*Review* **In Situ Delivery and Production System (***i***DPS) of Anti-Cancer Molecules with Gene-Engineered** *Bifidobacterium*

**Shun'ichiro Taniguchi**

Department of Hematology and Medical Oncology, Shinshu University School of Medicine, Matsumoto City 390-8621, Japan; stangch@shinshu-u.ac.jp

**Abstract:** To selectively and continuously produce anti-cancer molecules specifically in malignant tumors, we have established an in situ delivery and production system (*i*DPS) with *Bifidobacterium* as a micro-factory of various anti-cancer agents. By focusing on the characteristic hypoxia in cancer tissue for a tumor-specific target, we employed a gene-engineered obligate anaerobic and non-pathogenic bacterium, *Bifidobacterium*, as a tool for systemic drug administration. This review presents and discusses the anti-tumor effects and safety of the *i*DPS production of numerous anti-cancer molecules and addresses the problems to be improved by directing attention mainly to the hallmark vasculature and so-called enhanced permeability and retention effect of tumors.

**Keywords:** solid cancer; microenvironment; hypoxia; cancer therapy; DDS; anaerobic bacteria; *Bifidobacterium*; bacterial therapy; *i*DPS; EPR

## **1. Introduction**

## *1.1. Molecular Target Cancer Therapy and its Limitations*

One of the greatest advances in recent cancer research is the identification of driving genes, which are specific to cancer type and critically responsible for cellular growth. Many molecular targeting drugs have now been developed, leading to increased specificity to cancers and fewer side effects on bone marrow and digestive organs, the most common areas harmed by classic chemotherapeutic anti-cancer drugs [1]. However, there remain obstacles in cancer treatment, such as the appearance of drug-resistant cells from heterogeneous cancer cell populations, which leads to recurrence. There are also new types of side effects differing qualitatively from those of conventional cytotoxic anti-cancer drugs [2]. In the case of solid cancers, simple but troublesome problems exist in that the exposed dose of drugs is often insufficient to kill cancer cells as compared with hematopoietic cancers, which are more readily exposed to anti-cancer drugs. Accordingly, it is essential to develop a selective drug administration system to deliver large amounts of anti-cancer drugs to solid tumors and overcome the situation.

In their review on the hallmarks of recent cancer research leading to the identification of driving genes and development of molecular targeting drugs and antibody drugs, Hanahan and Weinberg pointed out the importance of the characteristic microenvironment of cancer, including low oxygen pressure (pO2) and immune avoidance conditions [3], as targets for emerging therapies [4]. Thus, it may be desirable to focus on the tumor microenvironment rather than attack individual cancers, which contain heterogenous populations of cells [5] that can produce drug resistance.

#### *1.2. The Enhanced Permeability and Retention (EPR) Effect*

To overcome the above difficulties, it was suggested that anti-cancer drugs of a high molecular weight should be employed to make use of the characteristic vasculature of tumor tissues [6]. In malignant tumors, leaky vasculature with 100–1000 nm pores is generally formed due to the rapid but immature formation of vessels in association with cancer growth. In addition to fragile blood vessels [7,8], there is usually poor lymphatic

**Citation:** Taniguchi, S. In Situ Delivery and Production System (*i*DPS) of Anti-Cancer Molecules with Gene-Engineered *Bifidobacterium*. *J. Pers. Med.* **2021**, *11*, 566. https:// doi.org/10.3390/jpm11060566

Academic Editor: Jun Fang

Received: 19 May 2021 Accepted: 15 June 2021 Published: 17 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

drainage, resulting in the retention of large molecules in tumor tissue. This phenomenon was discovered by Maeda, who named it the EPR effect. By focusing on EPR, many trials have managed to successfully target tumor tissues [9].

## *1.3. Our Approach Targeting the Low pO<sup>2</sup> in Solid Cancers with Bifidobacterium*

We have been working to establish a system for the selective and continuous production of anti-cancer molecules in tumors [10,11]. For this purpose, we directed our attention to the hypoxic conditions in solid cancers for a therapeutic target. As a tool for the local production of anti-cancer drugs, we adopted the obligate anaerobic and non-pathogenic bacterium, *Bifidobacterium*. Our first paper in 1980 showed the selective growth of *Bifidobacterium* in the cancer tissues of tumor-bearing mice where the bacilli were intravenously injected [12]. Currently, several gene-engineered *Bifidobacterium* lines have been established to produce numerous anti-cancer molecules in a process we have termed the in situ delivery and production system (*i*DPS). A new anti-tumor drug made with the *i*DPS is now undergoing clinical testing.

#### *1.4. Hypoxia and Immature Blood Vasculature in Malignant Tumors*

Tumor hypoxia is a well-known phenomenon [13–15]. The median pO<sup>2</sup> in tumor tissues is lower than that in normal tissues, which is never below 10 mm Hg. Hypoxia is generally observed in tumor tissues in spite of active angiogenesis. This paradoxical phenomenon has been attributed to impaired vascular communication and networks leading to functional, but chaotic, shunting and dysfunctional microcirculation [16]. Thus, even in the presence of blood vessels carrying fresh oxygen and nutrients, shunts to other blood vessels form easily, such that the downstream vessel will be not supplied, leading to hypoxia and/or necrosis [16].

Our project idea of the *i*DPS originally derived from Malmgren's work of injecting anaerobic *Clostridium tetani* spores into animals [17]. In his experiments, tumor-bearing mice died of tetanus due to the strongly toxic neuro-active substances produced by germinated *Clostridium tetani*, while normal mice survived. This was a strong piece of biological evidence for hypoxia existing in tumors; the spores of the anaerobic bacteria had germinated and produced strong toxins in the hypoxic conditions of the tumors, and the highly toxic poison leaked from the tumor tissues to kill the host despite very small amounts. This led us to the idea of non-pathogenic anaerobic bacteria as a tool for safely and selectively targeting solid cancers while sparing the host.

#### *1.5. Bacterial Therapy for Cancer*

Bacterial cancer therapy has a long history. The accidental tumor regression by clostridial infection clinically observed by Vautier in 1813 [18] launched a number of bacterial therapy experiments. Later, the recovery of a patient with inoperable lymphosarcoma by erysipelas prompted Coley to begin treating cancer patients with live erysipelas agents and/or bacterial toxins [19]. However, ensuing trials were limited, likely due to the difficulty in controlling toxicity and a shift to chemotherapy and radiation treatment. Recently, however, bacterial therapy has been revived with the use of genetic manipulation and is a promising method for cancer therapy [20–23].

Nowadays, several clinical studies on bacterial cancer therapies, including our own, are underway in the United States. In most cases, *Salmonella* or *Clostridium* is used. The earliest clinical trials approved by the American Food and Drug Administration (FDA) were carried out by Rosenberg's group at the National Institutes of Health (NIH) using *Salmonella* [24] and by a group at Johns Hopkins University with *Clostridium* [25]. Other trials have produced anti-tumor responses in both animals and human clinical studies [26–28]. In the above cases, the bacteria are attenuated due to their highly virulent nature. Despite concerns on the appearance of revertants and whether facultative anaerobic bacteria can be completely removed from normal aerobic tissues and cells, recent clinical trials have largely managed to maintain host safety [23].

#### The EPR Effect's Importance in Bacterial Therapy of bacteria localized in tumors increased by ten-fold as compared with controls [30].

*J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 3 of 21

trials have largely managed to maintain host safety [23].

The EPR Effect's Importance in Bacterial Therapy

Maeda's group noted that though EPR effect is applicable to particles of µm size (i.e., bacteria) or macromolecules of ∼1000 kDa, nanocarriers with diameters of ∼100 nm are known to achieve better or optimal EPR-based tumor accumulation [29]. Thus, the accumulation of even aerobic bacteria in tumors may be explained by the EPR effect [30]. Importantly, this effect can be augmented by vascular dynamic modifiers, such as nitroglycerin, from which nitric oxide is produced in the hypoxic conditions of tumors. In his attempts to target tumors with lactobacillus, Maeda's group showed that the number of bacteria localized in tumors increased by ten-fold as compared with controls [30]. **2. Our Trials for Bacterial Therapy**  *2.1. Selective Localization of i.v. Injected Bifidobacterium in Tumors* Figure 1 shows our first data reported in 1980 [12] on the growth of i.v. injected *Bifidobacterium* in tumors (Figure 1a) along with the results of our genetically engineered

Despite concerns on the appearance of revertants and whether facultative anaerobic bacteria can be completely removed from normal aerobic tissues and cells, recent clinical

Maeda's group noted that though EPR effect is applicable to particles of μm size (i.e.,

bacteria) or macromolecules of ∼1000 kDa, nanocarriers with diameters of ∼100 nm are known to achieve better or optimal EPR-based tumor accumulation [29]. Thus, the accumulation of even aerobic bacteria in tumors may be explained by the EPR effect [30].

his attempts to target tumors with lactobacillus, Maeda's group showed that the number

#### **2. Our Trials for Bacterial Therapy** *Bifidobacterium* (Figure 1b) (Farumashia,15, (5), 438–440, 2015 [in Japanese]). In both cases,

#### *2.1. Selective Localization of i.v. Injected Bifidobacterium in Tumors*

Figure 1 shows our first data reported in 1980 [12] on the growth of i.v. injected *Bifidobacterium* in tumors (Figure 1a) along with the results of our genetically engineered *Bifidobacterium* (Figure 1b) (Farumashia,15, (5), 438–440, 2015 [in Japanese]). In both cases, *Bifidobacterium* selectively grew in tumor tissues and became rapidly diminished in the blood and normal tissues, including the relatively hypoxic bone marrow known as a niche for hematopoietic stem cells. No acute toxicity was observed, and the survival of mice i.v. injected with *Bifidobacterium* was comparable with that of control animals, demonstrating the absence of chronic toxicity [12]. *Bifidobacterium* selectively grew in tumor tissues and became rapidly diminished in the blood and normal tissues, including the relatively hypoxic bone marrow known as a niche for hematopoietic stem cells. No acute toxicity was observed, and the survival of mice i.v. injected with *Bifidobacterium* was comparable with that of control animals, demonstrating the absence of chronic toxicity [12].

**Figure 1.** *Cont*.

**Figure 1.** (**a**) Specific distribution of *Bifidobacterium (bifidum)* in tumor tissues following a single *i.v.* injection of 5 × 10<sup>6</sup> c.f.u./mouse into Ehlich solid tumor-bearing mice. Each point represents the mean c.f.u./g tissue (*n* = 8). This figure was adopted from a previous study [12]. (**b**) Recombinant *Bifidobacterium (longum)* (6.9 × 10<sup>8</sup> c.f.u./mouse) was injected *i.v.* on day 1. Bacterial cells in each tissue were counted on day 2, 3, 4, and 7 by plating assays. \* Mean c.f.u./g tissue (*n* = 5). This figure was adopted from Farumashia,15, (5), 438–440 (in Japanese). **Figure 1.** (**a**) Specific distribution of *Bifidobacterium (bifidum)* in tumor tissues following a single *i.v.* injection of 5 <sup>×</sup> <sup>10</sup><sup>6</sup> c.f.u./mouse into Ehlich solid tumor-bearing mice. Each point represents the mean c.f.u./g tissue (*n* = 8). This figure was adopted from a previous study [12]. (**b**) Recombinant *Bifidobacterium (longum)* (6.9 <sup>×</sup> <sup>10</sup><sup>8</sup> c.f.u./mouse) was injected *i.v.* on day 1. Bacterial cells in each tissue were counted on day 2, 3, 4, and 7 by plating assays. \* Mean c.f.u./g tissue (*n* = 5). This figure was adopted from Farumashia,15, (5), 438–440 (in Japanese).

It was noteworthy that the survival of the normal animals in Malmgren's experiments [17] indicated that the *Clostridium* spores did not geminate in the bone marrow, thus demonstrating that bone marrow pO<sup>2</sup> was insufficiently low for obligate anaerobic bacteria germination and/or an inadequate EPR effect to trap the spores or bacteria. This phenomenon also likely occurred in our experiments. It was noteworthy that the survival of the normal animals in Malmgren's experiments [17] indicated that the *Clostridium* spores did not geminate in the bone marrow, thus demonstrating that bone marrow pO<sup>2</sup> was insufficiently low for obligate anaerobic bacteria germination and/or an inadequate EPR effect to trap the spores or bacteria. This phenomenon also likely occurred in our experiments.

