Development of Brain-Tumor-Targeted Benzothiazole-Based Boron Complex for Boron Neutron Capture Therapy

Abstract

Boron neutron capture therapy (BNCT) is a precision treatment technology that ideally damages only boron-accumulating cells. The effectiveness of BNCT depends on the amount of boron in the tumor cells and the concentration ratio between normal and tumor tissues. Therefore, for successful brain-tumor treatment using BNCT, it is essential to develop a drug with high blood–brain barrier (BBB) permeability and high tumor accumulation. The benzothiazole-based boron complex 4-(benzo[d]thiazol-2-yl)phenylboronic acid (BTPB) is a hydrophobic, low-molecular-weight compound that has shown high BBB permeability and brain accumulation. The highest boron concentration of BTPB is 36.11 ± 2.73 μg/g (at 1 h post-injection) in the brain, and the highest brain/blood ratio is 3.94 ± 0.46 (at 2 h post-injection), which is sufficient for the BNCT drug condition. In addition, BTPB showed good tumor-targeting ability in vivo in a U87MG glioma tumor model. In this study, we conducted a biological evaluation of BTPB compared to boronophenylalanine as a novel drug for BNCT.

Boron neutron capture therapy (BNCT) is a precision treatment technology that selectively destroys only tumor cells that have accumulated boron drugs through irradiation with thermal neutrons.¹⁻⁵ It is mainly applied to malignant brain tumors, where surgery is difficult and chemotherapy or radiotherapy is ineffective. In developing a strategy to increase the treatment efficiency of BNCT, the selective accumulation of boron drugs in cancer cells is a very important factor. Especially in the brain region, the drug requires the additional ability to cross the blood–brain barrier (BBB),³⁻⁵ and the following conditions are required as a strategy to accomplish this: (i) it should be a non-ionized compound, (ii) the log P value is greater than 2, (iii) the molecular weight should be less than 400 Da, and (iv) the number of hydrogen bonds in the compound should not exceed 10.⁶

l-4-Dihydroxyborylphenylalanine (L-BPA) has been used clinically as a representative drug for BNCT.⁷⁻⁹ L-BPA is a phenylalanine derivative known to actively enter cells and cross the BBB via l-type amino acid transporter 1 (LAT-1).¹⁰ In the past few decades, various types of derivatives have been studied with the goal of achieving efficient drug delivery to target organs, but most of the candidates have difficulty passing through the BBB or the accumulation rate in the brain is not high.⁷

Benzothiazole, a heterocyclic compound, has been widely used as a scaffold in several theranostic agents for brain diseases (e.g., Alzheimer’s disease and brain tumors) owing to its high brain permeability and antitumor effects¹¹ and its good optical properties.¹² We previously reported that Gd-DO3A-benzothiazole (Gd-DO3A-BTA) exhibited antiproliferative activity while showing an MRI contrast effect by targeting tumors.¹³ Thereafter, it was applied to gadolinium neutron capture therapy (Gd-NCT) to demonstrate the therapeutic effect on tumors in vivo, but there was a limit to its application in brain diseases due to its low BBB permeability.

In light of the previous experience described above and with a goal of discovering new types of brain targeting agents, especially those that would be applicable to BNCT, we designed a novel compound containing benzothiazole and boric acid, 4-(benzo[d]thiazol-2-yl)phenylboronic acid (BTPB) (Figure 1).

Figure 1.

Chemical structures of BTPB and L-BPA.

BTPB is a hydrophobic complex with a log P of 0.32 (Clog P, 3.6) and a low molecular weight of 255 Da. Although BPA enters the cells through LAT-1, it has low BBB permeability as a result of its hydrophilicity, with a log P of −3.65 (Clog P, −2.103).¹⁴ This is well reflected in the results of a parallel artificial membrane permeability assay (PAMPA), which evaluated the BBB permeability of a substance using an in vitro method, considering passive and transcellular permeation excluding active transport.^15,16 As shown in Table 1, the BBB permeability was significantly higher for BTPB compared to L-BPA. This indicates that the brain uptake of L-BPA is a result of active, not passive, transport of LAT-1.¹⁰ Benzothiazole derivatives are known to penetrate the BBB and enter tumor cells through the aryl hydrocarbon receptor (AHR), which is expressed in the tumors.^11,12,17 However, even considering passive transport, except for active transport through receptors, BTPB showed very high BBB permeability. This resulted in high brain uptake in a biodistribution experiment, which is thought to be the effect of both passive and active transport through receptors in vivo.

