF oligomer to the bacterial RNA in vitro. We also evaluated the biodistribution and targeting of the Tc-99m-labeled Eub MORF in mice with a K pneumoniae infection in one thigh. In this study, we investigated species-specific rRNA probes designated as Eco and Kpn radiolabeled with Tc-99m for accumulation and binding to their specific bacterial strains, E coli and K. pneumoniae, respectively. The specific sequences for E. coli and K. pneumoniae were from the literature [2, 11] that demonstrated specific binding by either in situ hybridization (ISH) or fluorescence in situ hybridization (FISH) methods. This study evaluated these radiolabeled MORF oligomers specific to bacterial strains as agents to detect infection through imaging.
American Type Culture Collection (Rockville, MD) and grown in Nutrient broth. Both were grown at 37 °C. The Alexa Fluor 633 carboxylic acid succinimidyl ester (AF633) and the lipophilic membrane dye FM 1–43 were from Invitrogen (Eugene, OR). The [99mTc]pertechnetate was eluted from a Mo-99/Tc99m generator (Perkin-Elmer, Boston, MA). The S-acetyl NHSmercaptoacetyltriglycine (MAG3) was synthesized in house [12]. The HPLC system was equipped with a 515 pump, an in-line dual UV detector, and an in-line gamma-radioactivity detector under the control of Millennium 32 software (Waters, Milford, MA).
MORF Conjugation The 12-mer MORF sequences used in this study are shown in Table 1. Eub is a universal probe specific for a common sequence in the 16S rRNA present in most if not all bacteria; it served as a positive control [1]. The nonEub is a random sequence and used as a negative control. Eco and Kpn MORFs are species-specific bacterial probes. Eco is complementary to the 16S rRNA of E. coli [11], and Kpn is complementary to the 23S rRNA of K. pneumoniae [2]. The MORF sequences were synthesized by Gene Tools (Philomath, OR) with a primary amine attached via a 6-carbon linker on the 3′ equivalent end for conjugation to the MAG3 chelator for radiolabeling with Tc-99m or to the fluorophore AF633 for fluorescence microscopy. The MORF oligomers were conjugated with MAG3 using methods standard in this laboratory [12]. In brief, a solution of 300 μg oligomer in 200 μl of 0.3 M HEPES buffer, pH 8, was added to 0.7–1.0 mg of MAG3 powder and the suspension immediately mixed on a vortex to form a clear solution. The sample was left undisturbed for 1 h at room temperature before the addition of 50 μl of 1 M ammonium acetate and 120 μl of freshly prepared 20 mg/ml stannous chloride (SnCl2-2H2O) in tartrate buffer (100 mg/ml sodium tartrate in 0.5 M ammonium bicarbonate, 0.25 M ammonium acetate, and 0.18 M ammonium hydroxide, pH 9.2). The sample was mixed and heated to 95 °C for 20 min, then cooled to room temperature. Absolute ethanol was added for a final concentration of 20 % (v/v) and the sample was purified on a 1 × 20 cm Biogel P2 size exclusion column (Bio-Rad, Hercules, CA) using 0.25 M ammonium acetate pH 7 as eluant. The MORF concentration was determined by OD at 265 nm and stored at −80 °C for further use. For fluorescence microscopy, the MORFs were conjugated with the fluorophore AF633. Briefly, MORF (200 μg) in 0.1 M sodium bicarbonate buffer pH 8.4 was mixed with AF633 (10 mg/ml in N-methyl-2-pyrrolidone, Sigma-Aldrich, St. Louis, MO) with an AF633 to a MORF molar ratio of 10:1. After 45 min in the dark, the mixture was purified on a 1 × 20 cm P2 column with eluant 0.25 M ammonium acetate buffer pH 7. The MORF concentration was determined at 265 nm and stored in the dark at −20 °C.
