Introduction

Currently, infection diseases are cause of many mortality in the world. Those bacteria that can cause infections are including Streptococcus, Staphylococcus and E. coli [1–3]. Streptococcus is a Gram-positive bacteria and responsible for many cases of disease like meningitis, bacterial pneumonia and endocarditis. Staphylococcus is a genus of Gram-positive bacteria can also cause a wide variety of diseases in humans and animals through either toxin production or penetration. E. coli is a gram-negative, rod-shaped bacterium and the virulent strains of which can cause various diseases such as gastroenteritis, urinary tract infections, and neonatal meningitis and in some rare cases, haemolytic-uremic syndrome, peritonitis, mastitis, septicemia and gram-negative pneumonia in human [4]. Intestinal mucosa-associated E. coli are observed in increased numbers in the inflammatory bowel diseases, Crohn’s disease and ulcerative colitis [5]. Invasive strains of E. coli exist in high numbers in the inflamed tissue, and the number of bacteria in the inflamed regions correlates to the severity of the bowel inflammation [6]. E. coli is one of the varieties of bacterial species that have been reported to play a major role in acute appendicitis which is one of the most frequent conditions that leads to emergency abdominal surgery [7–9].

Nuclear imaging techniques as non-invasive diagnostic procedures in infection diseases can be used prior to invasive methods. Early detection is the other advantage of these techniques especially in cases like appendicitis, even in the beginning of infectious process that delay can lead to perforation, abdominal abscess, peritonitis, sepsis, and even death [10].

Several radiopharmaceuticals such as 111In- or 99mTc-labeled white blood cells, 67Ga-citrate and 111In- or 99mTc labelled antibodies or antibody fragments, chemotactic peptides, liposomes and certain cytokines are used for SPECT imaging of infection and inflammation [11–15]. However, although these agents are widely accepted, none is considered satisfactory. In this way the research is now more focused on the development of particular tracers with more accuracy, sensitivity and specificity profile for the investigation of a particular infection or inflammation rather than searching for any ideal agent.

Recently the 99mTc labeled cationic antimicrobial peptide derived from human ubiquicidine (UBI) has been proposed as an infection imaging agent [16, 17]. The results of the studies indicated that 99mTc-HYNIC-UBI29–41 binds to bacteria and is accumulated at sites of infection [18–21]. As advancement in this field radiolabeled antibiotics have been developed which are highly specific for the microbial infections. [22, 23]. In different studies using labeled flouroquinolones, second and third generation cephalosporins, and also some antituberculous antibiotics their diagnostic specificity for the infection region have been showed [24–30]. However, antibiotics meet with the problem of antibiotic resistant bacteria [31, 32].

GMX [7-[(4Z)-3-(aminomethyl)-4-methoxyimino-pyrrolidin-1-yl]-1-yclopropyl-6-fluoro-4-oxo- 1,8-naphthyridine-3-carboxylic acid] is an oral broad spectrum quinolone antibacterial agent which has been shown to be active against most strains of the aerobic gram positive, aerobic gram negative and some other microorganisms [33]. It is used in the treatment of acute bacterial exacerbation of chronic bronchitis and mild-to-moderate pneumonia [34]. It acts by inhibiting DNA synthesis through the inhibition of both DNA gyrase and topoisomerase IV (TOPO IV), which are essential for bacterial growth [35]. Radiolabeling of GMX with 99mTc was performed by Shah et al. in 2011 for detecting Streptococcus pneumoniae infection with favorable results [36]. As early and rapid detection plays a vital role in the treatment outcome of some major diseases caused by E. Coli, in the present study our focus was on the development of a particular radiocomplex for imaging these infections. So, in order to achieve a radiocomplex with high accumulation in E. Coli infection, we hereby survey the optimized condition in preparation of 99mTc-GMX, stability in normal saline and human serum, lipophilicity, binding to E. Coli and its biological evaluation in artificially E. Coli infected mice as an in vivo scintigraphic agent.