Table 1. Effective Permeability (P_e) and BCS Codes of BTPB and L-BPA.

		Pe (×10–6 cm/s)
name	pH	1st	2nd	3rd	average	SD	BCS codea,b
BTPB	7.4	56.41	54.04	68.87	59.77	7.96	High (CNS+)
L-BPA	7.4	0	0	0	0	0	Low (CNS−)

^aBCS = Biopharmaceutical Classification System.

^b“High” indicates P_e > 0.4 and “Low” indicates P_e < 0.4; CNS+ indicates P_e > 10 and CNS– indicates P_e < 10.

A drug’s usefulness as a therapeutic agent is highly dependent on its ability to accumulate abundantly in the target organ and less so in other organs. Especially in BNCT, sufficient concentrations of a boron drug must be present and maintained in the target organ to produce the characteristic effects. The recommended boron concentration for BNCT is >20–35 μg ¹⁰B/g for the target tissue and <5 μg ¹⁰B/g for normal tissue, with a target/normal (T/N) ratio >3.^7,10 Therefore, various injection methods and improvements were tried to satisfy the conditions during the treatment time. As part of that effort, L-BPA was administered by intragastric (i.g.), intraperitoneal (i.p.), subcutaneous (s.c.), or intravenous (i.v.) infusion for several hours.¹⁸⁻²¹ In a study comparing L-BPA administered by s.c. and i.v. injection using infusion for 3 h, the brain/blood ratio was significantly higher following the former than the latter (p < 0.05), and most other tissues showed no difference.¹⁹ This indicates that s.c. injection may provide a sustained effect with slower absorption of the drug than i.v. using infusion. In addition, since the therapeutic dose is very high compared to the diagnostic dose, the solvent for administering an appropriate amount should be made selected as carefully as that of the injection method. In an effort to improve the solubility of L-BPA, a study on complex formation with monosaccharides was conducted.²² As a result, currently, L-BPA is dissolved in a fructose-based solution and used in clinical practice as an L-BPA-fructose complex (L-BPA-f).¹⁹ In order to apply BTPB in vivo, DMSO and Tween 80/saline were used as solvents to dissolve and disperse it. Future studies are needed to improve the solubility and stability of BTPB.

The data on the biodistribution of BTPB and L-BPA-f are shown in Figure 2 and Tables S1 and S2. BTPB exhibited the highest concentration in the brain among all the organs. The boron concentration of BTPB in the brain was maintained above 25 μg/g from 30 min to 4 h post-injection (p.i.). The highest value was 36.11 ± 5.37 at 2 h p.i., which was about 7 times the value of L-BPA-f (4.87 ± 0.86) for the same time period. In particular, the concentration of boron in the brain was higher than that of any previously developed boron drug for BNCT.¹⁸⁻²¹ Even at 4 h p.i., the boron value in the brain was 27.24 ± 9.53 μg/g, which was a concentration sufficient for BNCT. This may be due to the high boron concentration in the blood. Nevertheless, the brain/blood ratio of BTPB was maintained at >3.6 (up to 2 h p.i.), which is within the expected range for efficient treatment in BNCT. Also the brain/blood ratio of BTPB remained high until 7 h p.i. compared to L-BPA-f, as shown in Table 2.

Figure 2.

Boron concentrations in blood, brain, kidney, muscle and U87MG tumor (xenograft thigh) after i.p. injection of 500 mg/kg dose of (A) BTPB and (B) L-BPA-f.

Table 2. Organ-to-Blood Ratio of BTPB and L-BPA-f in Nude Mice Bearing U87MG Tumor (n = 4) at Different Times Post-injection.