MORF Radiolabeling The MORF oligomers were radiolabeled with Tc-99m using methods standard in this laboratory [12]. In brief, the MAG3-
Materials and Methods
Table 1. MORF oligomer sequences and specificity
Bacterial Cultures and Materials
Name
12-mer sequence (5′ → 3′)
Specificity
Eub NonEub Eco Kpn
GCT GCC TCC CGT AGG GCA TCC TCA GCA AAG GTA TTA CAC CAG CGT GCC
Universal Random control E. coli K. pneumoniae
E. coli strain K12 was purchased from the E. coli Genetic Stock Center (Yale University, New Haven, CT) and grown in LuriaBertani (LB) medium. K. pneumoniae was purchased from the
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
conjugated MORF (about 1 μg in 4 μl) was added to a solution of 45 μl 0.25 M ammonium acetate and 15 μl 50 mg/ml tartrate solution followed by 2 μl of freshly prepared 10 mg/ml SnCl22H2O solution in 10 mM HCl with 1 mg/ml ascorbate. After mixing on a vortex, Tc-99m (2–5 μl with 200–500 μCi) was added with agitation, followed by heating at 95 °C for 20 min. Radiochemical purity was determined by size exclusion HPLC Superose 12 column (Amersham Pharmacia Biotech, Piscataway, NJ). The running solution was 20 % acetonitrile in 0.1 M Tris–HCl pH 8 at a flow rate of 0.6 ml/min. Radioactivity recovery was always greater than 95 % and radiochemical purity greater than 90 %. Specific activity was about 200 μCi/μg.
MORF Fluorescence In Situ Hybridization In preparation for FISH, K. pneumoniae and E. coli log-phase cultures were diluted 1:3 (v/v) with 4 % formaldehyde prepared in Dulbecco’s phosphate-buffered saline (D-PBS). The sample was mixed on a vortex and left at room temperature for at least 3 h, then spun at 12,000 rpm for 2 min at 4 °C. The sample was washed with D-PBS to remove residual formaldehyde, spun again, and the pellet resuspended in D-PBS with 108 to 109 cells per milliliter. Finally, the fixed cell suspension was mixed with an equal volume of cold absolute ethanol and stored at −20 °C overnight. For hybridization, 3 μl of the fixed bacterial cell suspension was deposited onto an eight-chambered glass slide (Lab-Tek, Rochester, NY) and air dried. The AF633-conjugated MORFs were added at 5 ng/μl in 150-μl buffer containing 750 mM NaCl, 100 mM Tris-Cl pH 7.8, 5 mM EDTA, 0.2 % bovine serum albumin, 0.01 % polyadenylic acid, 0.1 % sodium dodecyl sulfate (all from SigmaAldrich), and 10 % dextran sulfate (dextran MW 500 kD, Calbiochem, Gibbstown, NJ), as described by Ouverney [3], and incubated at 43 °C for 2 h. The chambers were washed with distilled water at 43 °C, and then washed with buffer containing 30 mM NaCl, 4 mM Tris-Cl pH 7.8, and 0.2 mM EDTA for 30 min at 43 °C with two changes of wash solution. The cell membrane dye FM1-43 (0.2 μl at 5 μg/μl) was added about 10 min before viewing the cells under oil immersion with ×100 objective on an Olympus IX-70 inverted fluorescence microscope (Olympus America, Inc., Center Valley, PA). The quantification of fluorescence in cells was measured by ImageJ software (NIH). At least five areas were drawn about the cells of interest and background, and the corrected total cell fluorescence was calculated.
Accumulation of Fluorescent and Radiolabeled MORFs in Live Bacteria An overnight culture of K. pneumoniae and E. coli were diluted 1:50 with their respective media, and 200 μl of the diluted cultures were mixed with 15 μl of AF633-MORF at 15 ng/μl. The samples were incubated for 2 h at 37 °C with rocking in the dark then washed with 0.85 % NaCl and resuspended in 200 μl 0.85 % NaCl. About 3 μl was placed into a single chamber of an eight-chambered glass slide followed by 0.2 μl (5 μg/μl) of FM1-43 membrane stain. The samples were then air dried and mounted with fluorescence mounting medium (Dako, Carpinteria, CA) and viewed with an oil immersion ×100 objective. The incorporation and binding of the Tc-99m-labeled MORFs to RNA were also evaluated in live cells. Overnight cultures of K.
pneumoniae and E. coli were diluted 1:50 with media and 5 ml were mixed with 50 nmol of Tc-99m-labeled Eub, Kpn, or Eco MORFs and incubated at 37 °C on a lab rocker for 2 h. Thereafter, the samples were spun and washed three times with 0.85 % NaCl. Total RNA was isolated using TRIzol® Max™ bacterial RNA isolation kit from Invitrogen (Eugene, OR) following the manufacturer’s instructions. The RNA concentration was determined by OD at 260 nm using 25 μl/ μg/cm extinction coefficient. The RNA fraction was carefully transferred to fresh tubes and measured for radioactivity in a gamma well counter, and results were reported as nanomoles bound per 107 cells. To determine the number of bacteria in the incubation mixture, 100 μl of the sample was serially diluted and each dilution was grown overnight on LB agar plates. The bacterial cell count was determined from the colony number on each plate and dilution factor.