#### *2.2. Transformation of Bifidobacterium with an Expression Vector for Cytosine Deaminase (CD) 2.2. Transformation of Bifidobacterium with an Expression Vector for Cytosine Deaminase (CD)*

Although we initially sought to transform the bacteria to produce anti-cancer molecules, no plasmid was available for *Bifidobacterium* in the 1980s. In 1997, however, an Although we initially sought to transform the bacteria to produce anti-cancer molecules, no plasmid was available for *Bifidobacterium* in the 1980s. In 1997, however, an expression vector developed by Kano's group [31] launched a series of collaborative trials for the creation of anti-cancer drugs by *Bifidobacterium*. We first inserted the CD gene of *E. coli*

into the vector. The CD enzyme can convert the low-toxic 5FC, a prodrug of 5FU, to the toxic anti-cancer drug 5FU. 5FC is a well-known drug for mycosis, with almost no systemic toxicity by oral administration. We transformed *Bifidobacterium* with the CD expression vector and began experiments on solid cancers using genetically engineered *Bifidobacterium* in combination with 5FC [32–35]. This was the first step towards our *i*DPS. expression vector and began experiments on solid cancers using genetically engineered *Bifidobacterium* in combination with 5FC [32–35]. This was the first step towards our *i*DPS. *2.3. Therapy Experiments on Solid Cancers Using Transformed Bifidobacterium with 5FC* The procedure for our cancer treatment with *Bifidobacterium* carrying the CD gene is

as follows [11,36]. First, we *i.v.* injected the transformed *Bifidobacterium* into tumor-bearing

expression vector developed by Kano's group [31] launched a series of collaborative trials for the creation of anti-cancer drugs by *Bifidobacterium*. We first inserted the CD gene of *E. coli* into the vector. The CD enzyme can convert the low-toxic 5FC, a prodrug of 5FU, to the toxic anti-cancer drug 5FU. 5FC is a well-known drug for mycosis, with almost no systemic toxicity by oral administration. We transformed *Bifidobacterium* with the CD

#### *2.3. Therapy Experiments on Solid Cancers Using Transformed Bifidobacterium with 5FC* animals. Several days later when the specific localization of the bacteria inside the tumor

*J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 5 of 21

The procedure for our cancer treatment with *Bifidobacterium* carrying the CD gene is as follows [11,36]. First, we *i.v.* injected the transformed *Bifidobacterium* into tumor-bearing animals. Several days later when the specific localization of the bacteria inside the tumor tissues was expected, we commenced oral 5FC administration to the animals. Although the prodrug spread throughout the body, it was converted to 5FU only in tumor tissues by *Bifidobacterium* expressing CD. We then checked for tumor growth suppression and systemic toxicity by 5FU. In the first experiment, we used an autochthonous DMBA-induced rat breast cancer system, and comparable results were obtained in various tumor-bearing animals. The selective localization of *Bifidobacterium* in tumors was confirmed by tissue homogenate cultures in vitro and Gram-positive staining of *Bifidobacterium* in the tumor tissues. CD expression in *Bifidobacterium* was ascertained by immunostaining with anti-CD antibodies. tissues was expected, we commenced oral 5FC administration to the animals. Although the prodrug spread throughout the body, it was converted to 5FU only in tumor tissues by *Bifidobacterium* expressing CD. We then checked for tumor growth suppression and systemic toxicity by 5FU. In the first experiment, we used an autochthonous DMBAinduced rat breast cancer system, and comparable results were obtained in various tumorbearing animals. The selective localization of *Bifidobacterium* in tumors was confirmed by tissue homogenate cultures in vitro and Gram-positive staining of *Bifidobacterium* in the tumor tissues. CD expression in *Bifidobacterium* was ascertained by immunostaining with anti-CD antibodies. The success of our therapy system, i.e., the suppression of tumor growth in chemically induced autochthonous rat breast cancer, can be seen in Figure 2a. When we

The success of our therapy system, i.e., the suppression of tumor growth in chemically induced autochthonous rat breast cancer, can be seen in Figure 2a. When we treated human breast carcinoma transplanted into immunodeficient nude mice, tumor suppression was again witnessed without systemic toxicity (Figure 2b). Indeed, relatively large 5FU production was detected exclusively in tumors and not in normal tissues (Figure 2c). An important point was that no apparent adverse effects were observed, which was also the case in dogs, monkeys, and other large animal tests. treated human breast carcinoma transplanted into immunodeficient nude mice, tumor suppression was again witnessed without systemic toxicity (Figure 2b). Indeed, relatively large 5FU production was detected exclusively in tumors and not in normal tissues (Figure 2c). An important point was that no apparent adverse effects were observed, which was also the case in dogs, monkeys, and other large animal tests.

**Figure 2.** *Cont*.

**Group Sensitized** 

**antigen**

0.05. These figures were adopted from a previous study [11,36].

**Figure 2.** Anti-tumor effects of i.v. injected cytosine deaminase of *Escherichia coli* (e-CD) transformed *Bifidobacterium longum* (*B. longum/e-CD*) combined with oral 5-fluorocytosine (5FC). (**a**) Comparison of tumor volumes of non-injected rats (*n =* 5) with those of *B. longum/e-CD i.v.*  injected rats (*n* = 15). Rats bearing 7,12-dimethylbenz(a)anthracene-induced mammary tumors received *i.v. B. longum/e-CD* and 500 mg/kg/day 5FC*.* \* *p <* 0.05; \*\* *p <* 0.01. (**b**) Anti-tumor assessment of *B. longum/e-CD* in nude mice transplanted with KPL-1 human mammary tumor cells. Tumor-bearing nude mice (*n =* 8) were given a dose of transformed bacteria cells *i.v.* (5.9 × 10<sup>9</sup> c.f.u./mouse), followed by oral 5FC for 21 days. (**c**) Measurement of 5-fluorouracil (5FU) concentration in various tissues in rats bearing MRMT-1 mammary gland carcinoma. Rats were given *B. longum/e-CD* at 1.1 × 10<sup>10</sup> c.f.u./rat *i.v.* and 5FC by intragastric gavage for 4 days starting on day 4 after bacterium injection. The concentration of 5FU in normal tissues and tumor tissues was measured. Rats given 5FC without injection of *B. longum/e-CD* were used as controls. \* *p <* **Figure 2.** Anti-tumor effects of i.v. injected cytosine deaminase of *Escherichia coli* (e-CD)-transformed *Bifidobacterium longum* (*B. longum/e-CD*) combined with oral 5-fluorocytosine (5FC). (**a**) Comparison of tumor volumes of non-injected rats (*n =* 5) with those of *B. longum/e-CD i.v.* injected rats (*n* = 15). Rats bearing 7,12-dimethylbenz(a)anthracene-induced mammary tumors received *i.v. B. longum/e-CD* and 500 mg/kg/day 5FC. \* *p <* 0.05; \*\* *p <* 0.01. (**b**) Anti-tumor assessment of *B. longum/e-CD* in nude mice transplanted with KPL-1 human mammary tumor cells. Tumor-bearing nude mice (*n =* 8) were given a dose of transformed bacteria cells *i.v.* (5.9 <sup>×</sup> <sup>10</sup><sup>9</sup> c.f.u./mouse), followed by oral 5FC for 21 days. (**c**) Measurement of 5-fluorouracil (5FU) concentration in various tissues in rats bearing MRMT-1 mammary gland carcinoma. Rats were given *B. longum/e-CD* at 1.1 <sup>×</sup> <sup>10</sup><sup>10</sup> c.f.u./rat *i.v.* and 5FC by intragastric gavage for 4 days starting on day 4 after bacterium injection. The concentration of 5FU in normal tissues and tumor tissues was measured. Rats given 5FC without injection of *B. longum/e-CD* were used as controls. \* *p <* 0.05. These figures were adopted from a previous study [11,36].

We later sought to increase enzyme activity by modifying the active site of CD according to Mahan's method [37]. Since the original substrate for CD is not 5FC, but rather cytosine, the enzymatic activity and affinity to 5FC was relatively low as compared with that to cytosine. When we changed the amino acid at position 314 of the active center of the enzyme from aspartic acid to alanine, the conversion rate of 5FC to 5FU was increased by approximately ten-fold. In clinical trials, the modified expression vector was further tailored by removing the resistance gene to an antibiotic, spectinomycin, to protect We later sought to increase enzyme activity by modifying the active site of CD according to Mahan's method [37]. Since the original substrate for CD is not 5FC, but rather cytosine, the enzymatic activity and affinity to 5FC was relatively low as compared with that to cytosine. When we changed the amino acid at position 314 of the active center of the enzyme from aspartic acid to alanine, the conversion rate of 5FC to 5FU was increased by approximately ten-fold. In clinical trials, the modified expression vector was further tailored by removing the resistance gene to an antibiotic, spectinomycin, to protect against horizontal transmission.

#### against horizontal transmission. *2.4. Immunological Safety*

*2.4. Immunological Safety* To test for immunological toxicity and possible severe anaphylaxis from repeated injection of the bacteria into animals, we evaluated for active systemic anaphylaxis (ASA) reactions and passive cutaneous anaphylaxis (PCA) caused by IgG with a sensitive guinea pig system (Table 1). In terms of ASA reactions, the positive control ovalbumin induced severe shock, whereas little, if any, reactions were seen for *Bifidobacterium*. Regarding PCA, although IgG antibody formation against *Bifidobacterium* had been suggested, To test for immunological toxicity and possible severe anaphylaxis from repeated injection of the bacteria into animals, we evaluated for active systemic anaphylaxis (ASA) reactions and passive cutaneous anaphylaxis (PCA) caused by IgG with a sensitive guinea pig system (Table 1). In terms of ASA reactions, the positive control ovalbumin induced severe shock, whereas little, if any, reactions were seen for *Bifidobacterium*. Regarding PCA, although IgG antibody formation against *Bifidobacterium* had been suggested, experiments using animals immunized with *Bifidobacterium* confirmed the safety and therapeutic efficiency of the system.

experiments using animals immunized with *Bifidobacterium* confirmed the safety and therapeutic efficiency of the system. **Table 1.** Antigenic tests to estimate the toxicity of *B. longum*/S-eCD. (**a**)**ASA reaction Cause antigen No.of animals Antigen challenge outcome** (-) (+/-) (+) (++) (+++) We further examined for the induction of inflammatory cytokines related to *Bifidobacterium* in collaboration with an expert of infectious immunity, Tsutsui, Hyogo College of Medicine in Japan. As shown in Figure 3, the *E. coli* control predictably induced the cytokines seen in sepsis, while *Bifidobacterium* did little [12]. Since cytokine production occurs through Toll-like receptors, which are the main players in innate immunity, those results suggested that *Bifidobacterium* was recognized neither by Toll-like receptors in vivo [38], nor by other innate immunity systems activating IL-1β through inflammasomes in the cytosol [39]. It is well known that inflammasomes are activated by the flagella of *Salmonella* to induce IL-1β.


**Table 1.** Antigenic tests to estimate the toxicity of *B. longum*/S-eCD.

Note: (a)Actively immunized guinea pigs were injected intravenously with *B. longum*/S-eCD or OVA 14 days after the final sensitization. Anaphylaxis symptoms were quantified by the following criteria: –, no symptoms; ±, scrub of face or ear and/or scratch of nose; +, coughing or locomotion ataxia; ++, convulsion or roll, but no death observed within 1 h; and +++, death observed within 1 h. (b) In the PCA reaction, immunized guinea pigs were killed and blood samples were collected 14days after final sensitization to obtain each antiserum. Normal guinea pigs were shaved, and 0.05 mL of each serum dilution was injected intradermally into the dorsal skin. After 4 h, the animals were injected *i.v*. with 1 mL of antigen (*B. longum/S-eCD* or OVA) and 0.5 mL of 1% Evans blue solution. After 30 min, the animals were killed, the dorsal skin was peeled off, and blue spots within the intradermal sites were measured. A PCA reaction was judged to be positive when the blue spot measured more than 5 mm<sup>2</sup> in dimension. This table was adopted from a previous study [36]. *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 8 of 21

**Figure 3.** Production of inflammatory cytokines in C57BL/6 mice injected *i.v.* with cytosine deaminase of *Escherichia coli* (e-CD) -transformed *Bifidobacterium longum* (*B*. *longum/e-CD*) or nonpathogenic *E. coli* (control). Cytokines were assessed by ELISA at 6 h after injection. Closed bars, blood of mice injected with *B. longum/e-CD*; dotted bars, blood of mice injected with nonpathogenic *E. coli*; open bars, normal blood. IFN, interferon; IL, interleukin; ND, not detected. This figure was adopted from a previous study [10,11]. **Figure 3.** Production of inflammatory cytokines in C57BL/6 mice injected *i.v.* with cytosine deaminase of *Escherichia coli* (e-CD) -transformed *Bifidobacterium longum* (*B*. *longum/e-CD*) or non-pathogenic *E. coli* (control). Cytokines were assessed by ELISA at 6 h after injection. Closed bars, blood of mice injected with *B. longum/e-CD*; dotted bars, blood of mice injected with non-pathogenic *E. coli*; open bars, normal blood. IFN, interferon; IL, interleukin; ND, not detected. This figure was adopted from a previous study [10,11].

