Read on Twitter
Looking into the literature, it seems like specifically targeting CD47 with an active IgG might maximize the ADCC of the IgG relative to other targets. There are 2 reasons for this: 1. CD47 surface retention and 2. CD47 movement relative to the phagocytic synapse. To the 1/n
first point, when many trans-membrane proteins are engaged by mAbs, the cytoplasmic domain of the TM protein drives endocytosis. This removes the TM protein from the cell surface, and thus the IgG as well, reducing macrophage/DC engagement. Classic examples of these kind of 2/n
targets are Her2 and EGFR. For example, blocking dynamin activity (which is necessary for target endocytosis) has been shown to increase ADCC activity of cetuximab and trastuzumab (Wrighton 2020). What’s interesting about CD47 is that it’s endocytosis appears not to be 3/n
driven by its own cytoplasmic domain, but by the cytoplasmic domain of an interacting TM protein present on a neighboring cell, Sirpa. A study done in CHO cells expressing mouse Sirpa/CD47 showed that amino acids 405-424 of the cytoplasmic portion of Sirpa drove CD47 4/n
endocytosis on an interacting cell (Matozaki et al. 2008). Interestingly, $TRIL 621 and 622 Sirpa-IgG fusion proteins only contain amino acids 31-148 of the extracellular domain of human Sirpa. The homologous region of this in a mouse Sirpa construct could not efficiently 5/n
generate CD47 endocytosis in the absence of the cytoplasmic domain (Matozaki et al. 2008). Thus, it is possible that 621/2 have greater retention on the cell surface than approved mABs such as Cetuximab and Trastuzumab, potentially resulting in greater ADCC activity per IgG. 6/n
To the second point, it has been found that CD47 turns off the formation of nascent phagocytic synapses on macrophages via ‘dragging’ Sirpa to the synapse, where Sirpa can inhibit synapse formation via local signaling pathways (Morrissey et al. 2020). Thus, it is plausible 7/n
that when an IgG effector domain is anchored to CD47, it will ‘trick’ CD47 to ‘drag’ it into the synapse, where it can directly act at the critical decision point juncture for the macrophage. In composite, CD47 seems to have unique properties that maximize the activity of 8/n
bound, active IgGs. Personally very interested in how this would play out when 622 combos with mAbs that contain IgG1. Will the IgG1 initiate a synapse that will then concentrate 622 at the synapse where the weak IgG4 in aggregate will signal for phagocytosis, 9/n
even after mAb endocytosis? 622 single agent activity has already demonstrated that 622 can activate phagocytosis in the absence of additional signals (although it is likely harder for the weak IgG4 to generate synapses de-novo). When you think in terms of the mechanism of 10/n
concentrating IgG4 at newly formed phagocytic synapses, it is hard to believe imo that combos with drugs that will form novel synapses won’t increase efficacy. We will see with the combo trial results later this year. 11/11