Experimental

Materials and methods

Gemifloxacin mesylate obtained from Life Sciences Pharmaceuticals (LSP). All other chemicals obtained from Sigma-Aldrich and were used without further purification. Pertechnetate (99mTc\({{\rm O}_{4}}^{-}\)) was eluted from the alumina based an in-house 99Mo/99mTc generator with saline solution (0.9 % NaCl), and its activity was determined in a dose calibrator (Isomed, Germany). Quantitative gamma counting was performed using an EG&G/ORTEC Model 4001 M Mini Bin & Power Supply NaI(Tl) counter.

Preparation of 99mTc-GMX

A stock solution of GMX (concentration 20 mg/mL) was prepared by dissolving GMX in double distilled water. From this stock solution different amounts of GMX (0.5–5 mg) were carefully transferred to a vial. To this solution 20–100 µL (2 mg/mL) of freshly prepared stannous chloride dihydrate (SnCl2·2H2O) in nitrogen purged 0.1 M HCl was added into vial. Effect of different amounts of stannous chloride and also change of pH factor in formation were investigated. Radiolabeling of the formulation involves initial warming up of the vial to room temperature followed by the addition of 740 MBq of freshly eluted 99mTc\({{\rm O}_{4}}^{-}\) in maximum 1 mL of normal saline and incubation of the vial for 15 min at room temperature.

99mTc-GMX analysis

The 99mTc-GMX was characterized in NaCl 0.9 % (W/V) and in terms of radiochemical purity using thin layer radiochromatography (TLRC) and high performance liquid radiochromatography (HPLRC) system. TLRC plates (silica gel) were developed in ethanol:water:ammonium hydroxide (2:5:1 v/v) as well as in acetone. In the first solvent, free 99mTcO4 − and 99mTc-GMX move with solvent front with R f = 1 and the reduced technetium remain at the point of application (R f = 0). In the second solvent, 99mTcO4 − move with solvent front with R f = 1 and the other species remain at the point of application. The radioactivity was quantified by cutting the strip (1.5 × 10 cm2) into 1 cm pieces and counting in a well type gamma counter.

For radiochemical analysis by HPLRC 20 µL aliquots were taken out at 1, 4 and 6 h post reconstitution at room temperature and injected into the system with a JASCO 880-PU intelligent pump (Tokyo, Japan) equipped with a multiwavelength detector and a flow-through Raytest-Gabi γ-detector. CC 250/4.6 Nucleosil 120-5 C-18 column from Teknokroma was used. 0.1 % trifluoroacetic acid/water (Solvent A) and 0.1 % trifluoroacetic acid/acetonitrile (Solvent B) were used as a mobile phase in the following gradient: 0 min 95 % A (5 % B), 5 min 95 % A (5 % B), 25 min 0 % A (100 % B), 30 min 0 % A (100 % B), flow = 1 mL/min, λ = 280 nm.

Serum stability

Serum stability and radiocomplex transferred to serum protein was studied using the reported method [37]. Briefly, 1 mL of freshly prepared human serum was added to 100 µL of 99mTc-GMX. The mixture was incubated in 37 °C for 6 h. During incubation, samples of 100 μL aliquots were removed from the mixture after 1, 2, 4 and 6 h and treated with 100 μL of absolute ethanol. After that, the samples were centrifuged at 3000 rpm in 4 °C for 5 min to precipitate serum proteins. Radiochemical stability in serum was determined by taking samples of 20 μL of supernatant at different times up to 6 h incubation that were analyzed by HPLRC.

Lipophilicity determination

0.5 mL of the 99mTc-GMX was mixed with 0.5 mL of octanol in a 2 mL micro tube. The tube was vigorously vortexed over a period of 10 min and centrifuged at 3000×g for 5 min. Three aliquots of 100 μl were sampled from each layer and counted in the γ counter. The averaged activities from the aqueous and the octanol layers were used to calculate the log P values. The octanol-to-water partition coefficient (P o/w) was calculated by dividing the counts of the octanol phase by that of the aqueous phase.

