In Vivo Monitoring of Viable Bacteria by SPECT using 99m Tc-HYNIC(GH) 2 -UBI 29-41 and 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 in Infected Mice

Background: Chronic bacterial refractory infections remain critical


99m
Tc-HYNIC(GH) 2 -UBI 29-41 showed high organ distribution and was 2-10 times more distributed than 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 in all organs at 0.5, 2, and 3 h after the injection of labeled peptides in normal mice.SPECT imaging revealed that the accumulation of labeled peptides in the bacterial infection site (left thigh) was higher than that in the non-infection site (right thigh) for both peptides.For 99m Tc-HYNIC(GH) 2 -UBI 29-41, the target-to-non-target ratio (T/NT) was 3.3 when the viable bacterial count was 10 8 colony-forming units (cfu)/thigh and decreased to 1.2 when the viable bacterial count was 10 3 cfu/thigh.For 99m Tc-HYNIC(Tricine) 2 -UBI 29-41, the T/NT ratio was significantly higher when the viable bacterial count was 10 8 cfu/thigh, with a value of 10.0.However, the T/NT ratio was 1.3 at 10 3 cfu/thigh, which is similar to the value for 99m Tc-HYNIC(GH) 2 -UBI 29-41 with the same bacterial count.A high correlation was found between the T/NT ratios of each labeled peptide and the viable bacterial count at the infection site.The correlation was confirmed under antibacterial treatment conditions.

Introduction
Bacterial refractory infections continue to pose a serious global health concern.In addition to the development of antibacterial resistance, even though pathogenic bacteria remain sensitive to antibiotics, failed eradication of causative bacteria sometimes leads to relapse [1].Both physiological and genetic changes in bacteria are thought to contribute to drug resistance or tolerance and persistence in chronic infections [2].Some phenotypes are fastidious, as they are only maintained in vivo and somehow disappear once the bacteria are removed from the host.Therefore, establishing a novel method to evaluate bacterial behavior during drug treatment in individual animals over time is essential to accelerate next-generation antibacterial drug discovery research.In conventional acute infection models, counting viable bacteria by euthanizing animals at each time point makes it difficult to monitor bacterial behavior in individual animals.In contrast, imaging technology provides a useful non-invasive method to evaluate bacterial behavior.Two imaging methods show potential for evaluating infections: optical imaging and radioactivity-based methods.Van Oosten et al. reported the use of luciferaseengineered Staphylococcus aureus and fluorescently labeled vancomycin (vanco-800CW) in a mouse myositis model [3].Bioluminescence was used to indicate the localization of S. aureus and allowed the overlap with vanco-800CW to be determined.Ning et al. demonstrated the in vivo detection of Escherichia coli, S. aureus, Pseudomonas aeruginosa, and Bacillus subtilis using maltodextrin-based imaging probes in a rat myositis model [4].Tang et al. reported infection imaging using concanavalin A as a bacteria-targeting ligand, a nanoparticle carrier, and a near-infrared fluorescent dye in a mouse wound model [5].Optical imaging is considered useful for detecting bacterial infections; however, it can only be used if the infection model is close to the surface, such as in subcutaneous and thigh infections.Furthermore, the optical imaging sensitivity of bacterial counts was reported to be approximately 10 5 cfu/thigh.Empirically, the number of viable bacteria is reduced to around 10 2 -10 3 cfu/ thigh following treatment with antibacterial agents [3][4][5].However, if relapse occurs when treatment is discontinued, optical imaging is considered to have insufficient sensitivity to monitor the bacteria.
The other imaging technology is based on radioactivity.Radioactivity-based methods can be used to evaluate deep infections, such as lung infections, in addition to those near the surface, and they generally have higher sensitivity than optical imaging.Therefore, we selected single-photon emission computed tomography (SPECT) with 99m Tc-labeled probes to monitor the same animal using radioactivity.
Several radiolabeled agents, such as antibodies, antibiotics, and peptides, have been evaluated for imaging infections [6].Among them, ubiquicidin (UBI) 29-41 was selected as the probe to be examined in this study.UBI 29-41 is one of the infection probes used clinically that binds to a broad spectrum of Gram-positive and Gram-negative bacteria and fungi [7].UBI 29-41 is a cationic human antimicrobial peptide fragment with six positively charged residues (5 Arg + 1 Lys) that accumulates at the negatively charged surfaces of microorganisms.Nibbering et al. reported in vivo studies using UBI 29-41 in mice and rats infected with S. aureus [8].They monitored the efficacy of antibiotics and reported a high correlation between the accumulation of UBI 29-41 at the infection site and the dose of antibiotics administered.
It has been reported that the nature of the coligand affects the biodistribution of 99m Tc-labeled 6-hydrazinonicotinic acid (HYNIC)-chemotactic peptides [9].In this report, we synthesized two types of labeled UBI 29-41 with α-Dglucoheptonic acid (GH) and tricine as coligands.Using these two labeled peptides, we evaluated the feasibility of in vivo monitoring of viable bacterial counts by SPECT.