Animal Studies The studies in mice were with the approval of the Institutional Animal Care and Use Committee. The biodistribution of the Tc99m-labeled Kpn was measured in CD-1 mice (Charles River Laboratories International, Inc., Wilmington, MA) with live or heat-killed K. pneumoniae or E. coli in one thigh. An overnight culture of K. pneumoniae or E. coli was diluted with culture medium to an OD of 0.6 at 600 nm (about 5 × 108 cells/ml). Half of the culture was used as the live preparation while the remaining half was heated in a boiling water bath for 30 min to kill the bacteria and to provide a sample for injection of bacterial debris and possibly intact rRNA [13]. Either the live or heat-killed preparation of K. pneumoniae or E. coli was injected subcutaneously (0.1 ml with about 5 × 107 cells) into the one thigh of mice (n = 4). After 2 h, about 1 μg of the [99mTc]Kpn in 0.1 ml saline was injected through a tail vein. Animals were euthanized at 60 min and the organs of interest and blood were removed, weighed, and counted in a gamma well counter. A similar set of mice (n = 3) were prepared as above for imaging on a NanoSPECT/CT (Bioscan, Washington, DC) small animal camera. Animals received live K. pneumoniae or E. coli subcutaneously in the right thigh and after 2 h [99mTc]Kpn, about 3 μg with 600 μCi, was delivered through a tail vein in 0.1 ml saline. Immediately after injection, the mice were anesthetized with 1–2 % isoflurane carried in oxygen, and whole-body scans were taken at 30, 60, and 90 min with 24 projections and 60 s per projection. The regions of interest (ROI) were drawn about the infected and noninfected thighs for the quantification of accumulated radioactivity for analysis using VivoQuant 2.1 software (inviCRO, Boston, MA).
Statistical Analysis The Student t test was used to test for significance where indicated.
Results MORF Fluorescence In Situ Hybridization The binding of AF633-MORFs to RNA of K. pneumoniae or E. coli was evaluated in fixed cells. Fig. 1 presents images by
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 1 Fluorescence in situ hybridization images of AF633-conjugated MORFs Eub, nonEub, Eco, and Kpn with fixed bacteria a E. coli and b K. pneumoniae. Top row (green) shows cells stained with the membrane stain FM1-43, and bottom row (red) shows cells with AF633-conjugated MORFs Eub, nonEub, Kpn, and Eco. (Magnification ×1000). c Quantification of cell fluorescence calculated for panels a E. coli and b K. pneumoniae (*p G 0.01).
FISH showing the binding of Eub, nonEub, Eco, and Kpn to E. coli and K. pneumoniae ( bottom row, panels a and b, respectively). The bacterial membrane stained with FM 1–43 is shown across the top row. Fig. 1c shows the the corrected quantification of total AF633-MORF cell fluorescence for Fig. 1a and b. A signal for Eub, the universal probe, is evident in both E. coli and K. pneumoniae indicating hybridization of the Eub sequence to RNA of both bacteria, while lower staining is evident for the nonEub (nonspecific probe) in both bacteria. Eco provides a positive signal in its specific bacteria E. coli, but also in nontarget K. pneumoniae. On the other hand, Kpn
shows near background signal to the nontarget bacteria E. coli, but a positive signal, significantly higher (* p G 0.01), to its specific bacteria K. pneumoniae, indicating specific binding.
Accumulation of Fluorescent and Radiolabeled MORFs in Live Bacteria Fluorescence microscopy was used to evaluate the accumulation of AF633-MORFs in live E. coli (Fig. 2a) and K. pneumoniae (Fig. 2b). Fig. 2c shows the corrected
Fig. 2 Fluorescence microscopy of AF633-conjugated Eub, Kpn, and Eco MORFs to live a E. coli and b K. pneumoniae. Top row (green) shows the cell membrane stained with FM1-43, bottom row (red) shows accumulations of AF633-conjugated MORFs in cells. (Magnification ×1000). c Quantification of cell fluorescence calculated for panels a E. coli and b K. pneumoniae (*p G 0.01).
