The mechanisms of HER2 targeted ADCs are dependent on Rab GTPases

Abstract

Introduction: In the era of personalized cancer therapy, antibody-drug conjugates (ADCs) have become one of the fastest-emerging groups of anticancer drugs. ADCs consist of an antibody coupled to a cytotoxic payload by a chemical linker, designed to be cleaved off intracellularly. Understanding the intracellular trafficking and processing of ADCs is crucial for elucidating their mechanism of action.

Objective: This study aimed to compare trastuzumab deruxtecan (T-DXd) to ado-trastuzumab emtansine (T-DM1) with emphasis on Rab GTPase-regulated intracellular trafficking and its impact on ADC efficacy.

Methods: The efficacy of T-DXd and T-DM1 was assessed in a panel of HER2-positive cell lines. Correlations between ADC efficacy and the expression of HER2 and Rab GTPases were evaluated. Functional studies, including knockdown (KD), overexpression, and microscopy, were performed to evaluate the impact of Rab GTPases on ADC cytotoxicity.

Results: In contrast to T-DM1, T-DXd efficacy was found not to correlate to HER2 expression in a panel of HER2-positive cell lines. However, a correlation to RAB5A expression was found for T-DXd efficacy, although not as strong as for T-DM1. Altering the expression of RAB5 in our model system confirmed RAB5 to have an impact on both T-DXd and T-DM1 cytotoxicity, but more on T-DM1. In addition, RAB4a was found to influence T-DXd sensitivity, but not T-DM1, indicating differences in intracellular processing between T-DXd and T-DM1.

Conclusion: The study demonstrates that ADC design significantly influences intracellular trafficking and processing. The linker design, in particular, plays a major role in determining the intracellular fate of an ADC.

Keywords: ADC; HER2; RAB5; T-DM1; T-DXd; breast cancer; mechanism of action.

Graphical abstract

Figure 1.

Cellular sensitivity to T-DXd does not follow T-DM1. Relative viability (MTT) of SK-BR-3, SKOV-3, AU-565, HCC1954, and MDA-MB-453 following 72 h treatments with T-DXd and T-DM1. The curves represent the average of the minimum of three independent experiments represented by the data points. Data on T-DM1 is based on source data from Engebraaten et al. T-DM1, ado-trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.

Figure 2.

T-DXd sensitivity correlates to RAB5A protein expression, but not HER2. Linear regression analysis curves between T-DXd sensitivity and HER2 (a), RAB5A (b), and HER2 × RAB5A (c) protein expression. The log of the 1/IC₅₀ value for each cell line represents the mean of three experiments. Enhancement effect was compared by dividing the IC₅₀ of the payload equivalent on the IC₅₀ of the relevant ADC (d). Raw data of HER2 and RAB5A protein expression is collected from Engebraaten et al. and quantification of the western blot relative to γ-tubulin is shown in (e). ADC, antibody-drug conjugates; T-DXd, trastuzumab deruxtecan.

Figure 3.

T-DXd and T-DM1 display similar cell binding. Furthermore, AF488-trastuzumab and mCherry-RAB5A colocalize in endocytic vesicles. (a) SK-BR-3 was incubated with 10 nM of T-DXd or T-DM1 and subsequently fixed and detected using an AF488 labeled-anti-human antibody (green). Scale bar = 5 µm. (b) In vitro sensitivity of SK-BR-3 cells to T-DM1 and T-DXd (beige) in combination with 100 nM trastuzumab (green). The ADCs were subjected to 24 h incubation in a cell-free medium prior to incubation, and cell viability was measured after 72 h incubation by MTT. (c,d) Live cell time laps of SK-BR-3—mCherry-Rab5A cells incubated with AF488-trastuzumab. Images were captured in 10-min intervals for 3 h starting 5 min after AF488-trastuzumab incubation. Mean fluorescence intensity of AF488-trastuzumab colocalized with mCherry-RAB5a is shown in (c). The mean is denoted in bold, with each experiment (n = 3) shown as stippled lines. Representative sequential images are displayed with the time in minutes shown in the top right corner (d). Scale bar = 10 µm. ADC, antibody-drug conjugates; T-DM1, ado-trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.

Figure 4.

RAB5 impact T-DXd and T-DM1 cytotoxicity but not uptake of trastuzumab. Relative viability (MTT) of cells after KD of RAB5ABC in SK-BR-3 cells and subsequent 72 h treatment of 1 nM T-DXd (a) or 0.1 nM T-DM1 (d). Relative viability is shown with the bars representing the mean of the data and points representing the individual experiments. Statistical significance was determined by a one-way repeated measures analysis of variance followed by a Bonferroni ad hoc test. KD efficiency was verified by western blots and RAB5 was quantified relative to total protein ((c) and (f)). The bars represent the mean of three experiments. Representative western blots are shown and complete blots and full timeline of the KD can be found in Figure S3. (b, e) The relative viability (MTT) of SK-BR-3/Rab5a-high cells after 72 h incubation with 0.01 nM T-DXd or T-DM1. Relative viability is shown with the bars representing the mean of the three indicated data points from each experiment. Statistical significance is determined by a Student’s t-test. (g) Western blot and quantification verifying increased RAB5A expression in SK-BR-3/Rab5-high cells. The bars represent the mean of three experiments and representative western blots are shown. (h) RAB5ABC KD in SK-BR-3 cells were verified by western blot and left for 48 h prior to Alexa-568-Transtuzumab incubation (15 min incubation followed by 2 h chase) before they were investigated by flow cytometry and fluorescence microscopy. The KD efficiency at 48 h is presented as the average of three independent experiments and a representative western blot is shown. Median cellular fluorescence of Alexa Fluor 594 is shown as the average of three experiments with each experiment denoted as data points. Representative microscope images are displayed. Error bar: SD. Scale bar = 10 µm. *p ⩽ 0.05, **p ⩽ 0.01. ***p ⩽ 0.001. n.s., p > 0.05. ADC, antibody-drug conjugates; KD, knockdown; T-DM1, ado-trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.

Figure 5.

RAB4A impacts T-DXd, but not T-DM1 cytotoxicity. Relative viability (MTT) of SK-BR-3 cells after KD of RAB5ABC and RAB4A and 72 h treatment of 1 nM T-DXd (a) or 0.1 nM T-DM1 (b). Bars represent the mean of three independent experiments indicated as data points. Statistical significance was determined by a one-way repeated measures analysis of variance followed by a Bonferroni ad hoc test. (c) KD efficiency verified by western blots (representative blots shown) quantified relative to total protein. Complete blots can be found in Figure S3. The bars represent the mean of three experiments, and the error bar is the SD. *p ⩽ 0.05. **p ⩽ 0.01. n.s., p > 0.05. KD, knock down; T-DM1, ado-trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.

Figure 6.

Effect of RAB5 silencing on T-DM1 and T-DXd effect. Graphical suggestion of how RAB5 silencing could affect intracellular trafficking of T-DXd and T-DM1 and hence efficacy. RAB5 silencing can lead to halted or incomplete maturation of endosomes. The effect of T-DXd will be less influenced than T-DM1 due to the nature of the linker. T-DM1 is reliant on trafficking to the lysosome for the payload to escape and have an effect. T-DXd is exposed to cathepsins throughout the endosomal pathway, and the payload will still be able to act upon its target even when endosomal maturation is faulted. Source: Created with BioRender.com. T-DM1, ado-trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.