Trastuzumab sensitivity is dependent on HER2 status and HER2-low breast cancer cells are not sensitive to trastuzumab

We determined the HER2 status of a panel of different breast cancer cell lines by IHC and FISH, which revealed considerable heterogeneity amongst cell lines (Fig. S1A, S1B). Both SK-BR-3 and BT474 showed strong IHC staining 3+ and increased HER2 gene amplification, whereas BT20 and MCF-7 cells showed weak to undetectable IHC staining (1+ and 0) and a normal FISH status. The other four cell lines (MDA-MB-361, HCC-1569, ZR-75-1, MDA-MB-453) had an intermediate IHC 2+ staining and among these cell lines MDA-MB-361 and HCC1569 had a FISH ratio of ≥ 2 and therefore were considered as HER2 positive for FISH amplification. Interestingly, MDA-MB-453 showed weaker IHC staining than ZR-75-1 but slightly higher FISH/CEP17 ratio (Fig. S1A, S1B).

The growth inhibitory effects of anti-HER2 treatment in IHC 2+ breast cancer cell lines were assessed via cell viability and compared to anti-HER2 treatments in IHC 3 + SK-BR-3 and BT474 cells (Fig. S1C). The MCF-7 cells that express HER2 at a negligible level were used as a negative control when necessary. Increased trastuzumab dosages showed no effect on cell viability in MCF7 (IHC 0) and BT20 cells (IHC 1 + ) (Fig. S1C). Trastuzumab, on the other hand, resulted in a slight to moderate reduction in cell viability in four HER2 IHC2+ breast cancer cell lines compared to IHC 3 + BT474 and SKBR3 cells that were sensitive to trastuzumab (IC50 < 10 μg/ml) (Figure S1C). Due to intermediate response to trastuzumab, MDA-MB-361 (IHC HER2 2 + /FISH HER2 + ) and MDA-MB-453 (IHC HER2 2 + /FISH HER2-) breast cancer cell lines were chosen for our subsequent studies.

Effect of ADAM10/17 inhibitor on MDA-MB-361 (IHC HER2 2 + /FISH + ) and MDA-MB-453 (IHC HER2 2 + /FISH-)

We assessed the effect of an increasing dose of trastuzumab for 24 h in HER2-low MDA-MB-453 cells to investigate the reason of their insensitivity to trastuzumab. We found an increased activation of EGFR and HER2 as well as HER3 and HER4 receptors at higher concentrations of trastuzumab (especially 100ug/ml which was statistically significant) in MDA-MB-453 cells compared to the control group (UT) (Fig. 1A). To investigate whether the HER receptor activation is ligand-dependent, we assessed the levels of HER ligands in the media of MDA-MB-453 cells treated with an increasing dose of trastuzumab. We showed an increase in several HER ligands (NRG-1/heregulin, betacellulin and TGF-α) with an increased concentration of trastuzumab treatment for 24 h (Fig. 1B, C). Our laboratory has previously established ADAM10 and ADAM17 as the two important metalloproteases that cleave HER ligands and counteract the action of trastuzumab in HER2 overexpressing BT474 and SK-BR-3 cells [21, 22]. Therefore, we proceeded to assess the effect of a dual ADAM 10/17 inhibitor (INCB7839) in HER2-low MDA-MB-453 breast cancer cells. We showed that the dual ADAM10/17 inhibitor monotherapy slightly decreased the cell viability to 70% compared to the control and a minimally additive effect was seen when treated in combination with trastuzumab (Fig. 1D). However, the treatment effect was moderate, insufficient to decrease the cell viability below 50%, and this was not statistically significant.

Fig. 1: Trastuzumab induces activation of HER receptors as well as ADAMs and ligands.