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
quantification of total AF633-MORF cell fluorescence for Fig. 2a and b. Accumulation for the universal Eub is observed in both E. coli and K. pneumoniae. In comparison, Kpn shows significantly higher accumulation in its specific bacteria K. pneumoniae and only background signal in nontarget E. coli (*p G 0.01), indicating specificity for Kpn. In contrast, Eco shows only minimal accumulation in its specific bacteria E. coli, which is similar to its accumulation observed in K. pneumoniae. That the Eco shows binding to the nonspecific target and binding that is at a level similar to that obtained for the other probes suggests loss of RNA binding specificity. To further establish binding to bacterial rRNA, the Tc99m-labeled Eub, Kpn, and Eco MORFs were incubated with E. coli or K. pneumoniae and the RNA was isolated and counted. As shown in Fig. 3, Kpn shows a higher percentage of radioactivity bound to RNA from its specific bacteria K. pneumoniae compared to Eco (*p = 0.04) and Eub for the same target. Not surprising is that the difference was not significant in the latter case, which is expected as Eub is a universal probe. In addition, Kpn shows significantly higher binding to its specific bacteria K. pneumoniae compared to nontarget E. coli (**p = 0.03), strongly suggesting the specificity of Kpn MORF. On the other hand, Eco shows a higher percentage of radioactivity bound to RNA from its specific bacteria E. coli in comparison to Kpn and Eub, but then similar binding to RNA from both E. coli and K. pneumoniae, indicating lack of specificity.
Animal Studies The biodistribution of [99mTc]Kpn (1 μg, 200 μCi) was measured at 60 min in mice with either live or heat-killed K. pneumoniae or E. coli in one thigh (Fig. 4). The time of 60 min post administration was based on the experience of
this laboratory with other radiolabeled MORFs for pretargeting applications. It is known that similar oligomers clear from circulation within 30 min [7]. The delay to 60 min was to minimize background. Fig. 4 shows the percent injected dose per gram of [99mTc]Kpn in tissues for the four study groups. Each organ shows similar accumulation of probes regardless of study group. The kidneys show the highest accumulation and are presented on a separate scale in the inset. High accumulation was also found in the small intestine. The [99mTc]Kpn shows the highest accumulation in the target thigh with live K. pneumoniae, which is significantly higher than for mice with the nontarget live E. coli (*p = 0.05) and significantly higher than for mice with heat-killed K. pneumoniae (**p = 0.02). However, when comparing both heat-killed bacterial preparations, [99mTc]Kpn accumulation in the thigh with K. pneumoniae was not significantly different from the thigh with E. coli (#p = 0.16). Fig. 5 shows the SPECT/CT images of [99mTc]Kpn (3 μg, 600 μCi) obtained at 30, 60, and 90 min in mice with K. pneumoniae (panels a, b, c) or E. coli (panels d, e, f) in one thigh (left in images). As in the biodistribution study, high accumulations are observed in the kidneys and small intestines. The bladder is the primary route of excretion and is evident in all images. [99mTc]Kpn shows obvious accumulation in the thigh with its target bacteria K. pneumoniae (panels a, b, c) that does not decrease over the course of 90 min. In comparison, the animals with nontarget E. coli in the thigh (panels d, e, f) show much lower infected thigh accumulations at all times and a gradual decrease (washout) through 90 min. The corresponding radioactivities in the target thighs and normal thighs from the ROI of a similar area are shown in Fig. 6a. The resultant target to nontarget thigh ratios (T:NT), shown in Fig. 6b, for K. pneumoniae, increase gradually reaching 5.7 at 90 min, whereas with E. coli as target, the ratios are lower at all time points and decrease at the latest time. A study to block
Fig. 3 Evaluation of the incorporation and binding of Tc-99m-labeled MORFs into live bacteria. Tc-99m-labeled MORFs Eub, Eco, and Kpn were incubated with cultures of E. coli and K. pneumoniae, and the percent of radioactivity bound to RNA per 107 cells is shown. Mean of n = 3 with SD (*p = 0.04, **p = 0.03).
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 4 Biodistribution 60 min post administration of [99mTc]Kpn MORF (1 μg, 200 μCi) to mice (n = 4) receiving either live or heat-killed K. pneumoniae or E. coli in one thigh 2 h earlier (*p = 0.05, **p = 0.02, # not significant).
specific binding by delivering excess unlabeled material was not performed. It was reasoned that since bacteria grow at an exponential rate, together with the associated increase in target RNA, it would be too unpredictable to determine the mass of material for blocking.