Concerning the above findings, an interesting report found that extracellular vesicles in the blood inhibited the induction of NFB expression by *Bifidobacterium* [40]. *NFκB* is a well-known master gene for inflammatory cytokines. When *Bifidobacterium* was allowed to act on cultured human embryonic kidney cells, NF*κ*B was induced in the absence of Concerning the above findings, an interesting report found that extracellular vesicles in the blood inhibited the induction of NF*κ*B expression by *Bifidobacterium* [40]. *NFκB* is a well-known master gene for inflammatory cytokines. When *Bifidobacterium* was allowed

any, little of inflammatory cytokine induction in mice by *Bifidobacterium* was partially

In the field of probiotic research, human intestinal flora changes have been examined for relationships between flora variety and health conditions. Shortly after birth*, Bifidobacterium* becomes the main gut flora and coexists with the organism throughout life at gradually decreasing amounts [41,42]. It is likely that humans and other mammals have some immune tolerance against *Bifidobacterium*. These facts strengthen our notion that *Bifidobacterium* can be safely used for an *i*DPS to produce anti-cancer molecules in humans. Moreover, *Bifidobacterium* has been included in commensal microbiome work to enhance

For the purpose of applying *i*DPS/5FC for clinical cancer treatment in humans, rigorous testing was performed according to Chemistry, Manufacturing and Control (CMC) and Good Manufacturing Practice (GMP) guidelines. CMC requires showing the physicochemical properties of the product in detail, while GMP necessitates the provision of quality assurance that products are consistently produced and controlled to quality standards. Since CMC and GMP are generally difficult processes even with simple chemical compounds, certification for living bacteria has been much more challenging.

B expression in a

the cancer therapy efficiency of anti-PD-1 antibodies [43].

attributed to extracellular vesicles in the blood.

*2.5. Translational Research*

serum. However, serum addition to the medium suppressed NF

to act on cultured human embryonic kidney cells, NF*κ*B was induced in the absence of serum. However, serum addition to the medium suppressed NF*κ*B expression in a concentration-dependent manner. It was also shown that extracellular vesicles in the serum played a key role in suppressing NF*κ*B. Those results suggested that the lack, if any, little of inflammatory cytokine induction in mice by *Bifidobacterium* was partially attributed to extracellular vesicles in the blood.

In the field of probiotic research, human intestinal flora changes have been examined for relationships between flora variety and health conditions. Shortly after birth, *Bifidobacterium* becomes the main gut flora and coexists with the organism throughout life at gradually decreasing amounts [41,42]. It is likely that humans and other mammals have some immune tolerance against *Bifidobacterium*. These facts strengthen our notion that *Bifidobacterium* can be safely used for an *i*DPS to produce anti-cancer molecules in humans. Moreover, *Bifidobacterium* has been included in commensal microbiome work to enhance the cancer therapy efficiency of anti-PD-1 antibodies [43].

#### *2.5. Translational Research*

For the purpose of applying *i*DPS/5FC for clinical cancer treatment in humans, rigorous testing was performed according to Chemistry, Manufacturing and Control (CMC) and Good Manufacturing Practice (GMP) guidelines. CMC requires showing the physicochemical properties of the product in detail, while GMP necessitates the provision of quality assurance that products are consistently produced and controlled to quality standards. Since CMC and GMP are generally difficult processes even with simple chemical compounds, certification for living bacteria has been much more challenging.

We have performed the precise characterization of various aspects of *Bifidobacterium*, including its membrane composition, stability of the expression vector in which the antibiotic resistance gene for selective pressure and other sequences were removed for safety, and bacterial survival measurement methods. Every attempt has been made to establish virus-free and sterile preparations for GMP. Finally, through investigational new drug discussion, our proposed first-in-man clinical trial was approved by the American FDA, with NIH Recombinant DNA Advisory Committee approval of our protocol on biosafety. The first-in-man phase 1 and 2 tests were approved in 2013 and carried out sponsored by a bio-venture company, Anaeropharma Science Inc., Tokyo, Japan. Whereas the phase 1 test is almost completed, phase 2 has regrettably been postponed by the current COVID-19 pandemic.

The concept for applying the *i*DPS using genetically modified *Bifidobacterium* on human cancer therapy is much the same as that with animals. Since the line of CDexpressing *Bifidobacterium* for application on humans is named APS001F, the clinical trial has been entitled "A Phase I/II Safety, Pharmacokinetic, and Pharmacodynamic Study of APS001F with Flucytosine (5FC) and Maltose for the Treatment of Advanced and/or Metastatic Solid Tumors" [44].

#### *2.6. Combination Therapy of APS001F Plus 5FC (APS001F/5FC) in Combination with the Immune Checkpoint Inhibitor (ICPI) Anti-PD-1*

With recent developments in cancer treatment, the prominent anti-tumor effects of ICPIs, such as anti-CTLA4 and anti-PD-1 antibodies, have been demonstrated worldwide [45–47]. To augment the effects of ICPIs, combination treatments with chemotherapy, molecular targeting anti-cancer agents, and/or radiation therapy have been tested. However, the associated side effects often increased in tandem with anti-tumor action. We expected our *i*DPS using *Bifidobacterium* to enhance ICPI treatment by 5FU in tumors without raising side effects by enhancing immune reactions through innate immunity locally stimulated by *Bifidobacterium*.

Combination therapy experiments of the *i*DPS with APS001F/5FC and anti-mPD-1 antibodies have already yielded promising results [48]. The almost completed clinical phase 1 test of APS001F/5FC also serves for prechecking whether the combination of anti-PD-1 antibodies, which are already available for clinical use, with APS001F/5FC can be a potential treatment candidate. The rationality for this next step was made through investigation of the literature [49–52] as follows: (1) combining ICPIs and anti-cancer drugs, including 5FU, reportedly enhances anti-tumor effects, (2) 5FU has the potential to suppress myeloid-derived suppressor cell inhibition of anti-tumor immune reactions, and (3) the systemic administration of 5FU at a high dose rather impairs anti-tumor immunity. The third consideration may be attributed to the systemic toxicity of 5FU, which is improved by the *i*DPS with APS001F/5FC.

Combining APS001F/5FC with anti-PD-1 antibodies in therapeutic experiments enhanced treatment effects (Figure 4a) without increasing side effects. We first observed that the tumor growth in animals treated with anti-PD-1 antibodies was slightly suppressed. When combined with APS001F/5FC, however, this effect was significantly augmented. Animal survival was also prolonged, including a complete remission case [48].

By analyzing the immune cells in tumors during combination therapy, we witnessed a remarkable decrease in regulatory T (Treg) cells, which inhibit anti-tumor immune activity, and consequently the ratio of CD8 cells to Treg cells was greatly increased (Figure 4b) through a yet unspecified mechanism. In this therapy system, 5FU was produced in situ in tumor tissues and Treg cells were suppressed without a change in CD8 activity, likely raising the anti-tumor effect of combination therapy. We strongly believe that APS001F/5FC in future clinical trials will exert promising effects in combination with anti-PD-1 antibodies and other ICPIs. Furthermore, combined treatment with anti-tumor drugs and ICPIs may be possible with a single gene-engineered *Bifidobacterium* clone. Although technically challenging, we have already succeeded in establishing such a co-expression system with *Bifidobacterium. J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 10 of 21

**Figure 4.** *Cont*.

(**b**)

significantly different. This figure was adopted from a previous study [48].

**Figure 4.** Anti-tumor effects of APS001F/5FC, anti-PD-1 mAb, and their combination were

evaluated in a syngeneic CT26 mouse model. Experiment schedule. Mice were *s.c.* inoculated with 1 × 105 CT26 cells (day -13), and stratification was done on day 0. APS001F was administered *i.v.* at 1.0 × 109 c.f.u./mouse via the tail vein on day 3 with 200 mg maltose supplementation (days 3 through 16). 5-FC was administered *i.p.* at 250 mg/kg twice a day (500 mg/kg/day) on days 5 through 9 and days 12 through 16. Anti-PD-1 mAb was administered *i.p.* at 200 µg/mouse on days 1, 4, 8, and 11. Tumor volume was measured twice a week. (**a**) Change in tumor volume after treatment with APS001F/5-FC, anti-PD-1 mAb, and their combination. Results are the mean ± standard error of the mean of 9 mice. Analyses of 4 groups on day 21 were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not significantly different. Tumor Growth Inhibition was calculated on day 21. (**b**) Flow cytometric analysis of tumor cells in mice engrafted with CT26 cells. Results are the mean ± standard deviation of 8 mice. %CD45+ of total cells, %CD4+ of total cells, %Tregs (CD45+ CD4+ CD25+ Foxp3+ cells) of CD4+ cells, %CD8+ of total cells, CD8/Treg ratio, and %neutrophils (CD45+ CD11b+ Ly-6G+ cells) of total cells, and %TAMs (CD45+ CD11b+ Ly-6G−Ly-6Clow cells) of total cells were analyzed. Analyses of 4 groups were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not

(**a**)

**Figure 4.** Anti-tumor effects of APS001F/5FC, anti-PD-1 mAb, and their combination were evaluated in a syngeneic CT26 mouse model. Experiment schedule. Mice were *s.c.* inoculated with 1 × 105 CT26 cells (day -13), and stratification was done on day 0. APS001F was administered *i.v.* at 1.0 × 109 c.f.u./mouse via the tail vein on day 3 with 200 mg maltose supplementation (days 3 through 16). 5-FC was administered *i.p.* at 250 mg/kg twice a day (500 mg/kg/day) on days 5 through 9 and days 12 through 16. Anti-PD-1 mAb was administered *i.p.* at 200 µg/mouse on days 1, 4, 8, and 11. Tumor volume was measured twice a week. (**a**) Change in tumor volume after treatment with APS001F/5-FC, anti-PD-1 mAb, and their combination. Results are the mean ± standard error of the mean of 9 mice. Analyses of 4 groups on day 21 were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not significantly different. Tumor Growth Inhibition was calculated on day 21. (**b**) Flow cytometric analysis of tumor cells in mice engrafted with CT26 cells. Results are the mean ± standard deviation of 8 mice. %CD45+ of total cells, %CD4+ of total cells, %Tregs (CD45+ CD4+ CD25+ Foxp3+ cells) of CD4+ cells, %CD8+ of total cells, CD8/Treg ratio, and %neutrophils (CD45+ CD11b+ Ly-6G+ cells) of total cells, and %TAMs (CD45+ CD11b+ Ly-6G−Ly-6Clow cells) of total cells were analyzed. Analyses of 4 groups were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not significantly different. This figure was adopted from a previous study [48]. **Figure 4.** Anti-tumor effects of APS001F/5FC, anti-PD-1 mAb, and their combination were evaluated in a syngeneic CT26 mouse model. Experiment schedule. Mice were *s.c.* inoculated with 1 <sup>×</sup> <sup>10</sup><sup>5</sup> CT26 cells (day -13), and stratification was done on day 0. APS001F was administered *i.v.* at 1.0 <sup>×</sup> <sup>10</sup><sup>9</sup> c.f.u./mouse via the tail vein on day 3 with 200 mg maltose supplementation (days 3 through 16). 5-FC was administered *i.p.* at 250 mg/kg twice a day (500 mg/kg/day) on days 5 through 9 and days 12 through 16. Anti-PD-1 mAb was administered *i.p.* at 200 µg/mouse on days 1, 4, 8, and 11. Tumor volume was measured twice a week. (**a**) Change in tumor volume after treatment with APS001F/5-FC, anti-PD-1 mAb, and their combination. Results are the mean ± standard error of the mean of 9 mice. Analyses of 4 groups on day 21 were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not significantly different. Tumor Growth Inhibition was calculated on day 21. (**b**) Flow cytometric analysis of tumor cells in mice engrafted with CT26 cells. Results are the mean <sup>±</sup> standard deviation of 8 mice. %CD45<sup>+</sup> of total cells, %CD4<sup>+</sup> of total cells, %Tregs (CD45<sup>+</sup> CD4<sup>+</sup> CD25<sup>+</sup> Foxp3<sup>+</sup> cells) of CD4<sup>+</sup> cells, %CD8<sup>+</sup> of total cells, CD8/Treg ratio, and %neutrophils (CD45<sup>+</sup> CD11b<sup>+</sup> Ly-6G<sup>+</sup> cells) of total cells, and %TAMs (CD45<sup>+</sup> CD11b<sup>+</sup> Ly-6G−Ly-6Clow cells) of total cells were analyzed. Analyses of 4 groups were conducted using Tukey's method of multiple comparisons. Means sharing a letter are not significantly different. This figure was adopted from a previous study [48].