A MDA-MB-453 cells were treated with increasing doses of trastuzumab and control (IgG) for 24 h and cell lysates used for western blot and quantification of blots from three independent experiments is shown as indicated (phosphorylated proteins relative to the respective total proteins). Two-way Anova with Turkey multiple comparison test was done to determine statistically significant changes represented as *p ≤ 0.05 and **p ≤ 0.01. B MDA-MB-453 and MDA-MB361 cells were given same treatment as indicated, media was concentrated and used for ELISA and (C) western blot as indicated. D For cell viability experiment, 10,000 cells were seeded per well in 96 well plate and left to settle overnight before being treated with 40 µg/ml of trastuzumab, 100 nM of the anti-ADAM10/17 inhibitor, 10 nM of neratinib and or the indicated combination for 5 days in serum-reduced media. Differences of the treatment groups in comparison to control were analysed using Kruskal-Wallis-Test and p-values are denoted as *p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001.

Thus, our results indicate that trastuzumab induces the release of HER ligands and activation of HER receptors and trastuzumab monotherapy is not effective in decreasing cell growth in HER2-low MDA-MB-453 breast cancer cells. The dual ADAM10/17 inhibitor as a monotherapy or in combination with trastuzumab showed only a slight decrease in the cell viability of these HER2-low breast cancer cells compared to trastuzumab alone.

PanHER inhibitor neratinib monotherapy is effective in trastuzumab insensitive MDA-MB-361 and MDA-MB-453 and shows additive inhibitory effect with trastuzumab

Since we showed that trastuzumab induces HER ligand release, resulting in the activation of HER receptors in MDA-MB-453 cells, we hypothesized that a panHER inhibitor such as neratinib could be effective in these cells due to its effectiveness against HER2 positive breast cancer cells [24]. Neratinib monotherapy was able to significantly inhibit cell viability compared to control or trastuzumab, and its combination with trastuzumab was also the most potent combination (Fig. 1D). Next, we examined the effect of neratinib on HER receptors and the downstream pathways in MDA-MB-361 and MDA-MB-453 cells, treated with increasing doses of neratinib for 24 h. We found decreased activation of EGFR, HER2 and HER3 at 10 nM concentrations but to a lesser extent with 5 nM and highest inhibition was seen at 30 nM (Fig. S2A and S2B). However, there were only minimal inhibitory effects of neratinib on the Akt and Erk activation pathways at 10 nM concentration, but greater inhibition was observed at 30 nM concentration (Fig. S2A, S2B). There was also a gradual decrease of total HER2 with increased concentrations of neratinib, which is consistent with a previously published report that neratinib induced HER2 ubiquitylation and endocytic degradation via its dissociation with HSP90 in HER2 positive breast cancer cells [25]. We further showed that HSP90 dissociation with HER2 was also induced by neratinib in HER2-low breast cancer cells (Fig. S2C), similar to those reported in HER2 positive breast cancer cells [25].

Since HER ligand release would result in HER2 dimerization with other HER receptors in addition to the activation of HER receptors, we compared the effect of pertuzumab that inhibits HER2 dimerization with neratinib or trastuzumab treatment, either alone or in combination with each other. To investigate the best drug combination, we examined the effect of trastuzumab at 40 µg/ml, pertuzumab at 20 µg/ml, neratinib at 10 nM either alone or in combination with each other for 24 h on HER receptors and downstream pathways. In both cell lines, there was evidence of HER2 downregulation with neratinib or neratinib containing combinations (Figs. 2A, B). Neratinib alone or its combination with trastuzumab and/or pertuzumab decreased pHER2 in both cell lines (Fig. 2A, B). In addition, there was deactivation of EGFR, HER3 and Akt by neratinib alone or in combination with other anti-HER2 treatments, although there were differences between the two cell lines. The greatest inhibition of EGFR, HER3 and Akt activation was seen with neratinib with trastuzumab and/or pertuzumab in MDA-MB-453 cells (Fig. 2B). In MDA-MB-361 cells, neratinib alone or in any combination could decrease pEGFR, pHER3 and pAkt but the greatest inhibitory effect of pEGFR was seen with the triple combination (Fig. 2A). There was a decrease of pErk in neratinib containing treatment conditions in MDA-MB-361 cells but the greatest inhibition of Erk was observed with the triple combinations in MDA-MB-453 cells (Fig. 2A, B).