Discussion The bacterial rRNA has regions that have remained unchanged over time while other regions have undergone random mutations leading to bacterial diversity. These conserved regions have been used as targets for oligomer probes in the identification of bacterial pathogens by in vitro methods such as ISH and FISH [4, 5]. The Eub sequence is one such probe and has been demonstrated by others to bind to a conserved region of the bacterial rRNA and thus useful to detect bacterial contamination [1]. However, it cannot distinguish between bacterial species [1]. Previously, we demonstrated that the Tc99m-labeled Eub probe may be useful to detect bacterial infection in vivo [9]. Alternatively, the rRNA variant region is an ideal site for species-specific probes. The objective of this study was to evaluate Tc-99m-labeled species-specific probes with targets in the variant region by in vitro methods and in mouse infection models. Probes against E. coli and K. pneumoniae were chosen for the study since both are the source of common, clinically relevant infections each with increasing multidrug resistance strains [14, 15]. The MORF sequences from which Eub, nonEub, Eco, and Kpn were derived were reported previously [1, 2, 11] for in vitro bacterial identification. Based on findings by Deere et al. [16], the optimum length for oligomers to cross the bacterial cell wall was determined to be 9 to 12-mer. Therefore, we reduced the sequences to 12-mer by removing 6 bases from the 3′ equivalent end for Eub and Eco and from 5′ equivalent end for nonEub and Kpn. The results in fixed bacteria by FISH presented in Fig. 1 demonstrate that Kpn shows specific hybridization to the RNA of K. pneumoniae. However, this was not the case for Eco as a positive signal was obtained in both E. coli and K. pneumoniae.
Fig. 5 SPECT/CT images after administration of [99mTc]Kpn MORF (3 μg, 600 μCi) in mice (n = 3) with either K. pneumoniae in the left thigh obtained at a 30, b 60, or c 90 min, or E. coli in the left thigh obtained at d 30, e 60, or f 90 min.
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
Fig. 6 a Radioactivity (μCi) in the target thighs and normal thighs. b Target/nontarget (T:NT) ratios when K. pneumoniae or E. coli infected thigh is the target organism at 30, 60, and 90 min.
The accumulation of MORFs into live bacteria was confirmed by fluorescence microscopy. Fig. 2 shows the positive accumulation of the Eub MORF in both live K. pneumoniae and E. coli. Specific accumulation was observed for Kpn in K. pneumoniae. However, for Eco, the fluorescent signal in its target bacteria E. coli was equivalent to that observed with Kpn and lower than with Eub. Also, Eco is bound to nontarget K. pneumoniae to a level greater than the Eub. Both observations in Figs. 1 and 2 suggest the lack of specificity for this probe. The absence of specificity is not likely a problem associated with membrane penetration. First of all, Deere J et al. [16] showed that MORFs 9 to 12 bases in length were effective in inhibiting gene expression in live bacterial cultures. Thus, the oligomers were able to pass through the membrane. Secondly, in our previous study [9], the 12-mer Eub showed accumulation in both E. coli K12 and a membrane permeable E. coli SM101, indicating that the 12-mer Eub oligomer has no problem penetrating the cell membrane. The results obtained with the radiolabeled MORFs in live E. coli or K. pneumoniae, presented in Fig. 3, are similar. [99mTc]Kpn shows specific binding only to the RNA from K. pneumoniae while [99mTc]Eco shows binding to RNA from both E. coli and K. pneumoniae. Again, these negative results with Eco suggest the lack of specificity. As stated before, the loss of specificity may be due to shortening the sequence or perhaps the state of bacterial growth, or possibly hindrance due to the secondary structure of the RNA complementary site. Since Kpn shows specific accumulation in live cells with the both fluorescent tag and radiolabel, it was studied in animal models. The biodistribution at 60 min post administration of the [99mTc]Kpn in mice with either live or heatkilled K. pneumoniae or E. coli, presented in Fig. 4, shows the significantly higher accumulation in the thigh of mice receiving live K. pneumoniae, in comparison to mice with nontarget E. coli (*p = 0.05), and significantly higher than mice receiving heat-killed K. pneumoniae (**p = 0.02). It has been reported that the portions of the bacterial rRNA can remain stable even at 100 °C [13, 17, 18]. Therefore, the
higher accumulation of Kpn in the thigh with either heatkilled K. pneumoniae or E. coli compared to the normal thigh may be in part the reason for this observation. Similar to that seen by necropsy, imaging shows obvious accumulation of [99mTc]Kpn in the thigh infected with its specific bacteria K. pneumoniae compared to the thigh with nontarget E. coli. The target thigh to nontarget thigh ratios for K. pneumoniae-infected mice increase over time reaching 5.7 at 90 min, whereas for [99mTc]Kpn in E. coli-infected mice, the ratios remained lower and decreased to 2.8 at 90 min, suggesting washout of the probe with non-target bacteria.