In vitro binding with Escherichia coli

100 µL of the 99mTc-GMX was transferred to a test tube. Then, 0.9 mL of 50 % (v/v) 0.01 M acetic acid in phosphate buffer (pH 7.5) containing approximately 1 × 108 colony forming units (CFU) per mL viable E. Coli was added. The mixture was incubated for 1 h at 4 °C and thereafter the vials were centrifuged in a pre-cooled centrifuge at 2000×g at 4 °C for 5 min. The supernatant was removed, and analyzed by HPLRC method for evaluation of stability of 99mTc-GMX in incubated medium and binding condition. The bacterial pellet was gently re-suspended in 1 mL of buffer and re-centrifuged as above. The supernatant was again removed leaving the bacterial pellets in the test tube, which were counted for percent uptake by a gamma counter.

Biodistribution in E. Coli infected mice model

Animal experiments were performed in compliance with the regulations of our institution and with generally accepted guidelines governing such work (No 642). Male Balb/c mice (25–30 g) were infected by injection 0.1 mL of saline containing 1 × 108 colony-forming unit (CFU) E. Coli bacteria into right thigh muscle through needle gauge 19. Mice also were inflamed by injection of 0.1 mL of autoclaved turpentine oil into right thigh muscle. After 24 h, a group of three balb/c mice (25–30 g, 6 week) received 0.37 MBq of 99mTc-GMX in volume of 0.1 mL via a tail vein. The mice were sacrificed at different post injection times (1, 4 and 6 h) and the tissues and organs of interest were collected, immediately weighed and counted in a NaI well-type γ-counter. Subsequently, percentage uptake of the 99mTc-GMX in one gram of the blood, heart, lung, stomach, intestine, thyroid, liver, spleen, kidney, infected muscle (T) and normal muscle (NT) of balb/c mice was calculated as the percentage of the injected dose per gram tissue (%ID/g tissue). The values are expressed as mean ± SD.

Imaging studies

At time points of 1 and 4 h after injection, accumulation of the 99mTc-GMX in infected area was assessed by planar scintigraphy using the single head gamma camera (small area mobile, Siemens, 140 keV high sensitivity parallel whole collimator and 10 % window around 140 keV). Before the imaging, mice were anesthetized with 0.05 mL ketamine 10 % (3.3 mg) and 0.05 mL xylazine 2 % (1.33 mg) intra-peritoneally. After about 5 min, the animal was fixed on a board by covering with pieces of cloth for immobilization during the scanning.

Results and discussion

Radiocomplexation

The 99mTc-GMX structure as shown in Fig. 1 was prepared by reducing 99mTc in 99mTc\({{\rm O}_{4}}^{-}\) with stannous chloride. Technetium can have a number of oxidation states but the +5 oxidation state is the most common in Tc = O complexes, with d2 configuration. Crystal structural studies have shown that when the atoms implicated in coordination are π-donor Lewis bases as in the present case then the square pyramidal structure is very favored [38, 39]. Based on the above description, this Tc+5 = O 99mTc-GMX will have a square pyramidal configuration having 99mTc to ligand ratio of 1:2.

Fig. 1
figure 1

Proposed structure of the 99mTc-GMX

Full size image

In order to identify an easy to use synthetic route with a high radiolabeling yield of 99mTc-GMX, radiolabeling was performed by varying reaction parameters such as the amount of ligand, amount of reducing agent and the pH value of reaction system.

Highly radioactive ligands can suffer radiation damage and the presence of radioactive impurities will almost certainly lead to a reduction of the ratio of specific to non specific labeling therefore reducing the amount of ligand used in the formulation is still highly desirable. Also increasing the amount of ligand, the issue of solubility’s is raised. Our observation shows that if we increase the ligand concentration, solubility’s decrease. Therefore, one of the objectives was to reduce the amount of ligand used in the labeling process. As Table 1 shows 2 mg of ligand was the optimum amount of ligand in formulation to reach a high labeling yield for formation 99mTc-GMX.