Fig. 2: Reduced activation of HER receptors and downstream signalling proteins in MDA-MB-453 and MDA-MB361 cells with the combination anti- HER2 treatments.

A MDA-MB-361 and (B) MDA-MB-453 cells were treated with 10 nM Neratinib, 40 µg/ml trastuzumab and 20 µg/ml of Pertuzumab for 24 h either alone or in combination. Cell lysates were used for western blot as indicated and quantification of four blots is shown relative to the respective total proteins. C MDA-MD-361 and MDA-MB-453 cells treated with 10 nM neratinib, 40 µg/ml trastuzumab and 20 µg/ml of Pertuzumab for 5 days in serum-reduced media. Cell proliferation was assessed by cell titre Glo 5 days post treatments. All Data is from 3 independent experiments with 3 technical replicates and is normalized to the control (untreated or DMSO + IgG). Graphs plotted represent mean and error bars ± SEM. Difference in the mean between groups was analysed by one-way Anova with Turkey multiple comparison test; statistically significant changes are represented by asterisks as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p < 0.0001.

Since we were unable to observe any major differences in the inhibition of EGFR, HER2 and HER3 receptors with neratinib monotherapy at 10 nM or in combination with trastuzumab ± pertuzumab, we hypothesized that the inhibitory effect due to higher concentration of neratinib is most likely to be too effective and not enhanced by its combination with other anti-HER2 treatments. Therefore, we proceeded to investigate the effect of neratinib at a lower concentration at 5 nM alone and in combination with trastuzumab ± pertuzumab. We found that the decreased activation of EGFR, HER3 and Akt with the combination of neratinib and trastuzumab or triple combination to be more effective than neratinib 5 nM monotherapy in MDAMB-361 cells and to a lesser extent in MDA-MB-453 cells. We only observed a decrease in pErk with triple combination in both cell lines (Fig. S3A, B).

Next, we tested the effect of neratinib, pertuzumab or trastuzumab either alone or in combinations on the cell proliferation/viability. In both MDA-MB-361 and MDA-MB-453 cells, the more effective combinations were neratinib with trastuzumab or the triple combination, but neratinib with pertuzumab was also effective in FISH positive MDA-MB-361 cells (Fig. 2C, S4A, S4B, 4C). The combination of pertuzumab with trastuzumab was not effective in decreasing cell growth below 50% compared to DMSO + IgG control in both cell lines, which is in contrast with their effect on HER2 over-expressed breast cancer cell lines [26], correlated with the inability to decrease pHER2 and to downregulate HER2 at 24 h in these cells (Fig. 2A, B). Overall, neratinib and trastuzumab combination was more effective than neratinib or trastuzumab monotherapy in both cell lines, which is consistent with our biochemical data in combination with lower concentrations of neratinib at 5 nM (Fig. S3A, S3B). The additive effect of combination treatment appeared to be reduced with high doses of neratinib in MDA-MB-453 cells (Fig. S4D).

Collectively, the above observations show that neratinib but not trastuzumab, is effective in decreasing the activation of HER receptors and downstream pathways as well as inhibiting cell growth in MDA-MB-453 and MDA-MB-361 breast cancer cell lines. However, the combination of neratinib with trastuzumab (with or without pertuzumab) was more effective than neratinib monotherapy in inhibiting cell growth, consistent with greater inhibition of HER receptor activation and HER2 downregulation in both cell lines. Thus, our data provide a rational for the use of neratinib in combination with trastuzumab (with or without pertuzumab) in the treatment of HER2-low breast cancer cells.