#### *2.7. Establishment of a Protein-Secreting System*

We are underway to engineer *Bifidobacterium* that not only express, but also secrete, proteins such as anti-tumor antibodies and cytokines (Scheme 1). In spite of the prominent effects of ICPIs, there remain problems including autoimmune diseases and other severe side effects. Immune checkpoint molecules, such as CTLA4 and PD-1, inactivate T-cell killing to terminate excessive inflammatory reactions in the body. It is physiologically important to halt unnecessary immunoreactions, and blocking those molecules tends to induce immune toxicities in the host [53–55].

To improve this situation, it will be useful to establish an *i*DPS for immune checkpointmodifying antibodies, including anti-CTLA4 and anti-PD-1 antibodies, and immune modifiers, such as anti-tumor cytokines, with *Bifidobacterium.* We have therefore been attempting to establish *Bifidobacterium* that both express and secrete immunological anti-cancer molecules (Scheme 1), such as anti-CTLA4 and/or anti-PD-1 antibodies, in addition to such immune-stimulating anti-tumor cytokines as TNFα and INFγ.

expression system with *Bifidobacterium.*

*2.7. Establishment of a Protein-Secreting System* 

induce immune toxicities in the host [53–55].

**Scheme 1.** The in situ delivery and production system (*i*DPS) as a platform technology for producing various anti-tumor scFvs and cytokines. **Scheme 1.** The in situ delivery and production system (*i*DPS) as a platform technology for producing various anti-tumor scFvs and cytokines.

#### 2.7.1. Anti-HER2 scFv

As it is generally difficult to produce the original structure of antibodies with bacteria, we firstly tried to create single-chain variable fragments (scFvs), which were fusion proteins of variable light and heavy chains connected with a linker for each antibody, such that the scFv could be called a single-chain antibody.

By analyzing the immune cells in tumors during combination therapy, we witnessed a remarkable decrease in regulatory T (Treg) cells, which inhibit anti-tumor immune activity, and consequently the ratio of CD8 cells to Treg cells was greatly increased (Figure 4b) through a yet unspecified mechanism. In this therapy system, 5FU was produced in situ in tumor tissues and Treg cells were suppressed without a change in CD8 activity, likely raising the anti-tumor effect of combination therapy. We strongly believe that APS001F/5FC in future clinical trials will exert promising effects in combination with anti-PD-1 antibodies and other ICPIs. Furthermore, combined treatment with anti-tumor drugs and ICPIs may be possible with a single gene-engineered *Bifidobacterium* clone. Although technically challenging, we have already succeeded in establishing such a co-

We are underway to engineer *Bifidobacterium* that not only express, but also secrete, proteins such as anti-tumor antibodies and cytokines (Scheme 1). In spite of the prominent effects of ICPIs, there remain problems including autoimmune diseases and other severe side effects. Immune checkpoint molecules, such as CTLA4 and PD-1, inactivate T-cell killing to terminate excessive inflammatory reactions in the body. It is physiologically important to halt unnecessary immunoreactions, and blocking those molecules tends to

To improve this situation, it will be useful to establish an *i*DPS for immune checkpoint-modifying antibodies, including anti-CTLA4 and anti-PD-1 antibodies, and immune modifiers, such as anti-tumor cytokines, with *Bifidobacterium.* We have therefore been attempting to establish *Bifidobacterium* that both express and secrete immunological anti-cancer molecules (Scheme 1), such as anti-CTLA4 and/or anti-PD-1 antibodies, in

addition to such immune-stimulating anti-tumor cytokines as TNFα and INFγ.

In our attempts to produce scFvs for immune checkpoint molecules, we started by expressing and secreting a biologically active scFv already made and confirmed by another system. For this purpose, we turned to an anti-HER2 antibody, trastuzumab, since a biologically active scFv for trastuzumab had been developed by Akiyama at the Shizuoka Cancer Center in Japan. Trastuzumab is well known and widely used as a molecular targeting antibody for human breast cancer, but occasionally induces cardiotoxicities [56,57]. Thus, we sought to establish *Bifidobacterium* secretion of the biologically active scFv for anti-HER2 antibodies.

We succeeded in making not only an expression, but also a secretion, system for a biologically active scFv derived from the anti-HER2 trastuzumab with *Bifidobacterium* (Figure 5) [58]. Production of the scFv to human HER2 was confirmed by Western blot analysis. We also verified biochemical activity by FACS and immunological staining [58]. The genetically modified bacteria were injected into nude mice bearing the human HER2 positive breast cancer, NCI-N87. We witnessed selective localization of the bacteria inside the tumor, secretion of anti-HER2 single-chain antibodies, and ultimately a suppressive effect on tumor growth (Figure 5) [58]. This success in creating *Bifidobacterium* to express and secrete the biologically active trastuzumab scFv prompted us to establish *i*DPS with *Bifidobacterium* for ICPIs.

**Figure 5.** (**a**) Structure of the pH1 and pH2 plasmids. (**b**) Molecular size of the trastuzumab scFV produced by *Bifidobacterium*. Results for H1 scFv are shown. Regarding H2, identically sized scFv was confirmed at a markedly higher expression amount. (**c**) FACS analysis. SK-MEL-28 (HER2−), BT-474 (HER2 +/−), and SK-BR-3 (HER2+) cells were stained with His-tag-purified trastuzumab scFv. Blue line: control (buffer alone). Red line: stained with trastuzumab scFv from H2. (**d**) Immunostaining of cultured cells by His-tag-purified trastuzumab scFv from *B. longum* H2. Immunofluorescent staining. Blue: nucleus. Green: stained with trastuzumab scFv from H2. Right panels: stained with trastuzumab scFv. Left panels: negative control (without trastuzumab scFv). Original magnification of all images was x400. (**e**) Growth suppression of a human HER2(+) carcinoma transplanted into nude mice by recombinant *Bifidobacterium* H2. *B. longum* mock and H2 were *i.v.* administered to NCI-N87 human gastric cancer tumor-bearing mice twice a week. Mean ± standard deviation values of 8 mice. \*: *p* < 0.05 versus non-treated group, #: *p* < 0.05 versus mock-treated group. This figure was adopted from a previous study [58]. **Figure 5.** (**a**) Structure of the pH1 and pH2 plasmids. (**b**) Molecular size of the trastuzumab scFV produced by *Bifidobacterium*. Results for H1 scFv are shown. Regarding H2, identically sized scFv was confirmed at a markedly higher expression amount. (**c**) FACS analysis. SK-MEL-28 (HER2−), BT-474 (HER2 +/−), and SK-BR-3 (HER2+) cells were stained with His-tag-purified trastuzumab scFv. Blue line: control (buffer alone). Red line: stained with trastuzumab scFv from H2. (**d**) Immunostaining of cultured cells by His-tag-purified trastuzumab scFv from *B. longum* H2. Immunofluorescent staining. Blue: nucleus. Green: stained with trastuzumab scFv from H2. Right panels: stained with trastuzumab scFv. Left panels: negative control (without trastuzumab scFv). Original magnification of all images was x400. (**e**) Growth suppression of a human HER2(+) carcinoma transplanted into nude mice by recombinant *Bifidobacterium* H2. *B. longum* mock and H2 were *i.v.* administered to NCI-N87 human gastric cancer tumor-bearing mice twice a week. Mean ± standard deviation values of 8 mice. \*: *p* < 0.05 versus non-treated group, #: *p* < 0.05 versus mock-treated group. This figure was adopted from a previous study [58].

CTLA4, Anti-41BB, and Anti-Tumor Cytokines We next established *Bifidobacterium* to secrete scFvs of anti-PD-1 [59], anti-CTLA4 2.7.2. scFvs for ICP Antagonistic or Agonistic Antibodies, including Anti-PD-1, Anti-CTLA4, Anti-41BB, and Anti-Tumor Cytokines

[60], and anti-41BB (an immune checkpoint agonist) [61] antibodies. All scFvs produced by the *i*DPS were detected exclusively in tumor tissues and exhibited immunological activity and anti-tumor effects without remarkable side effects. In addition to scFvs, we have also established *Bifidobacterium* expressing and We next established *Bifidobacterium* to secrete scFvs of anti-PD-1 [59], anti-CTLA4 [60], and anti-41BB (an immune checkpoint agonist) [61] antibodies. All scFvs produced by the *i*DPS were detected exclusively in tumor tissues and exhibited immunological activity and anti-tumor effects without remarkable side effects.

secreting INFγ and/or TNFα, which displayed notable anti-tumor effects [62,63]. It is well

2.7.2. scFvs for ICP Antagonistic or Agonistic Antibodies, including Anti-PD-1, Anti-

In addition to scFvs, we have also established *Bifidobacterium* expressing and secreting INFγ and/or TNFα, which displayed notable anti-tumor effects [62,63]. It is well known that the clinical application of INFγ and TNFα is difficult due to their severe systemic toxicity in spite of strong anti-tumor effects [64]. However, this was not the case in our *i*DPS using *Bifidobacterium*. When we systemically *i.v.* injected *Bifidobacterium* that could secrete INFγ and/or TNFα into tumor-bearing animals, high amounts of cytokines were detected in tumors, with little in the blood and thus no systemic toxicity. In addition to the anti-tumor properties of *Bifidobacterium* expressing and secreting the cytokines, the enhancement of anti-tumor effects by combination with ICPI antibodies or a popular anticancer drug, Adriamycin, was observed as well. Taken together, drugs that have been unsuitable for clinical use due to strong systemic toxicity despite formidable anti-cancer properties may be revisited through the use of our *i*DPS. Most recently, a *Bifidobacterium* clone, APS002, has been established to secrete diabodies against EGFR/HER3 and CD3 and redirect T cells to EGFR/HER3-positive cancer cells. This clone inhibited the growth of human EGFR-positive cancer cells transplanted into humanized immunodeficient mice [65], indicating a possible clinical application of this engineered *Bifidobacterium*.

#### **3. For Further Improvement of the** *i***DPS**

#### *3.1. Notes for the Presnt iDPS*

## 3.1.1. Animal Experiments

The established methods for *i*DPS so far are detailed in our previous papers, especially in [11] and patent information [63]. In animal experiments with *Bifidobacteria*, we used chemically induced autochthonous tumor system, mouse tumor model, and allogenic transplanted human tumors in immunodeficient mice (Figures 1 and 2). Generally, autochthonous cancer is relatively difficult to cure compared with a transplanted tumor, because the tumor is comprised of cells which have escaped from host immune surveillance. In this sense, it is thought to mimic the human cancer system. In transplantation of cancer cells, the number of cells should be as small as possible to mimic a human tumor system where one nodule is produced from single or a few cancer cells. In such systems, we have experienced that *Bifidobacteria* tended to be localized even in small tumors. While *Bifidobacterium* could be safely administered to immunodeficient mice, immunodeficient mice are always infected with various bacteria as compared with normal mice because they are immunodeficient. Thus, during the assay of the number of *Bifidobacteria* in various tissues, it was required to remove intrinsically contaminated germs in in vitro bacterial culture system. Our genetically engineered *Bifidobacteria* have 5FU resistance in addition to spectinomycin, so that we were able to eliminate such germs and to identify the colony of *Bifidobacterium* by using both drugs in vitro bacterial culture.

In the assay for inflammatory cytokines, there was little or no induction of such cytokines in our mouse system (Figure 3), however, it probably depends on the type of animal, including human. More detailed analysis is needed from the viewpoint of molecular immunology.

In our present chemotherapy system (Figure 2), 5FC have been used as the prodrug. Since a *Bifidobacterium* clone expressing β-glucuronidase has also been established to activate prodrugs inactivated by glucuronic acid conjugation. Such clone will be useful to reuse drugs that have strong anti-tumor properties but severe systemic side effects.

In the protein secretory system (Figure 5), the optimum secretory signal peptide depended on the secretory protein. We have searched various secretory signals of the *Bifidobacteriu*m's own secretory proteins and adequate promoter for the gene as well. In order to improve combination therapy, we have established various vectors to coexpress/secrete anti-tumor cytokines and/or scFvs for various antitumor antibodies, which will be useful for combination therapies of cancer in the future.

At the preclinical level, there are some studies using *Bifidobacteria,* though there seems to be only our group that has advanced to clinical trials. One of the preclinical studies similar to ours demonstrated the successful delivery and efficient expression of Tumstatin, a powerful angiostatin, with genetically engineered *Bifidobacterium*, leading to antitumor effects through inducing apoptosis of tumorous vascular endothelial cells [66]. They observed that *Bifidobacterium longum*, selectively localizes to and proliferates in the hypoxia location within solid tumor. The other investigated the therapeutic effect of new recombinant *Bifidobacterium breve* strain expressing interleukin (IL)-24 gene on head and neck tumor xenograft in mice and reported that new recombinant bacterium has the capability of targeting tumor tissue *in vivo* [67]. These reports are consistent with our results in terms of selective localization of *Bifidobacterium* in tumors.