	0.5 h		1 h		2 h		4 h		7 h
	BTPB	L-BPA-f	BTPB	L-BPA-f	BTPB	L-BPA-f	BTPB	L-BPA-f	BTPB	L-BPA-f
brain	3.60 ± 0.27	0.15 ± 0.02	3.61 ± 0.40	0.41 ± 0.07	3.94 ± 0.46	0.59 ± 0.04	2.44 ± 0.62	0.88 ± 0.09	1.31 ± 0.44	1.14 ± 0.21
liver	3.28 ± 0.44	0.91 ± 0.07	2.85 ± 0.28	0.90 ± 0.10	2.63 ± 0.42	0.89 ± 0.11	1.42 ± 0.22	0.86 ± 0.11	1.08 ± 0.34	0.83 ± 0.18
kidney	3.09 ± 0.27	3.70 ± 0.72	2.64 ± 0.36	3.48 ± 1.11	2.71 ± 0.07	3.49 ± 0.98	1.92 ± 0.57	2.71 ± 0.75	1.70 ± 0.30	1.90 ± 0.33
tumor	1.44 ± 0.30	0.68 ± 0.10	1.33 ± 0.18	1.58 ± 0.32	1.99 ± 0.13	1.90 ± 0.11	1.58 ± 0.16	2.55 ± 0.31	3.20 ± 2.97	2.27 ± 0.27

The boron concentration of BTPB in tumors was higher than that of L-BPA-f except at 1 h p.i. The tumor values of BTPB (vs L-BPA-f) were determined to be 12.39 ± 3.68 (vs 7.20 ± 1.83), 11.89 ± 3.37 (vs 15.63 ± 2.85), 18.29 ± 0.54 (vs 15.61 ± 1.42), 17.52 ± 5.41 (vs 8.79 ± 2.37), and 16.86 ± 11.19 (vs 1.47 ± 0.40) μg/g at 0.5, 1, 2, 4, and 7 h p.i., respectively.

The concentrations of boron in the blood for BTPB and L-BPA-f were similar, at about 9, up to 2 h. However, the blood boron concentration of BTPB was maintained up to 4 h, while that of L-BPA-f decreased. As a result, the tumor/blood ratio of BTPB (1.58 ± 0.16) was lower than that of L-BPA-f (2.55 ± 0.31) at 4 h p.i. Since this study used a s.c. injection tumor model, it does not reflect the results that would be obtained in an orthotopic brain tumor model. Therefore, it will be necessary to study the brain tumor/normal brain ratio in an orthotopic brain tumor model. According to the BioD results, BTPB showed in higher boron concentrations than L-BPA-f in all the organs except the kidney, and the accumulation levels in the liver and kidney were similar. This suggests that L-BPA-f is eliminated primarily via the kidneys, whereas BTPB is eliminated via the renal and hepatic pathways, reflecting the lipophilic nature of BTPB.

For successful BNCT, a small amount of drug should be present in normal tissue and a large amount of drug in the target tissue. In this respect, BTPB may be suitable as a BNCT drug. Figure 3 showed the cytotoxic effects of BTPB on tumor cells (U87MG) and normal cells (HEK-293). In the U87MG cells, the survival rate was less than 43% even at a BTPB concentration of 100 μM. On the other hand, it was confirmed that HEK-293 cells showed a more than 95% survival rate for concentrations up to 100 μM, but there was toxicity at higher concentrations. The IC₅₀ concentration of BTPB for U87MG and HEK-293 were 112.02 ± 6.31 μM and 264.63 ± 14.68 μM, respectively.

Figure 3.

Relative proliferation (%) of (A) U87MG and (B) HEK-293 at various concentrations of BTPB.

Table 3 shows the boron concentrations in the brain of various BNCT drugs. In order to achieve the appropriate dose of boron drug for the target organ, various types of drugs have been developed and various methods (dissolution, dose, injection, etc.) have been tried. Even if the same amount is injected, the amount that accumulates in the tissue may vary depending on the injection method. For example, BPA-f 500 mg/g was administered by s.c. and i.v. injection (infusion for 3 h), and 3 h later, the amounts of boron in the brain were found to be different, at 8.73 and 4.61, respectively.¹⁹ Even if the dose is increased, the T/N ratio is similar or lower. When the total amount of BPA-f was increased from 125 to 500 mg/kg by i.v. injection (infusion for 2 h), there was a similar result, with a T/N ratio of 3.7.²⁰ In another case, with a total amount of BPA of 300, 600, and 1200 mg/g administered by i.p. injection, the boron accumulated in the brain after 6 h increased to 3, 5, and 21 μg/g, while the T/N ratio decreased to 3, 2.5, and 1.16, respectively.²¹