Comparing the effect of neratinib with tucatinib in HER2-low MDA-MB-453 cells and HER2 negative MCF-7 cells

Since tucatinib was shown to be a more HER2 selective inhibitor, we compared the effect of tucatinib with neratinib on HER receptor signalling and cell growth. At the same 10 nM concentration, we showed that neratinib was more effective than tucatinib in inhibiting pHER2, pEGFR and pHER3 at 4 h in MDA-MB-453 cells (Fig. S5A). We further compared neratinib and tucatinib at 10 nM concentration (± trastuzumab) and showed that neratinib ± trastuzumab was more effective in decreasing pHER2 than tucatinib ± trastuzumab at 24 h but the difference in inhibition was not significant for pEGFR, pHER3 and pHER4 in MDA-MB-453 cells (Fig. S5B). This is also consistent with the cell viability result with neratinib ± trastuzumab being more effective than tucatinib ± trastuzumab although the combination of neratinib with trastuzumab was the most effective among all the treatment conditions in inhibiting cell viability below 50% compared to the control (Fig. S5C). We could not detect any obvious differences between neratinib and tucatinib in inhibiting HER receptor signalling in MCF-7 cells which were only minimally responsive to either drugs or their combination with trastuzumab (Fig. S5B, Fig. S5C).

Development of a panel of patient-derived breast cancer organoids and histological characterization

In view of the limitations of 2D cell lines models in translating laboratory findings to clinical benefit for patients, we established a panel of breast cancer patient-derived organoids (PDOs) from surgically resected breast tumours in view of the potential of PDOs in predicting treatment response in patients [23]. Using a published protocol [23], we developed a panel of breast cancer PDOs with varying HER2 and ER/PR expression (Table 1) in order to test the effects of different anti-HER2 treatments and/or a ADAM10/17 inhibitor.

Table 1 The histological features of the human samples from which the organoids were derived from.

To test whether the breast cancer PDOs morphologically reflect the derived tumour, we performed histopathological analysis of H&E and IHC receptor status on the stained tissue and its corresponding organoid sections of PDOs. The phenotype of a PDO (T-S403276) showed similar morphology to the architecture of the derived tumour (Figure S6A). In addition, this breast cancer organoid represented clear cancerous features like the formation of tubules, enlarged pleomorphic nuclei and high mitotic activity as represented in H&E-stained sections (Fig. S6B). Compared with the cancerous tissue, the corresponding normal organoid (N-S403275) established from surrounding normal tissue of the tumour, displayed a well-organized structure and mildly complex cribriform architecture, and was histologically judged to display normal breast epithelium (Fig. S6A, S6C).

Beside histological conservation, PDOs tend to recapitulate the expression of some of the most important diagnostic and prognostic markers including oestrogen receptor (ER), progesterone receptor (PR), and HER2. We found that ER, PR and HER2 status was also retained in the breast cancer organoid as determined by IHC in tumour and its corresponding organoid (T-S403276). Tumour ER/PR and HER2 status corresponded to its derived organoid, and the majority of the PDO structures retained 70% of the staining (Fig. S6B). While the healthy organoids derived from the normal surrounding tissue with negative ER, PR and HER2 status also corresponded to their normal tissue (Fig. S6C). The bright field histological images of compact, coherent organoid structure of T-S403276 and the corresponding normal PDO (N-S403275) as well as another example of discohesive PDO (T-S396358) are shown in Fig. S6D.

In summary we found that the breast cancer organoid, TS403276, recapitulated its originating tumour tissue architecture and histological features, as well as hormone receptor and HER2 status.