Microorganism
S. aureus 25923 (American Type Culture Collection) susceptible to CPFX (minimum inhibitory concentration < 1 µg/mL) was used.S. aureus 25923 was cultured on Brain-Heart Infusion Agar for 24 h at 37°C.The colony suspension was washed, counted by optical density, and used in the in vitro binding assay.The viable cell count was also determined by plating on Brain-Heart Infusion Agar.For the in vivo assay, a stock solution of S. aureus 25923 stored at −80°C was used.

In Vitro binding to S. aureus
Ten microliters of the preparation containing each labeled peptide and 10 µL of suspension containing 5×10 10 cfu/mL S. aureus were added to 80 µL of a binding buffer (20 mM phosphate buffered saline containing 0.01% Tween80 and 5 mM acetic acid, pH=5).For the serum conditions, 10 µL of mouse serum were added instead of 10 µL of the binding buffer.The suspensions were gently mixed using a vortex mixer and incubated at 37°C for 1 h.After incubation, the tubes were centrifuged at 5000 ×g for 10 min.The supernatant was removed, and the pellet was resuspended in 100 µL of binding buffer and re-centrifuged at 5000 ×g for 10 min.The supernatant was removed, and the radioactivity of the pellet was counted using a γ-counter.The radioactivity associated with the bacteria pellet was expressed as a percentage of the total 99m Tc activity added.

Animals
All procedures for animal studies were approved by the Institutional Animal Care and Use Committee of Shionogi & Co., Ltd.(Osaka, Japan).Specific-pathogen-free male ICR mice (CLEA Japan Inc., 5 weeks old) were used in the infection and SPECT studies.

Biodistribution
A 0.2-mL solution containing 30 kBq of 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 or 10 kBq of 99m Tc-HYNIC(GH) 2 -UBI 29-41 was administered via the tail vein of mice.Animals were euthanized by exsanguination, and the thoracic cavity was opened at 0.5, 2, and 3 h post injection.Organs were excised and weighed, and the activity of probes was counted using a γ-counter.Organ uptake was calculated as a percentage of the injected dose per gram of wet tissue (%ID/ organ, Table 1).

Treatment of animal infections with antibacterial agents
Normal and immunosuppressed mice were used.To generate immunosuppressed mice, CY was injected by intraperitoneal administration at 4 days and 1 day before infection.Mice were anesthetized with isoflurane, and 6.0×10 6 -2.0×10 7 cfu of bacteria in 0.1 mL saline were aseptically injected into the left thigh muscle of each mouse.The antimicrobial procedures (original) were as follows: mice were subcutaneously administered 10-100 mg/kg of CPFX once a day or three times a day for 2 days after infection (Figure 2A, B).In experiments to monitor regrowth, mice were administered CPFX only once 2 h after infection (Figure 2C).