Table 1 Influence of different amount of ligand on preparation yield of 99mTc-GMX
Full size table

SnCl2·2H2O was used to reduce 99mTc in 99mTc\({{\rm O}_{4}}^{-}\) (VII) to its lower oxidation state (V) and enable 99mTc (V) to react with the ligand. The best amounts of SnCl2·2H2O is required were observed to be 80–120 µg. As has been shown in Fig. 2 employing concentration of reducing agent less than 80 µg will not reduce all of 99mTc\({{\rm O}_{4}}^{-}\)(VII) to its lower oxidation state 99mTc (V) which in turn cause to lower labeling yield. On the other hand choosing concentration more than 120 µg further reduction of 99mTc (VII) to its lower oxidation state 99mTc (IV) is possible in which a colloidal 99mTcO2 is formed in aqueous media and as a result labeling yield also was reduced. It can be concluded that a stable 99mTc-GMX with high radiochemical purity (98.60 ± 0.70 %) can be acquired with use of 100 µg SnCl2·2H2O (Fig. 2).

Fig. 2
figure 2

Effect of SnCl2·2H2O on preparation yield of 99mTc-GMX. Values are expressed as mean ± SD (n = 3)

Full size image

The effect of reaction pH was also investigated and the optimal pH range to produce a high labeling yield was found to be between pH 4–5 (Fig. 3). The yield decrease as pH varies from above range. As a coordination complex of 99mTc is formed by means of bonds between technetium acting as Lewis acid and functional groups which act as Lewis bases (donate electron pairs), in the basic medium carboxylic moiety of GMX is neutralized and is not able to produce a stable complex with 99mTc.

Fig. 3
figure 3

Effect of reaction pH on 99mTc-GMX labeling yield. Values are expressed as mean ± SD (n = 3)

Full size image

99mTc-GMX was prepared in optimized condition (2 mg/100 µL GMX, 100 µg/50 µL SnCl2·2H2O in 0.1 M HCl, 370 MBq/1 mL of 99mTcO4 −, pH 4–5) for 15 min at room temperature (specific activity of 64.3 MBq/nmol). In the formulation reported previously [36], in best condition 2 mg of GMX has been labeled with 111 MBq of 99mTcO4 − which is more than tree time lower in specific activity compared with our product. They used 125 µg SnCl2·2H2O as a reducing agent and final pH for their formulation was 5.4. The higher specific activity obtained in our radiocomplexation could be duo to our optimized condition.

Radiochemical purity

The radiochemical purity of 99mTc-GMX was determined by TLRC. In the first part a solvent system which consisted of ethanol: water: ammonium hydroxide (2:5:1 v/v) employed as a mobile phase. The activity remained in the origin was negligible and corresponding to reduced 99mTc (Fig. 4a). In the second part choosing acetone as a solvent, the 99mTc-GMX remain at the origin (Rf = 0) where as free 99mTc pertechnetate will migrate with solvent front (Rf = 1) and its activity was near the background (Fig. 4b). 99mTc-GMX showed 98.60 ± 0.70 % radiochemical yield at specific activity of 0.148 GBq/μmol. In characterization of 99mTc-GMX by HPLRC two different peaks with retention times of 4.51 min and 14.55 min were observed (Fig. 5). The 99mTc complex has been characterized by comparison with the corresponding ligand. The UV-HPLRC chromatogram retention time of ligand was observed to be 15.25 min, which matches well with 14.55 min γ-HPLRC chromatogram retention time of 99mTc-GMX.