Effects of neratinib, trastuzumab, ADAM10/17 inhibitor in a panel of PDOs

We established above that neratinib alone or in combination with trastuzumab significantly reduced the cell viability of MDA-MB-453 and MDA-MB-361 cells. Since PDOs have been shown to better recapitulate the tumour characteristics and may help to predict treatment response [23], we assessed the effects of neratinib with or without trastuzumab in a panel of PDOs. Firstly, the panel of PDOs with different HER2 and ER/PR expression were treated with the increasing doses of neratinib monotherapy for 5 days and DMSO was used as a control. We observed a dose dependent decrease in cell viability in a panel of PDOs (Fig. 3A–G), with the greatest sensitivity being observed in HER2 positive PDO (T- S403276) with an IC50 of 78.9 nM, followed by HER2 equivocal (PDO T-S386416) with an IC50 of 154.4 nM and then five HER2 IHC negative or 1+ PDOs (T-S396358, T-S396391, TS420880, TS410460 and TS358174) with IC50s 250-300 nM (Fig. 3A–G, Table 1). Furthermore, we also treated healthy organoids derived from the normal surrounding tissues with an increasing dose of neratinib (Fig. 3H–J). We observed no sensitivity of neratinib in these normal organoids even at higher dose of 1uM and no change was observed in cell viability when compared to the control group (Fig. 3H–J).

Fig. 3: Patient-derived breast cancer organoids with different HER2 status are sensitive to neratinib.

A panel of breast cancer organoids (A) T-S403276 (HER2 +ve, IHC 2+ and FISH +ve), (B) T-S386416 (HER2 equivocal, IHC 2+ and FISH -ve), (C) T-S396358 (HER2 -ve), (D) T-S396391 (HER2 -ve), (E) T-S410880 (HER2 –ve), (F) T-S410460 (TNBC) (G) T-S358174 (HER2-low)] and their corresponding normal organoids (H) N-S410879, (I) N-S403275 and (J) N-S388920) were treated with an increasing concentration of neratinib for 5 days to obtain dose response curves. Cell proliferation was assessed using 3D cell titre Glo and graph is normalized to control (DMSO + IgG). GraphPad Prism 8 software was used to plot the curves and obtain R-squared and EC50 values. Data presented from 3 independent experiments (n = 3) with 3 technical replicates; error bars are shown as the median ± range.

We assessed the effect of increasing doses of neratinib in combination with trastuzumab in another PDO T-S396358 (HER2 IHC 0 and ER/PR + , Table 1) and showed that trastuzumab enhanced the effect of neratinib at higher doses of 100 nM and 300 nM of neratinib (Fig. S7A). We also treated one less sensitive TNBC organoid (T-S358174, low HER2 and ER/PR negative) with high doses of neratinib and in combination with trastuzumab. The results showed that neratinib monotherapy 500 nM could decrease the cell viability to around 50% compared to the control, and the addition of trastuzumab did not result in any statistically significant difference (Fig. S7B). However, higher dose of 1uM neratinib, was toxic and displayed necrotic structures in less sensitive PDO (T-S358174), which was enhanced by trastuzumab treatment (Fig. S7C). We also compared the effect between neratinib and tucatinib in a HER2-low PDO (T-S485871, HER2 FISH 1+ and ER positive) and showed that neratinib was more effective than tucatinib (Fig. S5D).

In a further experiment, five additional PDOs (T-S403276, T-S396358, T-S396391, T-S420880, T-S410460) did not display any sensitivity to trastuzumab compared to the control (Fig. 4A–E). However, neratinib as a single agent at a dose of 300 nM (>IC50 doses for the five PDOs) decreased the cell viability by 50%. The addition of trastuzumab was slightly more effective than neratinib alone but this was not statistically significant compared to neratinib alone in all the PDOs (Fig. 4A–E). Furthermore, we observed no effects of any of the treatments in three normal PDOs as expected (Fig. 4F–H).

Fig. 4: Combination of trastuzumab, neratinib, ADAM10/17 inhibitor INCB7839 or their combination in a panel of patient-derived organoids.

A panel of breast cancer organoids some of their corresponding normal organoids were treated with neratinib (300 nM), trastuzumab (40 μg/ml) and ADAM10/17 inhibitor INCB7839 (300 nM) for 5 days and cell viability was assessed using 3D cell titre Glo and the results were normalized to control (DMSO + IgG). Data presented from 3 independent experiments (n = 3) with 3 technical replicates; data is shown as the median ± range; non-parametric Kruskal Wallis test was performed p values are denoted as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p < 0.0001.