Conclusions The [99mTc]Kpn MORF oligomer with a sequence complementary to the variant region of the K. pneumoniae rRNA demonstrated specific accumulation in live K. pneumoniae in vitro and in vivo. The [99mTc]Kpn MORF showed obvious accumulation in its specific infection target in mice and may be useful for K. pneumoniae infection detection through imaging. Acknowledgments. Funding was provided by a grant from NIH no. AI070857. Compliance with Ethical Standards Conflict of Interest The authors declare that they have no conflict of interest. Ethics Approval The studies in mice were with the approval of the Institutional Animal Care and Use Committee.
References 1. Amann RI, Krumholz L, Stahl DA (1990) Fluorescent-oligonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology. J Bacteriol 172(2):762–770 2. Kempf VA, Trebesius K, Autenrieth IB (2000) Fluorescent in situ hybridization allows rapid identification of microorganisms in blood cultures. J Clin Microbiol 38(2):830–838 3. Ouverney CC, Fuhrman JA (1997) Increase in fluorescence intensity of 16S rRNA in situ hybridization in natural samples treated with chloramphenicol. Appl Environ Microbiol 63(7):2735–2740
L. Chen et al.: An Oligomer for Detecting K. Pneumoniae Infection
4. Moter A, Gobel UB (2000) Fluorescence in situ hybridization (FISH) for direct visualization of microorganisms. J Microbio Meth 41:85–112 5. Giovannoni SJ, DeLong EF, Olsen GJ et al (1988) Phylogenetic groupspecific oligodeoxynucleotide probes for identification of single microbial cells. Bacteriol 170(2):720–726 6. Summerton JE (2007) Morpholino, siRNA, and S-DNA compared: impact of structure and mechanism of action on off-target effects and sequence specificity. Curr Top Med Chem 7:651–660 7. Liu G, Mang’era K, Liu N et al (2002) Tumor pretargeting in mice using (99m)Tc-labeled morpholino, a DNA analog. J Nucl Med 43(3):384–391 8. Liu G, He J, Dou S et al (2004) Pretargeting in tumored mice with radiolabeled morpholino oligomer showing low kidney uptake. Eur J Nucl Med Mol Imaging 31(3):417–424 9. Chen L, Wang Y, Cheng D et al (2013) 99mTc-MORF oligomers specific for bacterial ribosomal RNA as potential specific infection imaging agents. Bioorg Med Chem 21(21):6523–6530 10. Wang Y, Chen L, Liu X et al (2013) Detection of Aspergillus fumigatus pulmonary fungal infections in mice with (99m)Tc-labeled MORF oligomers targeting ribosomal RNA. Nucl Med Biol 40(1):89–96 11. Geller BL, Deere JD, Stein DA et al (2003) Inhibition of gene expression in Escherichia coli by antisense phosphorodiamidate morpholino oligomers. Antimicrob Agents Chemother 47(10):3233–3239
12. Wang Y, Liu G, Hnatowich DJ (2006) Methods for MAG3 conjugation and 99mTc radiolabeling of biomolecules. Nat Protocols 1:1477–1480 13. Sheridan GEC, Masters CI, Shallcross JA et al (1998) Detection of mRNA by reverse transcription-PCR as an indicator of viability in Escherichia coli cells. Appl Environ Microb 64(4):1313–1318 14. Kallen AJ, Hidron AI, Patel J et al (2010) Multidrug resistance among Gram-negative pathogens that caused healthcare-associated infections reported to the national healthcare safety network, 2006–2008. Infect Control Hosp Epidemiol 31(5):528–531 15. Sahm DF, Thornsberry C, Mayfield DC et al (2001) Multidrug-resistant urinary tract isolates of Escherichia coli: prevalence and patient demographics in the United States in 2000. Antimicrob Agents Chemother 45(5):1402–1406 16. Deere J, Iversen P, Geller BL (2005) Antisense phosphorodiamidate morpholino oligomer length and target position effects on gene-specific inhibition in Escherichia coli. Antimicrob Agents Chemother 49:249–255 17. Uyttendaele M, Bastiaansen A, Debevere J (1997) Evaluation of the NASBA nucleic acid amplification system for assessment of the viability of Campylobacter jejuni. Int J Food Microbiol 37:13–20 18. Aellen S, Que YA, Guignard B et al (2006) Detection of live and antibiotic-killed bacteria by quantitative real-time PCR of specific fragments of rRNA. Antimicrob Agents Chemother 50(6):1913–1920