SPECT imaging
At 46 h after infection, 0.2 mL of a solution containing 10-30 MBq of each labeled peptide was administered via the tail vein of mice (Figure 2A, B).In experiments for monitoring regrowth, the labeled peptide was administered at 22 h and 70 h after infection (Figure 2C).
The accumulation of each labeled peptide in the bacteriainfected site in mice was assessed by SPECT ISSN: 2581-7566 with surgical tape.Whole body images were acquired under the following conditions: energy window, 20% at 140 keV; projection limit, 20 s; projection count, 64; and rotation angle, 360 degrees.The total time for actual imaging was only ~20 min.After whole body imaging, the mice were euthanized, and the infected (left) and normal (right) legs were extracted.Leg-only images were then acquired using the same conditions.For image processing, adjusted regions of interest were drawn over the entire infected muscle (target [T]) and contralateral muscle (nontarget [NT]).The accumulation of each labeled peptide at the infection site was expressed as the ratio of the counts in the target and nontarget muscles (T/NT).

Determining the number of viable bacteria
After SPECT imaging, the entire infected thigh muscles were removed and individually homogenized in Mueller-Hinton broth.Serial dilutions of the thigh homogenate were plated on Brain-Heart Infusion Agar.The plates were then incubated for 24 h at 37°C, and the numbers of colonies were counted.The viable bacterial counts in thigh (log10 CFU/thigh) were calculated from the number of colony in the plate.

Statistical Analysis
The differences between log cfu before and after treatment with CPFX in mice were evaluated using the Student's t-test.The P values were calculated, and statistical significance was accepted within 95% confidence limits by SAS ® Version 9.4.All results were reported as means and SD.The Pearson correlation coefficient (r) was used to assess the correlation between the labeled peptide accumulation and the viable bacterial count.

In Vitro binding to S. aureus
The binding of 99m Tc-HYNIC(GH) 2 -UBI 29-41 to S. aureus (5×10 8 cfu/tube) without serum was 90.7±5.2% of the total 99m Tc activity, and in the presence of serum, the binding was 69.5±2.7%.The binding of 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 II SPECT 2H/XO SRI CT, TriFoil Imaging).Two hours after the labeled peptide injection, mice were anesthetized with isoflurane.Mice were then arranged lying face down on a SPECT/CT bed with both hind legs spread out and fixed   ISSN: 2581-7566 was 90.7±1.7%without serum and decreased to 13.2±1.0% in the presence of serum (Figure 4).

Biodistribution
The biodistributions of 99m Tc-HYNIC(GH) 2 -UBI 29-41 and 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 in normal mice at 0.5, 2, and 3 h are summarized in table 1.The data showed that the highest concentrations of radioactivity were measured in the kidney for both peptides.The accumulation of 99m Tc-HYNIC(GH) 2 -UBI 29-41 in the kidney increased over time, whereas 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 decreased.99m Tc-HYNIC(GH) 2 -UBI 29-41 showed high organ distribution and was 2-10 times more distributed than 99m Tc-HYNIC(Tricine) 2 -UBI 29-41 in all organs at all time points.This difference in organ distribution is attributed to the difference in coligands.The activity at the infection site in the thigh showed no further increase after 2 h; therefore, the SPECT images were acquired 2 h after peptide injection.

Effect of antibiotic administration
To evaluate the correlation between the labeled peptide accumulation and the viable bacterial count, it was necessary to establish various viable bacterial counts.In addition to the untreated (control) group, 2-day treatment with 10-100 mg/kg of CPFX was administered to mice.As a result, viable bacterial counts in the range of 10 3 -10 8 cfu/thigh were established (Figure 5).This range was considered sufficient to evaluate treatment with antibiotics in a chronic infection model.

Detection of labeled peptide accumulation at the infection site using SPECT
The bacterial infection site was imaged 2 h after the injection of each labeled peptide.Of the many tests performed with 10 3 -10 8 cfu/thigh, representative images for each labeled peptide in S. aureus-infected mice are shown in figure 6.For 99m Tc-HYNIC(GH) 2 -UBI 29-41, the T/NT ratio was 3.3 when the viable bacterial count was 10 8 cfu/thigh and decreased to 1.2 when the viable bacterial count was 10 3 cfu/thigh.For 99m Tc-HYNIC(Tricine) 2 -UBI 29-41, the T/ NT ratio was significantly higher when the viable bacterial count was 10 8 cfu/thigh, with a value of 10.0.However, the T/NT ratio was 1.3 at 10 3 cfu/thigh, which is similar to the value for 99m Tc-HYNIC(GH) 2 -UBI 29-41 with the same bacterial count.
The correlations between the accumulation of each labeled peptide and the viable counts of bacterial are shown in figure 7. The accumulation of each labeled peptide showed a high correlation with the viable bacterial count: r=0.906 and P=0.002 for 99m Tc-HYNIC(GH) 2 -UBI 29-41; r=0.857 and P=0.001 for 99m Tc-HYNIC(Tricine) 2 -UBI 29-41.