Table 3. Summary of ¹⁰B Concentrations in Various BNCT Drugs.

	injection routea	injection amount	brain 10B(mg/g)b	brain/blood ratio	tumor 10B(mg/g)b	tumor/blood ratio	ICP time (h)	ref
BTPB
Balb/c nude mouse	i.p.	500 mg/kg	36.11 ± 2.74	3.94	18.30 ± 0.55	1.99	2	–
BPA-Fc
Balb/c nude mouse	i.p.	500 mg/kg	4.88 ± 0.86	0.59	15.61 ± 1.42	1.90	2	–
NMRI nude mouse	i.p.	700 mg/kg	5.4 ± 2.6	0.47	–	–	1.5	(18)
C3H/He mouse	s.c.	500 mg/kg	8.73 ± 1.20	0.62	19.56 ± 2.63	1.39	3	(19)
	i.v. infusion for 3 h	500 mg/kg	4.61 ± 2.28	0.35	20.17 ± 6.49	1.52	3
GS-9L rat	i.v. infusion for 2 h	125 mg/kg/hd	–	–	–	3.6	1e	(20)
		250 mg/kg/hd	–	–	–	3.7	1e
		500 mg/kg/hd	–	–	–	3.6	1e
BPA
Wistar rat	i.p.	300 mg/kg	3 ± 2f	3	20 ± 5	20	6	(21)
		600 mg/kg	5 ± 3f	2.5	36 ± 4	18	6
		1200 mg/kg	21 ± 4f	1.16	47 ± 12	2.6	6

^ai.p. = intraperitoneal, i.v. = intravenous, s.c. = subcutaneous.

^bAssuming that ¹⁰B enriched was used.

^cBPA-F = p-boronophenylalanine-fructose complex.

^dmg/kg/h = injection amount per body weight per hour.

^e1 h after the end of i.v. infusion of drug.

^fBrain gray matter.

In summary, the benzothiazole-based boron complex BTPB showed high brain permeability and tumor accumulation in an in vivo glioma model. It demonstrated the highest absorption in the brain of any boron drug for BNCT reported to date. For clinical application, future studies are needed to establish conditions with better pharmacokinetics for efficiently delivering drugs to tumors.

Glossary

Abbreviations

BNCT

boron neutron capture therapy

BBB

blood–brain barrier

BSH

disodium mercaptoundecahydro-closo-dodecaborate

L-BPA

l-4-dihydroxyborylphenylalanine

EPR

enhanced permeability and retention

LAT-1

l-type amino acid transporter 1

BTPB

4-(benzo[d]thiazol-2-yl)phenylboronic acid

AHR

aryl hydrocarbon receptor

DO3A

1,4,7,10-tetraazacyclododecane-1,4,7-tris-tert-butyl acetate

MRI

magnetic resonance imaging

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.2c00284.

• Details regarding materials and methods, synthetic characterization, PAMPA, determination of binding constant with HSA, octanol–water partition coefficient, in vitro and ex vivo fluorescence imaging, cell culture, cell fractions, brain tumor model, and in vitro and in vivo MRI protocols (PDF)

Author Contributions

Experiments, data curation, and draft manuscript preparation, J.-u.Y.; synthesis and characterization, S.K. and K.C.L.; cell culture, preparing animal model, and biodistribution, J.-u.Y. and S.K.; manuscript review and editing, J.Y.K. and J.-A.P.; project supervision, Y.J.L. and J.-A.P.; funding acquisition, J.-A.P. All authors have read and agreed to the published version of the manuscript.

This work was supported by a National Research Foundation of Korea (NRF) grant, funded by the Korean government (MSIT) (No. 2020R1A2C200790611). This work was also supported by a grant of the Korea Institute of Radiological and Medical Sciences (KIRAMS), funded by MSIT, Republic of Korea (No. 50536-2021).

The authors declare the following competing financial interest(s): J.-A.P. has patent no. 10-2021-0073886 pending to 2-2007-014865-4.