Defucosylated Monoclonal Antibody (H₂Mab-139-mG_2a-f) Exerted Antitumor Activities in Mouse Xenograft Models of Breast Cancers against Human Epidermal Growth Factor Receptor 2

Round 1

Reviewer 1 Report

The authors have attempted to develop and characterize a class-switched HER2 targeting antibody for HER2 expressing tumors. While interesting, this work feels like an extension of their previously published work. The data are presented well with the exception of following points-

1. Flow cytometry is not an appropriate technique for determining the binding affinity and dissociation constants for antibodies. The authors should perform surface plasmon resonance for these assessments. Alternately, I suggest to remove the graph showing KD curve from Fig 2E and F and leave it in the text as suggested potential Kd for affinity

2. Figures 3 and 5 essentially show the exact same properties for the antibody in different cell lines and can be combined. Similarly, figures 4 and 6 can be combined. This will also streamline the text and flow of information.

3. A critical piece that is missing is benchmarking this antibody against the previous IgG1 and fucosylated version of it. The authors claim to have improved the ADCC and ACC by class switching and defucosylating the current ab. However, they have not shown any data as controls in their own experiments to demonstrate this.

3. Several sub-sections of the methods section do not describe the method but provide a reference for their previous work. At least a brief description needs to be added to provide adequate information to the readers without having to cross-reference other papers.

4. The discussion reads more like a literature review on various anti- HER2 antibodies. Lines 360-380 seem irrelevant to the current antibody and should be highly condensed or deleted entirely. 

Author Response

The authors have attempted to develop and characterize a class-switched HER2 targeting antibody for HER2 expressing tumors. While interesting, this work feels like an extension of their previously published work. The data are presented well with the exception of following points-

1. Flow cytometry is not an appropriate technique for determining the binding affinity and dissociation constants for antibodies. The authors should perform surface plasmon resonance for these assessments. Alternately, I suggest to remove the graph showing KD curve from Fig 2E and F and leave it in the text as suggested potential Kd for affinity

   We deleted the graph and left it in the text.

2. Figures 3 and 5 essentially show the exact same properties for the antibody in different cell lines and can be combined. Similarly, figures 4 and 6 can be combined. This will also streamline the text and flow of information.

   We demonstrated the data in HER2-overexpressed cells (Figure 3 and 4) and endogenous HER2-expressing cells (Figures 5 and 6), separately. Furthermore, the space is not enough to combine Figures 4 and 6. We hope that the reviewer understands the situation.

3. A critical piece that is missing is benchmarking this antibody against the previous IgG1 and fucosylated version of it. The authors claim to have improved the ADCC and ACC by class switching and defucosylating the current ab. However, they have not shown any data as controls in their own experiments to demonstrate this.

   We cited a review article about the function of mouse antibodies and Fc receptors (ref 24). According to the review, mouse IgG1 cannot bind to mouse FcγRIV which is essential for the activation of effector cells such as macrophages. In contrast, mouse IgG2a or IgG2b can bind to it with high affinity. In our previous research (ref 26), we already showed that a defucosylated IgG2a mAb exhibited a superior ability to activate effector cells compared to the original mAb.

   Furthermore, we compared the reactivity of H2Mab-139-mG2a-f and original H2Mab-139 against CHO/HER2 (Figure 1C), and confirmed that the class switch did not affect the recognition. We also showed that H2Mab-139-mG2a-f reacted with HEK-293T cells, but not with HER2-KO HEK293T (BINDS-23) cells (Figure 1D).

4. Several sub-sections of the methods section do not describe the method but provide a reference for their previous work. At least a brief description needs to be added to provide adequate information to the readers without having to cross- reference other papers.

   We first added the method. However, the MDPI office did not accept the result section because they have a high similarity to our previous published papers. Therefore, we cited them.

5. The discussion reads more like a literature review on various anti- HER2 antibodies. Lines 360-380 seem irrelevant to the current antibody and should be highly condensed or deleted entirely.

   We deleted.