#### 3.1.2. As for Safety of *i*DPS with *Bifidobacterium*

For the clinical trial, bacteria dosage was initiated at less than a NOAEL (no observed adverse effect level) in the most sensitive dogs, and the dose was gradually increased with the minimum homing dose for rat transplanted cancer as a guide [44].

Since the *Bifidobacterium* we use is derived from enterobacteria in human, which is also used as intestinal regulators and yogurt as a probiotics, and thus it has been considered to be a priori non-toxic from experience. However, to evaluate the toxicity of genetically modified *B. longum*, a number of preclinical studies have been also carried out in several animal species, including normal mice, nude mice, normal rats, nude rats, dogs and monkeys. Both pharmacological and preliminary general toxicity studies were done, none of which revealed serious unfavorable toxicities [10]. Various antibiotics were examined to eliminate excess *Bifidobacteria* after treatment, and many antibiotics were found to be effective. In particular, we have confirmed that it can be easily removed with commonly used penicillin antibiotics (data, not shown).

#### 3.1.3. The Specific Advantages of Using Bifidobacterium for Our *i*DPS

The specific advantages of using *Bifidobacterium* and the reason why we have been using *Bifidobacterium longum* for our *i*DPS are as follows: (1) It is an obligate anaerobic bacterium, so that it can discriminate hypoxic malignant tumor tissues for the colony formation from normal tissues. (2) It does neither produce toxic substances, nor has flagella which activate inflammasome to induce IL-1β. (3) It has been generally regarded as a good bacterium derived from human intestinal bacteria, so that it is easy to think about safety a priori even if it is administered into the blood, thus leading to a sense of security for the recipient. In addition, (4) it has also been shown to work positively in treatment with antitumor immune check point inhibitor, anti-PD-1 antibody.

Since an expression vector has become available firstly for *Bifidobacterium longum*, this *Bifidobacterium longum* has been used as a tool for *i*DPS and we came to apply for IND with *longum*. However, if safety is ensured by other probiotic species and the use of expression vectors becomes possible for them, they could become a better tool, so that it seems important to continue to search such species in the future.

#### *3.2. Seeking an Ideal Micro-Factory with Guaranteed Safety*

Nowadays, bacteria can be modified to endow new phenotypes using gene engineering [18,23]. It will be important to improve the efficiency of *i*DPSs by modifying *Bifidobacterium* to develop ideal delivery and in situ production systems. The success and safety of *i.v.* administration of APS001F with living *Bifidobacterium* in our clinical phase 1 trial was an encouraging first step. Based on the clinically confirmed safety of systemic *Bifidobacterium* injection, we will next modify *i*DPS setups to produce an ideal micro-factory of various anti-cancer molecules for selectively and continuously treating all types of solid cancer.

To strengthen the concept of *i*DPS tolerance in clinical applications, its molecular safety mechanism needs deeper understanding from the viewpoint of the innate and acquired immunological reactions to *i.v.* injected *Bifidobacterium.* We believe that we can ameliorate our system towards completely and safely eradicating cancer.

There are several issues to consider when improving and strengthening *i*DPS microfactories in tumors. First, it will be necessary to increase the number of bacteria in tumors to provide clear therapeutic effects. One way is the simple quantitative approach of increasing the inoculation size as we have not yet determined the tolerable maximal dose. Concerning qualitative modifications, it will be critical to consider two main factors: (1) The dynamics of tumoral blood flow preventing bacteria entrance into the tumor, and (2) The capture of bacteria by reticuloendothelial cells and neutrophils that rapidly reduces bacterial density in the blood.

#### 3.2.1. Modification of Tumor Hemodynamics with Vasodilators to Enhance the EPR Effect

In our research, the number of bacteria detected in each tumor-bearing animal varied and depended on the tumor type. In order to consistently obtain a large number of bacteria in any type of tumor, we will need to consider modifying the hemodynamics of lesions by directing attention to the EPR effect [9].

Inside the tumor, it is well known that blood vessels occasionally become blocked by poor blood flow dynamics, which may hinder the entrance of *Bifidobacterium* and other macromolecules. In future clinical trials, transiently exposing tumor blood vessels to vasodilators, such as nitroglycerin, may contribute to improved *i*DPSs. Since nitroglycerin reportedly augments the anti-tumor effects of chemotherapies [68–71], it seems logical to use this agent to enhance the accumulation of *Bifidobacterium* in tumors by increasing the EPR effect. Maeda's group targeted cancers with lactobacillus in animal experiments in combination with nitroglycerin and showed that the number of bacteria localized in tumors increased by ten-fold versus controls [30].

As an additional factor related to poor blood flow in tumors, we may have to consider thrombus inhibition of the intra-tumoral accessibility of anti-cancer drugs. Thrombosis can occur in cancer patients [72–74], in whom blood clots may form easily in tumor tissues [75]. The local administration of recombinant plasminogen activators may be effective to dissolve such clots [76]. If bacteria are equipped with such an enzyme by gene-engineering, their accumulation in tumors will likely become enhanced.

Another factor to potentially improve the bacterial accumulation is to make the bacteria smaller. In a liposome study of pancreatic cancer, smaller liposomes could deliver greater amounts of anti-cancer drugs to the lesion and augment therapeutic effects [77]. Tunability of bacterial size has been investigated [78]. Interestingly, it was reported that deletion of the *Bacillus subtilis ponA* gene encoding for PBP1 (a class A penicillin-binding protein), a bifunctional peptidoglycan synthase, led to thinner cells [79], indicating that it may be possible to change the size of the bacteria by gene engineering.

3.2.2. Transient Evasion from Bacteria Trapping by the Reticuloendothelial System (RES) and/or Neutrophils

Clinical trials using bacteria for cancer treatment have found that the amount of bacteria accumulation in tumors appears to be less than expected based on the results of animal experiments [23].

One reason may be a more rapid capture of injected bacteria by the RES and/or neutrophils than in animal trials. A similar phenomenon is seen for liposome-type drugs. The RES in the human liver and spleen is a major obstacle to the tumor delivery of macromolecular drugs and liposomes [80,81], the effect of which seems to be stronger than in animal systems. Therefore, higher doses may be needed to achieve satisfactory therapeutic effects in humans along with a method to temporarily avoid trapping by the RES.

In one study, covering liposomes with polyethylene glycol (PEG) by so-called PEGylation enabled a breakthrough in the field of liposomes to avoid RES trapping. PEGylation has been also attempted at the cellular level with promising results [82,83]. Additional trials may bring about the same and/or improved effects as PEGylation. In that way, the removal of membrane molecules on bacteria responsible for phagocytosis by the RES and/or neutrophils by gene-targeting [84] will help avoid trapping by the RES. Furthermore, the

genes encoding the antigens recognized by reticulocyte-endothelial cells and neutrophils can be replaced with ones produce anti-cancer molecules. As a result, the bacteria will evade detection by the RES and neutrophils, leading not only to an increase in bacterial number entering the tumor microcirculation, but also enabling the bacteria to more stably express anti-cancer molecules.

In addition to genome editing, a sophisticated method to control plasmid copy number has been reported [85], which will be a powerful tool for enhancing bacterial cancer therapy in the future.

#### 3.2.3. Other Factors for Improving the *i*DPS

Regarding other factors contributing to the improvement of *i*DPSs, bacterial proteolytic activity [86] might be able to widen the localization area of bacteria, thus also being effective to spread anti-cancer substances produced by bacterial micro-factories to the whole tumor region from the central necrosis and/or periphery of the necrotic region where the anaerobic bacteria are colonized. For better distribution of anti-cancer substances in tumor tissue, there may be other ways to conjugate anti-tumor substances with oligopeptides to penetrate tumor masses and consequently widen the diffusion area of therapeutic agents [87].

It is also possible that specific and effective energy sources or nutrients that selectively stimulate the growth of bacteria in solid cancers can increase the number of intra-tumoral bacteria, even if the initial localization number is small. In the case of *Bifidobacterium*, lactulose was firstly used to enhance bacterial number in animal experiments. Lactulose is a disaccharide made up of galactose and fructose that cannot be used by mammalian cells as an energy source. Although considered an ideal bacterial energy source, the *i.v.* administration of lactulose is not permitted in the clinical setting. Therefore, maltose has been used in phase 1 testing as an alternative energy source. The search for an ideal nutrient in clinical trials continues for *Bifidobacterium* [88]. If lactic acid can be used as an energy source, it will be abundantly supplied by the tumor as a metabolite of cancer cell glycolysis. Other bacteria, such as *Veillonella*, consume lactic acid as reported in a meta-omics analysis of elite athletes as a performance-enhancing microbe that functions via lactate metabolism [89]. Transferring this metabolic phenotype to *Bifidobacterium* through gene transfer techniques may create a better micro-factory for anti-cancer drug production.

#### **4. Conclusions**

The present review of our novel *i*DPS with *Bifidobacterium* demonstrates its strong potential to safely improve the current problems of solid cancer treatment. Further advances will lower medical expenses through the continuous production of anti-cancer agents and cost-effectively reintroduce discontinued drugs that have strong anti-tumor properties but severe systemic side effects.

With a concerted global effort, it will be possible to pursue and realize an ideal bacterial-based system, no matter what difficulties await; all that is needed is to remove the unnecessary and undesirable genes and replace them with beneficial ones. Our novel *i*DPS with *Bifidobacterium* represents a promising therapeutic candidate for solid tumors as an in situ self-propagating micro-factory.

**Funding:** These works were supported by grants from the Japan Science and Technology Agency and the New Energy and Industrial Technology Development Organization, Collaborative Research Funds from Anaeropharma Science Inc. to Shinshu University, a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science, and an award from the Kobayashi Foundation for Cancer Research.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** These works were done by graduate students and colleagues at Shinshu University School of Medicine and Anaeropharma Science Inc. I thank all of them for their technical support and useful discussions, and N. Kimura and the late T. Baba for their guidance when I was a graduate student at Kyushu University. Since autumn 2020, research concerning the *i*DPS has been transferred to Azusapharma Science Inc.

**Conflicts of Interest:** As a disclosure statement for conflicts of interest, S. Taniguchi is a former Science Advisor of Anaeropharma Science Inc.

## **References**


**Fatemah Bahman <sup>1</sup> , Valeria Pittalà 2,\* , Mohamed Haider 3,4 and Khaled Greish 5,\***


**Abstract:** Triple negative breast cancer (TNBC) is the most aggressive breast cancer accounting for around 15% of identified breast cancer cases. TNBC lacks human epidermal growth factor receptor 2 (HER2) amplification, is hormone independent estrogen (ER) and progesterone receptors (PR) negative, and is not reactive to current targeted therapies. Existing treatment relies on chemotherapeutic treatment, but in spite of an initial response to chemotherapy, the inception of resistance and relapse is unfortunately common. Dasatinib is an approved second-generation inhibitor of multiple tyrosine kinases, and literature data strongly support its use in the management of TNBC. However, dasatinib binds to plasma proteins and undergoes extensive metabolism through oxidation and conjugation. To protect dasatinib from fast pharmacokinetic degradation and to prolong its activity, it was encapsulated on poly(styrene-co-maleic acid) (SMA) micelles. The obtained SMA–dasatinib nanoparticles (NPs) were evaluated for their physicochemical properties, in vitro antiproliferative activity in different TNBC cell lines, and in vivo anticancer activity in a syngeneic model of breast cancer. Obtained results showed that SMA–dasatinib is more potent against 4T1 TNBC tumor growth in vivo compared to free drug. This enhanced effect was ascribed to the encapsulation of the drug protecting it from a rapid metabolism. Our finding highlights the often-overlooked value of nanoformulations in protecting its cargo from degradation. Overall, results may provide an alternative therapeutic strategy for TNBC management.

**Keywords:** TNBC; dasatinib; poly(styrene-co-maleic acid) micelles; nanoformulation; metabolism; EPR; nanomedicine; targeted therapy

#### **1. Introduction**

Breast cancers are the top widespread type of tumor among females in the U.S., and in 2021, it is predicted that 280,000 new breast cancers will be diagnosed [1,2]. The disease is globally affecting about 1 in 8 women in the U.S. during their lifetime. Breast cancer mortality could be attributed to metastasis by 80–90% [3].