We also proceeded to investigate the effect of dual ADAM10/17 inhibition using INCB7839 in the panel of PDOs in view of the results seen in HER2-low breast cancer cells. We assessed the effects of the ADAM10/17 inhibitor together with neratinib in five PDOs, (T- S403276, T-S396358, T-S396391, T-S420880, T-S410460) (Fig. 4A–E). The treatment with the ADAM10/17 inhibitor monotherapy showed no or minimal sensitivity compared to the control, and the combination of the ADAM10/17 inhibitor and trastuzumab displayed some additive effects but were not statistically significant (Fig. 4A–E). These results are consistent with the modest inhibitory effect observed in the dual ADAM10/17 inhibitor ± trastuzumab in HER2-low breast cancer cells.

Trastuzumab induces antibody-dependent cell cytotoxicity (ADCC) response mediated by natural killer (NK) cell in PDOs

Trastuzumab is a potent mediator of antibody-dependent cell-mediated cytotoxicity (ADCC). Studies have shown that the trastuzumab-induced ADCC response, mediated by NK cells, is preferentially exerted on HER2 overexpressing cancer cells compared to those that do not overexpress HER2 [27]. Therefore, we proceeded to investigate trastuzumab induced NK cell-mediated ADCC in two breast cancer PDOs (one was HER2 IHC 2+ and FISH positive, like in MDA-MB-361 cells; and another one IHC 0).

Firstly, the PDOs T-S403276 (HER2 IHC2+ and FISH positive, ER/PR + ) and T-S446358 (HER2 0, ER/PR + ) were successfully co-cultured with NK cells isolated from healthy donors. Furthermore, the PDOs were treated with increasing doses of trastuzumab and neratinib monotherapy for 5 days, and DMSO + IgG was used as a control. Here we observed a dose-dependent decrease in cell viability with trastuzumab or neratinib monotherapy in the presence of NK cells in both PDOs (Supplementary Fig. 8A, B). The PDO T-S403276 seemed to be more sensitive to neratinib or trastuzumab monotherapy than PDO T-S446358, showing 50% decrease in cell viability with 20 μg/ul trastuzumab and 50 nM of neratinib. In contrast, 50% decrease in cell viability was only observed in ≥100 nM neratinib and 100 μg/ul trastuzumab compared to the control group (DMSO + IgG) in T-S446358 PDO (Supplementary Fig. 8A, B).

Similarly, in a comparative study between neratinib and trastuzumab treatment groups, PDO T-S403276 displayed significant decrease in cell viability with the combination treatment, neratinib and trastuzumab, in the presence of NK cells compared to the combination treatment without NK cells (Supplementary Fig. 8C). While PDO T-S446358 still shows that the combination treatment of neratinib and trastuzumab was effective compared to trastuzuamb and neratinib monotherapy, however no significant difference was observed in the presence of NK-cells (Supplementary Fig. 8C).

In a further study, NK cells were treated with a CD16-specific blocking antibody to assess whether the anti-tumour effects of trastuzumab observed in this study were related to ADCC. We observed no significant change in cell viability in neratinib group treated with anti-CD16 compared to those without anti-CD16 (Supplementary Fig. 8D). However, in the groups of trastuzumab alone or in combination with neratinib, the anti-proliferation effect was decreased in the presence of a CD16-specific blocking antibody, suggesting a possible ADCC mechanism of response.

In summary, trastuzumab monotherapy and or neratinib monotherapy displayed a dose-dependent decrease in cell viability in the presence of NK cells. In addition, neratinib in combination with trastuzumab, was the most effective treatment in the presence of NK cells reducing the cell viability by more than 80%. Furthermore, antitumour effects of trastuzumab are mediated by NK-cell induced ADCC.