Reviewer 2 Report

Overall, the author presented a defucosylated anti-HER2 monoclonal antibody derived from the previous developed H2Mab-139. With extensive characterization using flow cytometry, western blotting and IHC, the H2Mab-139-mG2a-f has shown a high selectivity and binding affinity with HER2. In addition, increased ADCC was observed, which is in accordance with the expectation. However, the rationale of fusing the variable region on a different IgG subtype (IgG2) remains unclear. The reviewer believes this is a critical concern since various IgG subtypes can have very different effector functions, affecting the downstream process, e.g. ADCC. Minor comments are as the following:

1.       Even though it has been well-studied that core fucosylation on the Fc can significantly affect ADCC, the reviewer still recommends to include the demonstration of it in the introduction, explaining the rationale for the defucosylation.

2.       How was the defucosylation performed?

3.       Please emphasize the mIgG2a control group is the fused mAb. The current description could be confusing with the original mIgG2.

4.       The disulfide bond pattern of IgG 2 in Figure 1 A is not accurate and it can be misleading.

5.       The reviewer could not find the discussion of Figure 2B.

6.       The reviewer recommends to re-structure the whole discussion section (section 4). Most of the part has been used to talk about the current commercialized mAbs targeting HER2. It would be better to focus on the presented study itself and the finding as well as the perspective of it.

 

 

Author Response

Overall, the author presented a defucosylated anti-HER2 monoclonal antibody derived from the previous developed H2Mab- 139. With extensive characterization using flow cytometry, western blotting and IHC, the H2Mab-139-mG2a-f has shown a high selectivity and binding affinity with HER2. In addition, increased ADCC was observed, which is in accordance with the expectation. However, the rationale of fusing the variable region on a different IgG subtype (IgG2) remains unclear. The reviewer believes this is a critical concern since various IgG subtypes can have very different effector functions, affecting the downstream process, e.g. ADCC.

We cited a review article about the function of mouse antibodies and Fc receptors (ref 24). According to the review, mouse IgG1 cannot bind to mouse FcγRIV which is essential for the activation of effector cells such as macrophages. In contrast, mouse IgG2a or IgG2b can bind to it with high affinity. In our previous research (ref 26), we already showed that a defucosylated IgG2a mAb exhibited a superior ability to activate effector cells compared to the original mAb.

Furthermore, we compared the reactivity of H2Mab-139-mG2a-f and original H2Mab-139 against CHO/HER2 (Figure 1C), and confirmed that the class switch did not affect the recognition. We also showed that H2Mab-139-mG2a-f reacted with HEK-293T cells, but not with HER2-KO HEK293T (BINDS-23) cells (Figure 1D).

Minor comments are as the following:

1. Even though it has been well-studied that core fucosylation on the Fc can significantly affect ADCC, the reviewer still recommends to include the demonstration of it in the introduction, explaining the rationale for the defucosylation.

   We added in the introduction (ref 24 and 25).

2. How was the defucosylation performed?

   We used FUT8-knockout ExpiCHO-S (BINDS-09). FUT8 is the only α1,6-fucosyltransferase that transfers fucose via an α1,6 linkage to the innermost N-acetylglucosamine on N-glycans for core fucosylation.

3. Please emphasize the mIgG2a control group is the fused mAb. The current description could be confusing with the original mIgG2.

   We did not use the fused mAb. As mentioned in section 2.8, we used the control mouse IgG2a purchased from Sigma- Aldrich.

4. The disulfide bond pattern of IgG 2 in Figure 1 A is not accurate and it can be misleading.

   Thank you very much. We changed.

5. The reviewer could not find the discussion of Figure 2B.

   We added the explanation in section 3.2.

6. The reviewer recommends to re-structure the whole discussion section (section 4). Most of the part has been used to talk about the current commercialized mAbs targeting HER2. It would be better to focus on the presented study itself and the finding as well as the perspective of it.

   We deleted a paragraph, and focus on the presented study.

Round 2

Reviewer 1 Report

Changes and explanations provided are adequate