Monitoring for regrowth by SPECT
The mice were treated with CPFX 2 h after infection and first imaged at 24 h after infection when the bacteria had decreased.Afterward, individual mice were re-imaged at 72 h after infection when the remaining bacteria had regrown (Figure 2C).In the CPFX treatment group, the viable bacterial count was 10 5 cfu/thigh at 24 h after infection but increased to 10 8 cfu/thigh at 72 h.The T/NT ratio value increased from 1.8   at 24 h to 2.8 at 72 h after treatment with CPFX (Figure 8).
Next, we aimed to quantify the viable bacteria by measuring the amount of labeled peptide accumulated in the bacteria using SPECT.To this end, we investigated the correlation between the labeled peptide accumulation and the viable bacterial counts.To the best of our knowledge, this correlation has only been investigated by Lupetti et al. using 99m Tc-labeled fluconazole with Candida albicans at 10 6 -10 8 cfu/g tissue (about 10 5 -10 7 cfu/thigh), and a high correlation was observed [15].In this study, we used a dynamic range of bacterial counts from 10 3 -10 8 cfu/thigh.Empirically the number of viable bacteria is reduced to around 10 2 -10 3 cfu/thigh following treatment with antibacterial agents, the lower limit was set at 10 3 cfu/thigh.To achieve this bacterial count, mice were treated with CPFX.The bacterial count reached a minimum of 10 4 cfu/thigh on the first day of treatment and reached the target of 10 3 cfu/thigh on the second day of treatment (Figure 5).
In this report, it was possible to monitor decreases in the number of bacteria with antibacterial agents.This suggests that bacteria killed by antibacterial agents are eliminated from the site of infection by the immune system.Moreover, it has been reported that UBI does not accumulate in sterile inflammation, and its accumulation at the infected site is thought to occur only in viable bacteria [16,17].
This study determined that SPECT imaging with 99m Tc-HYNIC-UBI29-41 is a useful method to quantify viable bacteria in the range 10 3 -10 8 cfu/thigh by measuring the accumulation of labeled peptides.Furthermore, even with continuous SPECT imaging in individual mice, the accumulation of the labeled peptide showed a high correlation with the viable bacterial count.
In this study, we examined one bacterial species and model.Therefore, we aim to study other bacterial species and models for research on bacterial refractory infections involving persisters and biofilms in the future.

Conclusion
SPECT imaging can be used to quantify viable bacterial counts ranging from 10 3 to 10 8 cfu/thigh by measuring the accumulation of labeled antimicrobial peptides.This approach enables the monitoring of viable bacterial counts in live individual animals over time, which is required for the investigation of chronic bacterial refractory infections.

Figure 2 :
Figure 2: Schematic diagram of the experimental protocol infected with S. aureus in normal mice (A) in immunosuppressed mice (B) at monitoring for regrowth (C).

Figure 5 :
Figure 5: Viable bacterial count in thigh infected with S. aureus for normal mice treated with CPFX for 2 days.Results are expressed as the mean ± SD for three animals.*Values that were significantly (P < 0.05) different to the control.

Figure 8 :
Figure 8: Left axis (bar graph): Viable bacterial count in thigh infected with S. aureus for immunosuppressed mice treated with CPFX.Results are expressed as the mean ± SD for three animals.Right axis (line graph): The accumulation of 99m Tc-HYNIC(GH) 2 -UBI in the left thigh.Results are expressed as the mean for two animals.