Human epidermal growth factor receptor 2 (HER2 or ErbB2) plays a critical role in cell signaling and its deregulation is involved in many cancers. It is a member of the ErbB family of tyrosine kinase receptors. There are four receptors in the family: epidermal growth factor (EGF) receptor 1 (also termed ErbB1/HER1), ErbB2/Neu/HER2, ErbB3/HER3 and ErbB4/HER4. ErbB receptors can form homo- and heterodimers mediated by EGF-related peptide growth factors that bind to their extracellular domain (Olayioye et al., 2000). No ligand has been identified for HER2, and even if some papers suggest that it can form homodimers (Peckys et al., 2015) (Ghosh et al., 2011), it is more commonly described as the favored heterodimerization partner for the other ErbB receptors (Karunagaran et al., 1996) (Graus-Porta et al., 1997) (Ghosh et al., 2011). The heterodimers involving HER2 have higher signaling potency than heterodimers involving other ErbB receptors (Karunagaran et al., 1996). HER2-involving signaling is directly linked with cell growth and, when HER2 is overexpressed, with several cancers. Pro-oncogenic HER2 heterodimers with other HER receptors form through HER2 extracellular region (hereafter sHER2), leading to autophosphorylation of its intracellular region and downstream signaling. Dimerization and pro-oncogenic effects can be inhibited by antibodies directed against sHER2, prominently two monoclonal anti-sHER2 antibodies, trastuzumab (Herceptin) and pertuzumab (Perjeta). Trastuzumab binds to domain IV of sHER2 (see below) and was first approved by the Food and Drug Administration (FDA, 1998) and by the European Medicine Agency (EMA) in 2000 for its use against HER2-positive breast cancer (EMA, 2000). Fab (Fragment antigen binding) related effects of trastuzumab are manyfold and include cytostatic effect, angiogenic factor regulation, antibody-mediated aggregation (Ben-Kasus et al., 2009) and internalization of HER2. Trastuzumab also inhibits a proteolytic cleavage of HER2 to a shorter form p59HER2 that is active in signaling (Molina et al., 2001). Its Fc related effects are antibody-dependent cellular cytotoxicity (ADCC) and IgG1 half-life increasing, the latter being important for favorable trastuzumab pharmacokinetics. However, acquired trastuzumab resistance has appeared over time, which needs to be overcome (Pohlmann et al., 2009). Pertuzumab has taken this role. Like trastuzumab, pertuzumab is a HER2 targeting humanized monoclonal antibody. It was first approved by the FDA in 2012 (F. and D.A., 2012) and by the EMA in 2013 (EMA, 2013). Pertuzumab binds to domain II of sHER2 and its main Fab effect is to sterically block the HER2 dimerization process (Nami et al., 2018). Combination of trastuzumab and pertuzumab displays a synergistic effect (Nami et al., 2018, Nahta et al., 2004, Richard et al., 2016) and therefore is currently used as a very effective immunotherapy strategy against breast cancer (Scheuer et al., 2009, Blumenthal et al., 2013, Howie et al., 2019, Liu et al., 2022).

Accordingly, the molecular basis of trastuzumab and pertuzumab inhibition of HER2 has been the subject of much basic research. Particularly the atomic structures of complexes of sHER2 with Fabs have been reported (Table 1), first by X-ray crystallography for sHER2-trastuzumab (Cho et al., 2003) and for sHER2-pertuzumab (Franklin et al., 2004). In recent years HER2 structural biology has taken advantage of the ongoing revolution in cryogenic electron microscopy (cryo-EM) to provide atomic-level structures of large protein complexes. Thus, several important cryo-EM structures of HER2 complexes have been solved by single particle analysis (SPA), particularly the HER2-HER3 and HER2-HER3-trastuzumab (Fab) complexes (Diwanji et al., 2021). A major conclusion of these recent works is that dynamics are important for HER2 dimerization and for regulation of signaling (Diwanji et al., 2021, Bai et al., 2023). However, these dynamics also make HER2 complexes difficult objects for SPA: Not only is sHER2 only 630 residues (70 kDa, not counting glycosylations), but its complexes tend to display several kinds of heterogeneity. This includes continuous conformational variability, that is currently the object of much research in the data processing cryo-EM community (Vuillemot et al., 2023, Punjani and Fleet, 2021). Thus, the only structure of the tripartite complex between sHER2, trastuzumab Fab and pertuzumab Fab (hereafter called the HTP complex) is available only to lowish resolution (Table 1) and in particular does not allow visualizing to atomic resolution the HER2-trastuzumab interface, as this interface lies in the most flexible and least resolved part of HTP (Hao et al., 2019).

The architecture of HER extracellular regions includes four domains, consecutive in sequence, I to IV (Fig. 1a). In sHER2, domains I, II and III pack together while IV forms an elongated stem that connects to the transmembrane helix in full-length HER2. Accordingly, in maps of sHER2 but also of solubilized full-length HER2, the tip of IV is poorly defined and most residues beyond 570 are not resolved, including residues close to the trastuzumab paratope. Indeed, in cryo-EM SPA the most encountered limiting factors are particles distribution and orientation, ice thickness, and sample flexibility (Passmore and Russo, 2016, Weissenberger et al., 2021, Brown and Hanssen, 2022, Vinothkumar and Henderson, 2016, Rice et al., 2018).

As to particles' distribution and orientation, there is a well-known and unpredictable gap between sample preparation and the vitrification process. Often particles are strongly affected by the air–water interface and the more widely used approaches to overcome the associated problems of particle misdistribution and/or damage is to work on the water surface by adding detergent or to add a continuous thin carbon layer support (Chen et al., 2019, Kampjut et al., 2021, Glaeser, 2018).

Even if there is no foolproof method to predict the behavior of particles on the grids and during vitrification, an exhaustive quality assessment of the sample strongly increases success rates (Berrow et al., 2021, Raynal et al., 2021). Here we use such an approach to present a complete cryo-EM structure of the HTP complex that allows simultaneously resolving all HER2-trastuzumab and HER2-pertuzumab interactions. By using state-of-the-art image processing methods, we account for intrinsic flexibility of the complex and describe both its structure and motion.