Triple negative breast cancer (TNBC) is a long-lasting orphan disease and among the most clinically challenging breast cancer subtype. TNBC is the most aggressive and heterogeneous breast tumor that lacks all of three therapeutically relevant biomarkers including estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) [4]. The conventional treatment for TNBC involves surgical excision and radiotherapy with a combination of adjuvant chemotherapies [5,6]. Despite current

**Citation:** Bahman, F.; Pittalà, V.; Haider, M.; Greish, K. Enhanced Anticancer Activity of Nanoformulation of Dasatinib against Triple-Negative Breast Cancer. *J. Pers. Med.* **2021**, *11*, 559. https://doi.org/ 10.3390/jpm11060559

Academic Editor: Jun Fang

Received: 21 April 2021 Accepted: 11 June 2021 Published: 15 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

therapeutic regimens, patients affected by TNBC show frequently fatal prognosis and are exposed to early relapse and metastatic spread, as a result of resistance to chemotherapies [5]. Despite initially TNBC exhibiting more chemo-sensitivity than other groups of breast cancer, it shows high propensity to spread and metastasize to vital organs, for instance the lungs and brain, rendering the survival rate still significantly lower than patients with non TNBC across any phase of diagnosis [7–9]. These aggressive phenotypes can be at least to some extent ascribed to the incidence of breast cancer stem cells (BCSCs). In addition, the lack of targeted therapies increases the use of traditional chemotherapy often accompanied by severe side effects. The subclassification of TNBC based on gene expression profiling analysis includes basal like 1 (BL1) and basal like 2 (BL2), immunomodulatory (IM), mesenchymal (M), mesenchymal stem like (MSL), and luminal androgen receptor positive (LAR) [10–12]. This classification is paving the way to the identification of more specific molecular targets for TNBC treatment. In fact, these subtypes show different drug sensitivity profiles to anticancer treatments such as cisplatin for BL1 and BL2, PI3K, and proto-oncogene tyrosine-protein kinase (Src) inhibitors [13].

Src is a protein tyrosine kinase that regulates various cancerous events at an intracellular level, such as cellular adhesion, invasion, growth, survival, and vascular endothelial growth factor (VEGF) expression [14,15]. In addition, Src regulates an osteoclast function in normal bone and bone metastases [16]. A number of literature reports have evidenced in TNBC an abnormal activation and amplification of Src or Src-family kinases (SFK) and an involvement in metastasis regulation [17]. Not surprisingly, TNBC shows increased sensitivity to Src inhibitors compared to other cancer subgroups [18–20]. In addition, it has been demonstrated that ER and HER2 expression levels affect the beneficial effects of Src inhibitors in TNBC [21]. Therefore, Src can be considered as a new molecular target for TNBC therapy, and Src inhibitors have long been proposed as new antitumoral treatments, since they are able to prevent cell growth in liver, colon, breast, and ovarian cancers.

Dasatinib (Figure 1) is a Src, BCR-ABL, c-KIT, PDGFR-α and PDGFR-β, and ephrin receptor kinase blockers accepted by the Food and Drug Administration (FDA) for treating cases of Philadelphia chromosome positive leukemias (chronic myeloid leukemia; CML) [22,23]. Preclinical studies demonstrated significant inhibition of malignant breast cells growth through reducing the percentage of aldehyde dehydrogenase-positive (ALDH+) BCSCs within the BL-2 subtype of breast cancer [13,24]. Considering that BCSCs are often responsible for the onset of chemotherapy resistance, dasatinib has been considered for the treatment of TNBC [24,25]. Similarly, preclinical studies evidenced synergistic or additive dasatinib activity with chemotherapy, implying that this Src inhibitor can offer clinical benefit in TNBC [26]. Regrettably, patients suffering from TNBC have inadequate benefit from Src inhibitors treatment [27–29]. In fact, despite promising preclinical results, a phase II clinical trial by administering dasatinib as a single agent highlighted only a 9% clinical benefit rate, and other clinical trials terminated due to futility (e.g., NCT00817531, NCT00780676, etc.) [29]. Moreover, dasatinib suffers from some limitations related to its pharmacokinetic profile. Oral absorption of dasatinib is quick and produces around 80% of bioavailability; however, it is rapidly eliminated through CYP3A4-mediated metabolism, with a T1/2 of 3–4 h. In addition, dasatinib bioavailability has reduced the its ability to modify gastric pH value (antacids, H2-receptor blockers, proton pump inhibitors) and is modified according to the concomitant treatment with CYP3A4 inducers or inhibitors [30].

In recent years, considerable attention has been devoted to strategic application of nanoscience to pharmaceutical development with the aim of improving efficacy, delivery at the site of action, safety, physicochemical properties, and the absorption, distribution, metabolism, and excretion (ADME) profile of bioactive compounds [31]. In particular, nanoparticle formulations (NPs) can guarantee increased bioavailability of drugs administered orally, enhanced half-life of intravenous drugs (by reducing both metabolism and elimination), and augmented drug concentration in specific tissues [32]. Taking in account the dasatinib ADME profile, encapsulation of the drug into NPs may improve the drug efficacy, minimize side effects, and permit the active principle to assemble at the malignant

tumor site by means of the enhanced permeability and retention (EPR) effect [33,34]. In addition, using the SMA micellar system to generate dasatinib NPs has multiple advantages over other nanoformulations. It produces a micelle with a nearly neutral or slightly negative charge reducing opsonization of the micelles, recognition by the reticuloendothelial system, and elimination from the blood circulation [35]. In our previous work utilizing dasatinib micelles compared to free drug, we had shown enhanced anticancer activity both in vitro and in vivo against various glioblastoma cell lines and in animal model of the disease [36]. In this study, we encapsulated dasatinib into polystyrene co-melic acid (SMA) micelles to generate micellar dasatinib system (SMA–dasatinib) that has been characterized for physicochemical properties including size, loading, charge, and release rate. In addition, SMA–dasatinib has been assessed for their anticancer effect in vitro using 4T1, MDA-MB-231, and MCF-7 cell lines and in vivo in a syngenic model of TNBC. The cell lines chosen represents a spectrum of commonly used breast cancer cell lines of both hormonal responsive and TNBC of human and murine origin. Our choice of cell lines will allow the comparison of dasatinib formulation in different biological environments and further allows the comparison of our results to earlier and subsequent research in the field. Encouraging obtained results will pave the way for further study in the management of TNBC. *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 3 of 13

**Figure 1.** (**a**) Chemical structure of dasatinib; (**b**) SMA–dasatinib drug release studies. Cumulative release of the free drug from SMA–dasatinib micelles at pH 7.4 in PBS and FBS. **Figure 1.** (**a**) Chemical structure of dasatinib; (**b**) SMA–dasatinib drug release studies. Cumulative release of the free drug from SMA–dasatinib micelles at pH 7.4 in PBS and FBS.

#### In recent years, considerable attention has been devoted to strategic application of **2. Materials and Methods**

nanoscience to pharmaceutical development with the aim of improving efficacy, delivery at the site of action, safety, physicochemical properties, and the absorption, distribution, metabolism, and excretion (ADME) profile of bioactive compounds [31]. In particular, nanoparticle formulations (NPs) can guarantee increased bioavailability of drugs administered orally, enhanced half-life of intravenous drugs (by reducing both metabolism and elimination), and augmented drug concentration in specific tissues [32]. Taking in account the dasatinib ADME profile, encapsulation of the drug into NPs may improve the drug efficacy, minimize side effects, and permit the active principle to assemble at the malignant tumor site by means of the enhanced permeability and retention (EPR) effect [33,34]. Dasatinib were retained from LC Laboratories (Woburn, MA, USA). Polystyrene co-maleic anhydride (molecular weight~1600), Roswell Park Memorial Institute (RPMI) 1640 medium, Hank's balanced salt solution, fetal bovine serum (FBS), bovine serum albumin (BSA), and TrypLE express were bought from ThermoFisher Scientific (Dubai, United Arab Emirates). *N*-(3-dimethylaminopropyl)-*N*-ethylcarbodiimide hydrochloride (EDAC), L-glutamine, antibiotic solution of penicillin/streptomycin were acquired from (Merck Hertfordshire, UK). All consumable materials including petri dishes, conical tubes (15 mL and 50 mL), cell culture flasks (25 and 75 cm<sup>2</sup> ), and dialysis tubing were purchased from (Merck Hertfordshire, UK).

#### vantages over other nanoformulations. It produces a micelle with a nearly neutral or *2.1. SMA–Dasatinib Micelles Synthesis*

the management of TNBC.

**2. Materials and Methods**

slightly negative charge reducing opsonization of the micelles, recognition by the reticuloendothelial system, and elimination from the blood circulation [35]. In our previous work utilizing dasatinib micelles compared to free drug, we had shown enhanced anticancer activity both in vitro and in vivo against various glioblastoma cell lines and in animal model of the disease [36]. In this study, we encapsulated dasatinib into polystyrene SMA–micelles were synthesized as previously reported [36]. Briefly, SMA was hydrolyzed by adding the SMA powder to 1 M NaOH solution at 70 ◦C to reach a concentration of 10 mg/mL. After this time, the pH of the obtained solution of SMA was adjusted to pH 5.0. This was followed by adding EDAC (1:1 weight ratio with SMA). Dasatinib was dissolved in dimethyl sulfoxide (DMSO) at 25% weigh ratio to SMA. Dasatinib was

co-melic acid (SMA) micelles to generate micellar dasatinib system (SMA–dasatinib) that has been characterized for physicochemical properties including size, loading, charge,

in vitro using 4T1, MDA-MB-231, and MCF-7 cell lines and in vivo in a syngenic model of TNBC. The cell lines chosen represents a spectrum of commonly used breast cancer cell lines of both hormonal responsive and TNBC of human and murine origin. Our choice of cell lines will allow the comparison of dasatinib formulation in different biological environments and further allows the comparison of our results to earlier and subsequent research in the field. Encouraging obtained results will pave the way for further study in

Dasatinib were retained from LC Laboratories (Woburn, MA, USA). Polystyrene comaleic anhydride (molecular weight~1600), Roswell Park Memorial Institute (RPMI) 1640 medium, Hank's balanced salt solution, fetal bovine serum (FBS), bovine serum albumin (BSA) ,and TrypLE express were bought from ThermoFisher Scientific (Dubai, UAE). N- (3-dimethylaminopropyl)-N-ethylcarbodiimide hydrochloride (EDAC), L-glutamine, antibiotic solution of penicillin/streptomycin were acquired from (Merck Hertfordshire,

In addition, using the SMA micellar system to generate dasatinib NPs has multiple ad-

added to the solution, and pH was kept at 5 by adding 0.1 HCL until pH remained stable at 5.0. Then, the pH was raised up to reach 11.0 and kept until it become stable. The pH was then dropped to 7.4, and the solution was filtered 4 times by meand of a Millipore Labscale TFF system with a Pellicon XL 10 KDa cutoff membrane. Finally, the concentrated SMA–dasatinib micelles were frozen at −80 ◦C and the following day lyophilized (5 mTorr and −52 ◦C) to achieve a stable SMA–dasatinib powder.

#### *2.2. SMA–Dasatinib Micelles Characterization*

The SMA micelles loading was determined by using three different samples of 1.0 mg/mL of SMA–dasatinib micelles dissolved in DMSO for measuring absorbance at 320 nm of dasatinib to a previously prepared standard curve of the drug intending to determine the ratio between the micelle and the loaded dasatinib.

For size distribution and zeta potential determination of SMA–dasatinib micelles, a Malvern ZEN3600 Zetasizer Nano series was used (Malvern Instruments Inc., Westborough, MA, USA) by using 1 mg/mL of the SMA–dasatinib nanomicelles dissolved in both double DW as a solvent for size measurement or for charge measurement. Then, to measure the release rate of free drug (dasatinib) from SMA micellar system, two separate experiments have been performed by measuring the release in PBS and in FBS. A 2 mg of the SMA– dasatinib were dissolved in 2 mL of PBS or FBS, respectively, and inserted into a 10 kDa cutoff dialysis membrane that was flooded in 20 mL of PBS or FBS for 72 h. At specified time points, the surrounding water was collected from outside the dialysis bag and replaced with PBS or FBS, and the absorbance was measured at 320 nm.

#### *2.3. Cell Culture*

The cell lines 4T1, MDA-MB-231, and MCF-7 were obtained from American Type Culture Collection (ATCC) (Manassas, VA, USA). RPMI medium supplemented with 5% fetal bovine serum (FBS) was used to culture the cell lines while being maintained in a humidified atmosphere at 37 ◦C, 5% CO2.