Fig. 4
figure 4

TLRC chromatograms of 99mTc-GMX developed in ethanol:water:ammonium hydroxide (2:5:1 v/v) with 99mTc-GMX in R f = 1 and 99mTcO2 in R f = 0 (a) and in acetone solvent with 99mTc-GMX in R f = 0 and 99mTcO4 in R f = 1 (b). 99mTc-GMX showed 98.60 ± 0.70 % labeling yield

Full size image
Fig. 5
figure 5

Reversed phase HPLRC chromatogram for ligand in multiwavelength detector (λ = 280 nm) (a) and for 99mTc-GMX in Raytest-Gabi γ-detector (b). CC 250/4.6 Nucleosil 120-5 C-18 column from Teknokroma was used. 0.1 % trifluoroacetic acid/water (Solvent A) and 0.1 % trifluoroacetic acid/acetonitrile (Solvent B) were used as a mobile phase in the following gradient: 0 min 95 % A (5 % B), 5 min 95 % A (5 % B), 25 min 0 % A (100 % B), 30 min 0 % A (100 % B), flow = 1 mL/min

Full size image

Stability and binding to E. Coli

The radiochemical purity of the 99mTc-GMX was >95 % over the observed period of 6 h (Fig. 6). No decomposition of the complex was observed in this time period, suggesting its high stability in normal saline at room temperature. Radiochemical stability of 99mTc-GMX in human serum was >90 % up to 6 h indicating its quality of being appropriate from stability outlook (Fig. 6). So far the main drawback in labeling of ciprofloxacin and its similar structures are colloid impurity and their instability which were subject of discussion in previous studies by different groups [40, 41] but here with optimization of labeling condition high labeling yield and stability for 99mTc-GMX is achieved.

Fig. 6
figure 6

In vitro stability of the 99mTc-GMX at different time in saline and serum

Full size image

The radiocomplex transferred to human serum proteins was about 66 ± 7 % after 6 h. The partition coefficient of the 99mTc-GMX was determined by distribution in octanol and water, and its log P value was found to be −0.12 ± 0.04, signifying low lipophilicity which could explain the accumulation of the 99mTc-GMX in the kidney and liver. In vitro binding of 99mTc-GMX to E. Coli is shown in Fig. 7. The 99mTc-GMX radioactivity bound to bacteria was 34.6 % at 1 h after incubation while this value was significantly low for 99mTc\({{\rm O}_{4}}^{-}\) (9.0 %) which showed the binding was specific. No decomposition of 99mTc-GMX was observed in HPLRC which is also an indication of the stability of 99mTc-GMX in the evaluated condition.

Fig. 7
figure 7

In vitro binding of 99mTc-GMX with E. Coli in comparison with 99mTc\({{\rm O}_{4}}^{-}\) 1 h after incubation at 4 °C

Full size image

Biodistribution and imaging

Biological evaluation of 99mTc-GMX was performed in Balb/c mice. The results are shown in Tables 2, 3 and Fig. 8. The uptake value for blood was high and with mild clearance decreased from 8.42 ± 1.50 %ID/g at 1 h to 2.56 ± 1.55 %ID/g at 4 h. Similarly, the level of the 99mTc-GMX in liver decreased from 18.48 ± 2.45 %ID/g at 1 h to 17.73 ± 2.27 %ID/g at 4 h post injection. For the kidneys the 99mTc-GMX showed decreasing uptake from 14.26 ± 1.35 %ID/g at 1 h to 11.84 ± 1.51 %ID/g at 4 h post injection. The presence of the high activity in liver and kidney suggesting that the hepatobiliary and urinary systems is the major route of excretion of administered radioactivity.

Table 2 Biodistribution of 99mTc-GMX in artificially E. Coli infected mice (%ID/g ± SD, n = 3)
Full size table
Table 3 Biodistribution of 99mTc-GMX in aseptic inflamed mice (%ID/g ± SD, n = 3)
Full size table
Fig. 8
figure 8

The comparison uptake of 99mTc-GMX in E. Coli infected muscles with normal and inflamed muscles of mice at different time point (1, 4 and 6 h) after injection