#### In Vitro Anti-Proliferative Effect of Dasatinib and SMA–Dasatinib Micelles

Cells were seeded in 96-well plates (density: 4T1 5 <sup>×</sup> <sup>10</sup><sup>3</sup> , MDA-MB-231 5 <sup>×</sup> <sup>10</sup><sup>3</sup> , MCF-7 5 <sup>×</sup> <sup>10</sup><sup>3</sup> cells/well, respectively) and incubated for 24 h at 37 ◦C in 5% CO<sup>2</sup> and then treated with a different of concentrations of dasatinib 0 to 10 µM) or SMA–dasatinib (0 to 10 µM). The cytotoxicity was assessed after 48 h incubation using a sulforhodamine B (SRB) assay, as described previously [37]. Free SMA and DMSO at concentrations equivalent to that used for testing dasatinib were used as controls. Cells were fixed using 10% trichloroacetic acid and stained with SRB. The cytotoxicity experiments were performed in triplicate (*n =* 3). Then, the 50% growth inhibition (IC50) was assessed by using SRB assay after 48 h incubation. HepG-2 cells also were seeded at a density of 50,000 cells/cm<sup>2</sup> onto a 25 cm<sup>2</sup> flask. Then, cells were treated by various concentration of dasatinib and SMA–dasatinib. Twenty-four hours after incubation, the supernatants were collected and diluted accordingly to retreat 4T1 cells. Data were represented as mean ± SD of three independent experiments of each cell lines.

#### *2.4. Effect of Dasatinib and DMA-Dasatinib Treatment in In Vivo Syngeneic Model*

The Laboratory Animal Care Facility of the Arabian Gulf University (AGU), Bahrain, supplied the Female Balb/c mice (6–12 weeks old, mean weight 20–25 gm). Animals were maintained under standard conditions such as controlled temperature (25 ◦C), a 12 h photoperiod, and had access to food and drinking water ad libitum. All animal experiments were performed based on the rules and regulations of AGU Animal Care Policy and approved by the Research and Ethics Committee, REC approval No: G- E003- PI-04/17.

To propagate the tumor, female Balb/c mice (*n =* 3) were treated subcutaneously with 1 million 4T1 mammary carcinoma cells in both sides (right and left side) of the

mice back. The tumor then was collected and cut down into small pieces of average size 1–3 mm<sup>3</sup> in sterile PBS to sustain tumor viability. Following this, 5 mice of each group were cleanshaven, anesthetized, and injected with one small piece of the 4T1 tumor tissue subcutaneously. When the tumors reached 100 mm<sup>3</sup> in size, mice were casually divided into three groups *n =* 5 in each group (negative control, dasatinib, and SMA–dasatinib) subjected to drug treatment. Dasatinib was given once at a dose of 5 mg/kg via the tail vein, whereas SMA–dasatinib at a dose of 5 mg/kg (dasatinib equivalent dose) dissolved in PBS was given by IV injection. The first day of drug treatment was established as day 0. Tumor volume was measured by manual caliber; the size was assessed by using this formula:

V (mm<sup>3</sup> ) = ((transverse section (W)<sup>2</sup> <sup>×</sup> longitudinal cross section (L))/ 2).

Tumor sizes were normalized by using the original tumor measure and represented as mean ± standard error of the mean (SEM). Additionally, the body weight of mice was estimated every day and normalized daily for 10 days.

#### *2.5. In Vivo Biodistribution of Dasatinib and SMA–Dasatinib*

Cells of 4T1 were injected into female Balb/c mice, bilaterally on the flanks to obtain 1–3 mm<sup>3</sup> tumor size. When cancer volume reached 100 mm<sup>3</sup> , mice were casually divided into 2 groups (4 mice per group). Mice were injected with both dasatinib or the equivalent in SMA–dasatinib at 50 mg/kg via the tail vein. Mice were sacrificed 24 h after drugs injection and different organs were collected. Organs such as heart, lungs, liver, spleen, kidneys, and tumor tissue were examined for dasatinib content. SMA–dasatinib was taken out using the method reported earlier [38]. In brief, organs were crushed, weighed, and snap-frozen before pulverization. Obtained frozen tissue powder (1 mg) was treated with 67% ethanol and 1 mL of HCl 4M. The suspension was incubated at 70 ◦C for 30 min, sonicated, and centrifuged to take out dasatinib from tissue samples. Dasatinib amount was measured by absorbance at 332 nm and compared to a dasatinib calibration curve. The amount of dasatinib was standardized to the weight of tissue and to the whole weight of the organs from which it was extracted.

#### *2.6. Statistical Analysis*

The data from both experiments in vitro and in vivo were evaluated using GraphPad prism software. Tumor size measurements are expressed as group means ± SEM in the treatment groups. Cytotoxicity experiments with dasatinib and SMA–dasatinib are reported as means ± SD. The statistical significance of difference between groups were performed using a two-tailed t-test. Statistical differences were considered significant if the *p*-value was <0.05.

#### **3. Results**

#### *3.1. Synthesis and Characterization of SMA–Dasatinib*

SMA–dasatinib was synthesized and characterized by a low critical micelle concentration (CMC), as previously described [36]. Furthermore, the structural variation of hydrophobic styrene and hydrophilic maleic groups stimulates the quick construction of SMA micelles and facilitates the encapsulation of dasatinib. The loading of SMA–dasatinib was 11.5%, calculated as the weight ratio of the dasatinib over the total amount of SMA micelle weight. Micelles average size measuring showed that SMA–dasatinib micelles were 198 nm and had a polydispersity index (PDI) of 0.17, which was determined by dynamic light scattering (DLS). As shown in Table 1, the zeta potential of SMA–dasatinib is near neutral with a value of 0.0035 mV, which can sustain the micelle in the blood circulation for a long time by lowering the clearance by the reticuloendothelial system and allows accumulation in the tumor [39].



<sup>1</sup> Data are shown as mean values <sup>±</sup> standard deviation (SD). Values are the mean of triplicate experiments; <sup>2</sup> PDI = polydispersity index.

Thus, the average size of SMA–dasatinib is within the size range to facilitate its accumulation in tumor tissue by the effect of enhanced permeability and retention (EPR) [40]. Moreover, the release rate of the drug from the micelles was more efficient in an environment mimicking extracellular pH than in the blood (53 vs. 44%) following 96 h incubation).

The release of dasatinib from SMA micelles was assessed at physiological pH 7.4 in PBS and FBS, respectively, for 96 h (Figure 1). The SMA–dasatinib micelles were stable in solution with about half of the formulation released after 96 h. Moreover, in the first 2 h, the cumulative release was around 5%, as shown in Figure 1. The stability of the micellar system depends on the slow release in the blood circulation, which promotes the SMA–dasatinib accumulation at the tumor site through the EPR effect. A previous study has demonstrated the endocytosis of SMA micelle through caveolin-1 [41]. Therefore, SMA–dasatinib will be internalized by endocytosis and the release of dasatinib into the TNBC tumor cells.

#### *3.2. Cytotoxicity of Dasatinib and SMA–Dasatinib versus Breast Cancer Cell Lines*

The assessment of the cytotoxic effect of SMA–dasatinib and dasatinib on cell viability was achieved using different breast cancer cell lines, such as human MDA-MB-231, 4T1, and MCF-7 cells. A cell's cytotoxicity of dasatinib and SMA–dasatinib was determined by means of the SRB assay. Equivalent concentrations of free SMA and DMSO were used to dissolve the dasatinib and yielded no cytotoxic effect.

The treatment of MCF-7 cells (Figure 2A and Table 2) evidenced that either dasatinib and SMA–dasatinib showed no noteworthy difference in their cytotoxic activity after 48 h incubation and both displayed an IC<sup>50</sup> > 10 µM. An IC<sup>50</sup> value of 6.1 ± 2.2 µM was obtained for the dasatinib treatment of MDA-MB-231 cells, while SMA–dasatinib exhibited an IC<sup>50</sup> value of 8.16 ± 3.1 µM (Figure 2B and Table 2). An enhanced effect could be partially attributed to greater internalization capability of MDA-MB-231 cells compared to MCF-7 cells [26]. *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 7 of 13

**Table 2.** Experimental IC<sup>50</sup> values (μM) of free dasatinib and SMA–dasatinib towards human MDA-

MCF7 >10 >10

4T1 0.014 ± 0.003 0.083 ± 0.01 Hep-G2 >10 >10 4T1 after Hep-G2 0.21 ± 0.04 0.09 ± 0.012

IC<sup>50</sup> value determination was performed using GraphPad Prism. Data are reported as IC<sup>3</sup> values in μM ± standard deviation (SD). Values are the mean of triplicate experiments.<sup>2</sup> The IC50 value calculations were calculated according to GraphPad prism algorithm and were included to have a numerical reference value of comparison, although it is clear that a plateau is reached after certain

The anticancer activity of dasatinib and SMA–dasatinib was evaluated using Balb/c mice harboring 4T1 tumor over a treatment period of 10 days. Figure 3A shows that during the first days' treatment with free dasatinib (5 mg/kg) tumor growth seems to be delayed, while overall tumor size did not change significantly after 10 days in comparison to control-treated mice. Very differently, treatment with SMA–dasatinib almost entirely

**Dasatinib SMA–Dasatinib**

**Cell Line IC<sup>50</sup> (µM) 1,2**

*3.3. Effect of Dasatinib and SMA–Dasatinib on the Development of 4T1 Tumors*

**Figure 2.** Cytotoxicity of dasatinib and SMA–dasatinib (**A**) against MCF-7, (**B**) MDA-MB-231, (**C**) and 4T1 cells. The cells were treated for 72 h with specific concentrations of dasatinib and SMA–dasatinib micelles. The cell number was determined using the SRB assay. Data are expressed as mean ± SEM (*n =* 3). **Figure 2.** Cytotoxicity of dasatinib and SMA–dasatinib (**A**) against MCF-7, (**B**) MDA-MB-231, (**C**) and 4T1 cells. The cells were treated for 72 h with specific concentrations of dasatinib and SMA–dasatinib micelles. The cell number was determined using the SRB assay. Data are expressed as mean ± SEM (*n =* 3).

MB-231, 4T1, and MCF-7 cells.

1

concentrations in MCF-7 and MDA-MB 231 cell lines.

stopped the tumor growth for the duration of the study.


**Table 2.** Experimental IC<sup>50</sup> values (µM) of free dasatinib and SMA–dasatinib towards human MDA-MB-231, 4T1, and MCF-7 cells.

1 IC<sup>50</sup> value determination was performed using GraphPad Prism. Data are reported as IC<sup>3</sup> values in <sup>µ</sup><sup>M</sup> <sup>±</sup> standard deviation (SD). Values are the mean of triplicate experiments.<sup>2</sup> The IC<sup>50</sup> value calculations were calculated according to GraphPad prism algorithm and were included to have a numerical reference value of comparison, although it is clear that a plateau is reached after certain concentrations in MCF-7 and MDA-MB 231 cell lines.

Both MDA-MB-231 and MCF-7 reached a plateau, which can be explained by the inherent dependence of the breast cancer cell lines on tyrosine kinases signaling for growth and division. Cells of 4T1 treated with free dasatinib and SMA–dasatinib showed a significant cytotoxic effect when compared to MCF-7 and MDA-MB-231 cells with IC<sup>50</sup> of 0.014 ± 0.003 and 0.083 ± 0.01 µM, respectively (Figure 2C and Table 2).

#### *3.3. Effect of Dasatinib and SMA–Dasatinib on the Development of 4T1 Tumors*

The anticancer activity of dasatinib and SMA–dasatinib was evaluated using Balb/c mice harboring 4T1 tumor over a treatment period of 10 days. Figure 3A shows that during the first days' treatment with free dasatinib (5 mg/kg) tumor growth seems to be delayed, while overall tumor size did not change significantly after 10 days in comparison to control-treated mice. Very differently, treatment with SMA–dasatinib almost entirely stopped the tumor growth for the duration of the study. *J. Pers. Med.* **2021**, *11*, x FOR PEER REVIEW 8 of 13

**Figure 3.** In vivo antitumor activity of dasatinib and SMA–dasatinib on 4T1tumor bearing Balb/c mice. Mice were treated for 10 days with single dose of either dasatinib 5 mg/kg and SMA–dasatinib 5 mg/kg. Control group was injected with PBS (pH 7.4). Tumor volume changes (**A**) and body weight changes (**B**) were monitored over the treatment period. Data are presented as the mean of triplicate experiments ± standard error. **Figure 3.** In vivo antitumor activity of dasatinib and SMA–dasatinib on 4T1tumor bearing Balb/c mice. Mice were treated for 10 days with single dose of either dasatinib 5 mg/kg and SMA–dasatinib 5 mg/kg. Control group was injected with PBS (pH 7.4). Tumor volume changes (**A**) and body weight changes (**B**) were monitored over the treatment period. Data are presented as the mean of triplicate experiments ± standard error.