Full size image

The radioactivity concentration of infected muscle by E. Coli at 1 h post injection was 1.64 ± 0.15 %ID/g which decreased to 1.27 ± 0.12 %ID/g at 4 h post injection. The ratio uptake of T/NT was 1.49 at 1 h post injection and 1.89 at 4 h post injection also ratio for infected muscle to inflamed muscle was 1.43 and 1.89 at 1 and 4 h post injection respectively. The increase of the T/NT ratio as time elapsed from 1 to 4 h is may be due to the clearance of non specific uptake from normal and inflamed regions and on the other hand this high value for retention shows 99mTc-GMX have specific affinity to Gram-negative bacterial infection site, although the radioactivity concentration in infectious muscle at 4 h post injection is lower than 1 h post injection. The uptake of 99mTc-GMX in inflamed muscle was similar to normal muscle which showed that its uptake was not specific in aseptic inflammation regions and confirmed the affinity of 99mTc-GMX for infected muscles due to its specific binding.

In previous study Streptococcus pneumonia infected rats was used to assess the in vivo behavior of the 99mTc-Gemifloxacin [36]. The ratio of T/NT activity which accumulated at the site of infectious muscle compared to normal muscle was 3.25 in the Streptococcus pneumonia model. In the present study although 99mTc-GMX was prepared with high specific activity, accumulation of 99mTc-GMX was lower in the E. Coli induced infection model compared to the Streptococcus pneumonia model which show lower susceptibility of 99mTc-GMX for E. Coli induced abscess. In another study by Sah et al. [26] 99mTc-fleroxacin has been evaluated for E. Coli infection imaging. Their results showed uptake ratio of 2.78 for infected to inflamed muscle at 1 h post injection while the blood and stomach-intestines uptake were 11.2 ± 1.50 and 7.25 ± 1.50 %ID/g at 1 h post injection respectively. Compared to our results although the ratio for infected to inflamed muscle 99mTc-fleroxacin was higher than that of 99mTc-GMX (2.78 vs 1.49) however values for blood, stomach and intestines for 99mTc-GMX was lower than that of for 99mTc-fleroxacin (8.42 ± 1.50, 1.68 ± 1.00 (means uptake values for stomach and intestines) %ID/g at 1 h versus 11.2 ± 1.50 and 7.25 ± 1.50 %ID/g at 1 h post injection respectively). This high uptake for 99mTc-fleroxacin especially in stomach and intestine is an indication of the in vivo instability of the radiocomplex which subsequently could prevent accurate diagnosis of E. Coli infection in the intestine, especially in the case of intestinal infections such as appendicitis. On the other hand 99mTc-GMX low stomach and intestine activity uptake would be an advantage for imaging of abdominal infection diseases.

An imaging study at 4 h post injection showed uptake for 99mTc-GMX in the site of infection (Fig. 9). The uptake in all organs was decreased significantly after 6 h which shows that elimination is time depended and the early time up to 4 h is the best time for infection detection.

Fig. 9
figure 9

Anterior scintigraphy of mice with right thigh muscle artificially E. Coli infected 4 h post injection of 99mTc-GMX after masking of abdominal region while infection site has been shown with arrow

Full size image

Conclusion

Gemifloxacin belongs to the class of medicines known as quinolone antibiotics. It works by killing bacteria or preventing their growth. Gemifloxacin is used to treat bronchitis and pneumonia caused by bacterial infections. Since Gemifloxacin is active against bacteria such as E. Coli it is expected that 99mTc-GMX can play a role in diagnose of infection diseases with E. Coli origin. In this study, we have shown preparation and evaluation of a 99mTc-GMX with high labeling yield. Based on the data obtained from this study, the 99mTc-GMX was stable, reproducible with high labeling efficiency with desirable characteristics making it a promising agent for imaging of Gram-negative infectious lesions. According to the results of in vivo biodistribution studies, we found that the uptake by infectious sites is high and has a good retention time. These promising characteristics make 99mTc-GMX as a very suitable candidate for diagnostic of E. Coli infectious sites in nuclear medicine. Further investigation on the ability of 99mTc-GMX to bind and detect other bacteria strains should be pursued.