The therapeutic efficacy of dasatinib and SMA–dasatinib treatments were not associated with any statistically significant weight loss during the treatment period, as shown in Figure 3B and Table 3. The therapeutic efficacy of dasatinib and SMA–dasatinib treatments were not associated with any statistically significant weight loss during the treatment period, as shown in Figure 3B and Table 3.

Mean weight 23.8 24.2 23.9 Std. deviation 0.4637 0.7537 0.7987

<sup>1</sup> Data are presented as the mean of triplicate experiments ± standard error.

*3.4. In Vivo Biodistribution of Dasatinib and SMA–Dasatinib*

**Day Control Dasatinib SMA–Dasatinib**

The biodistribution of dasatinib and SMA–dasatinib were measured in vivo; to this extent, immunocompetent Balb/c mice harboring 4 T1 tumors were intravenously injected with equivalent doses of dasatinib or SMA–dasatinib, and the concentration of dasatinib in various organs and tumor has been measured. As reported in Figure 4, dasatinib and SMA–dasatinib are distributed to the heart, liver, lung, kidney, and spleen. There was an increased accumulation of dasatinib following SMA–dasatinib injection in the spleen, kidney, and lung when compared to the free dasatinib injection (Figure 4A). No significant statistical difference was observed in the heart and liver. Additionally, in the tumor when comparing SMA–dasatinib to the free dasatinib injection (Figure 4B), no statistically sig-

**Table 3.** Body weight changes upon treatment with dasatinib and SMA–dasatinib were monitored

over the treatment period <sup>1</sup>

nificant difference was observed.

.

1


**Table 3.** Body weight changes upon treatment with dasatinib and SMA–dasatinib were monitored over the treatment period <sup>1</sup> .

<sup>1</sup> Data are presented as the mean of triplicate experiments <sup>±</sup> standard error.

#### *3.4. In Vivo Biodistribution of Dasatinib and SMA–Dasatinib*

The biodistribution of dasatinib and SMA–dasatinib were measured in vivo; to this extent, immunocompetent Balb/c mice harboring 4 T1 tumors were intravenously injected with equivalent doses of dasatinib or SMA–dasatinib, and the concentration of dasatinib in various organs and tumor has been measured. As reported in Figure 4, dasatinib and SMA–dasatinib are distributed to the heart, liver, lung, kidney, and spleen. There was an increased accumulation of dasatinib following SMA–dasatinib injection in the spleen, kidney, and lung when compared to the free dasatinib injection (Figure 4A). No significant statistical difference was observed in the heart and liver. Additionally, in the tumor when comparing SMA–dasatinib to the free dasatinib injection (Figure 4B), no statistically significant difference was observed.

**Figure 4.** (**A**) Tissue and (**B**) tumor distribution of free dasatinib and SMA–dasatinib at 24 h after intravenous injection of dasatinib or SMA–dasatinib (50 mg/kg) to Balb/c mice bearing 4T1 tumors (*n =* 5). Representation of the relative content of dasatinib per 100 mg tissue expressed in free and micellar dasatinib.

#### *3.5. Cytotoxicity of Dasatinib and SMA–Dasatinib Versus HepG2 Cell Line and 4T1 after Passage in HepG2*

The effect of SMA–dasatinib and dasatinib on HepG2 cell viability was assessed by using SRB assay. HepG2 cells upon treatment with dasatinib and SMA–dasatinib micelles did not show any significant toxicity (Figure 5A) after 48 h incubation and both displayed an IC<sup>50</sup> > 10 µM (Table 2). However, when the supernatants obtained from HepG2 treatment with dasatinib and SMA–dasatinib, respectively, were added to 4T1 cells the correspondent IC50s varied noticeably. While the SMA–dasatinib cytotoxicity did not change (IC<sup>50</sup> = 0.09 vs. 0.083 µM, respectively, Table 2); dasatinib cytotoxicity resulted in a 15-fold decrease in cytotoxicity (IC<sup>50</sup> = 0.21 vs. 0.014 µM, respectively, Table 2).

(**A**) (**B**)

decrease in cytotoxicity (IC<sup>50</sup> = 0.21 vs. 0.014 µM, respectively, Table 2).

*3.5. Cytotoxicity of Dasatinib and SMA–Dasatinib Versus HepG2 Cell Line and 4T1 after* 

The effect of SMA–dasatinib and dasatinib on HepG2 cell viability was assessed by using SRB assay. HepG2 cells upon treatment with dasatinib and SMA–dasatinib micelles did not show any significant toxicity (Figure 5A) after 48 h incubation and both displayed an IC<sup>50</sup> > 10 µM (Table 2). However, when the supernatants obtained from HepG2 treatment with dasatinib and SMA–dasatinib, respectively, were added to 4T1 cells the correspondent IC50s varied noticeably. While the SMA–dasatinib cytotoxicity did not change (IC<sup>50</sup> = 0.09 vs. 0.083 µM, respectively, Table 2); dasatinib cytotoxicity resulted in a 15-fold

**Figure 4.** (**A**) Tissue and (**B**) tumor distribution of free dasatinib and SMA–dasatinib at 24 h after intravenous injection of dasatinib or SMA–dasatinib (50 mg/kg) to Balb/c mice bearing 4T1 tumors (*n =* 5). Representation of the relative content

**Figure 5.** Cytotoxicity of dasatinib and SMA–dasatinib (**A**) against HepG2 cells, (**B**) 4T1 cells after treatment with HepG2. The cells were treated for 48 h with specific concentrations of dasatinib and SMA–dasatinib micelles. The cell number was determined using the SRB assay. Data are expressed as mean ± SEM (*n =* 8). **Figure 5.** Cytotoxicity of dasatinib and SMA–dasatinib (**A**) against HepG2 cells, (**B**) 4T1 cells after treatment with HepG2. The cells were treated for 48 h with specific concentrations of dasatinib and SMA–dasatinib micelles. The cell number was determined using the SRB assay. Data are expressed as mean ± SEM (*n =* 8).

#### **4. Discussion 4. Discussion**

of dasatinib per 100 mg tissue expressed in free and micellar dasatinib.

*Passage in HepG2*

Dasatinib is a multi-target kinase inhibitor, including BCR/ABL kinases and Src family kinases (SFK) that are closely linked to multiple signal pathways that regulate prolif-Dasatinib is a multi-target kinase inhibitor, including BCR/ABL kinases and Src family kinases (SFK) that are closely linked to multiple signal pathways that regulate proliferation, invasion, survival, metastasis, and angiogenesis [42]. Dasatinib showed promising results in the treatment of TNBC as a single agent or as a neoadjuvant; nevertheless, its use is limited by its poor aqueous solubility (6.49 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mg/mL). Moreover, after oral administration, dasatinib is subjected to extensive first pass metabolism, where multiple CYP enzymes appear to have the potential to metabolize the drug [43]. Our current work aims at encapsulating dasatinib into SMA micelles to generate an SMA–dasatinib micellar system that can improve its solubility in water, protect the drug against enzymatic degradation, potentiate its chemotherapeutic effect, and minimize the rate of drug resistance.

The characterization of SMA–dasatinib micelles showed successful encapsulation of the drug with a loading capacity of 11.5%. Given that effective molecular size for EPR is 20–200 nm, the micellar size of 198 nm favors the accumulation of the nanoparticles in the tumor cells. In addition, the particle size of the prepared drug-loaded micelles should improve their circulation time and extend their plasma half-life by avoiding their rapid elimination from the kidney. The surface charge of the obtained prepared SMA–dasatinib micelles was almost neutrall, which is desirable to limit their interaction with active plasma constituents such as complement system and coagulation factors. Further, a near neutral charge will ensure selective EPR-based extravasation through tumor vasculature with minimal interaction with normal endothelial cell membrane. The micellar formulations showed a sustained slow-release rate of the drug for 96 h in both PBS and FBS (Figure 2), which shows that they can function as a reservoir for delivering a consistent level of dasatinib once concentrated extracellularly at tumor tissues and, hence, prolong the exposure of tumor cells to effective doses of the drug.

The examination of dasatinib-induced inhibition of metabolic activity on three commonly studied TNBC cell lines showed different responses. The MCF7 cell line was the least sensitive to treatment with SMA–dasatinib and free drug (IC<sup>50</sup> > 10 µM) compared to MDA-MB-231 (IC<sup>50</sup> 8.16 and 6.1 µM, respectively). This correlates with previous studies, which suggested that MDA-MB-231 are more sensitive due to the presence of active ABL kinase and their greater drug internalization capacity [26,44]. The 4T1 cell line exhibited a significantly high sensitivity to dasatinib and SMA–dasatinib (IC<sup>50</sup> = 0.014 and 0.083 µM, respectively) compared to MDA-MB-231 and MCF7, which may be due to their sensitivity to Src (Kin-2) receptor tyrosine kinase blockade [45]. Interestingly, there was no significant difference in the cytotoxic effect of the SMA–dasatinib and the free drug on the different types of TNBC cell lines in vitro. Nevertheless, Figure 4 showed that treatment with SMA– dasatinib significantly inhibited the tumor growth in vivo compared to animals treated

with the free drug. Both treatments resulted in no significant weight loss in treated animals, indicating that it is relatively safe to use dasatinib and SMA–dasatinib micelles in this animal model.

The biodistribution after IV administration showed a significantly high accumulation of SMA–dasatinib in the spleen compared to the free drug. This could possibly be due to the fact the size of SMA–dasatinib micelles is larger than the fenestration of the liver vasculature, which can reduce the hepatic uptake of the micelles and may decrease the metabolism of the drug. On the other hand, there was no significant difference between the tumor distribution of dasatinib and SMA–dasatinib. Dasatinib is characterized by a large volume of distribution and human plasma protein binding. In vitro studies showed that plasma protein binding of dasatinib can reach 96%, creating a depot from which the drug slowly releases its free form. It may also increase the molecular size of the drug and enhance its accumulation at the tumor site by EPR effect similar to SMA–dasatinib [46].

Treatment of HepG2 cells with dasatinib and SMA–dasatinib micelles did not show significant toxicity (IC<sup>50</sup> > 10 µM). This is probably due to low expression levels of Src kinase, which reduced the sensitivity of the cell line to the drug [42]. The passage of dasatinib and SMA–dasatinib through HepG2 before treatment of 4T1 cells was carried out to check the effect of metabolism on the cytotoxic ability of the treatments. Dasatinib is significantly metabolized by CYP3A4 in the liver generating an active metabolite with similar potency to the drug; however, it represents only 5% of dasatinib in plasma. The co-administration of potent CYP3A4 inducer results in a considerable reduction on Cmax and AUC of the drug [43]. Treatment of 4T1 cells with supernatants obtained from HepG2 treatment showed a significant decrease in cytotoxicity of the free dasatinib, while the cytotoxic effect of SMA–dasatinib remained unchanged. The encapsulation of dasatinib offered protection for the drug against enzymatic degradation. The size of the produced micelles enhanced its accumulation at the tumor site by EPR effect and reduced its liver uptake. Our work is an emphasis of the overlooked advantage of nano-delivery systems in terms of cargo protection against degradation. This potential advantage was first described by Maeda back in 1991 [47]. Neocarzinostatin (NCS) is a very potent anticancer pretentious agent; however, NCS half-life is almost 1.9 min in tested mice. Using the nanoformulation of SMANCS protected the drug from the proteolytic activities in the plasma as well as extending its half-life by one order of magnitude. Further, this early work proved the safety of clinical use of SMA as a polymeric carrier for various biological payloads. Overall, our work further emphasizes the metabolic advantages of SMA–dasatinib nanosystems with a potential application for treating TNBC.

#### **5. Conclusions**

In this work, we have successfully synthesized and characterized an SMA nanomicellar system encapsulating the TKI dasatinib. Both the free drug and its nanoformulation have shown comparable cytotoxic activity in vitro against an array of breast cancer cell lines. The TKI and its nanoformulation proved to be more effective against TNBC cell lines compared to a hormone-sensitive cell line. In an animal model of 4T1 TNBC, the nanoformulations was about seven-fold more effective in controlling 4T1 implanted tumors. This pronounced in vivo activity was attributed to the protection of an SMA micellar system of TKI from the enzymatic degradation. Overall, our work can renew the interest in dasatinib as an effective treatment modality against TNBC.

**Author Contributions:** Conceptualization, funding acquisition, project administration, supervision, writing—review and editing, K.G.; methodology, validation, F.B.; software, formal analysis, data curation, writing—original draft preparation, supervision, writing—review and editing, M.H. and V.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Arabian Gulf University Research grant to KG, grant No: G003-PI-04/17.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research and Ethics Committee of Arabian Gulf university, approval No E003-PI-04/17 approved on 11 December 2017 and extended to November 2022.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding authors.

**Acknowledgments:** We sincerely acknowledge the technical support of Reem Al Zahrani and Sebastian Taurin.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

