LAUSR

444 | JulY 2020 | volume 19 Delivery of protein therapies to the brain has proven difficult because of the need to cross the blood–brain barrier (BBB). Now, two recent papers from researchers at Denali Therapeutics describe a novel modular platform capable of trafficking protein therapeutics to the brain through binding to transferrin receptor (TfR). Approaches that exploit TfR to enable proteins to cross the BBB through receptormediated transcytosis (RMT) have previously been investigated using constructs such as bispecific antibodies, with one ‘arm’ binding to TfR and one arm binding to the desired target in the central nervous system (CNS). In their study, Kariolis et al. describe the develop ment of a novel BBB transport vehicle (TV) that exploits RMT by TfR, based on the Fc domain of a human IgG1 antibody that has been modified to allow specific binding to TfR. Binding to TfR and to a CNS target can be achieved through fusion of antigenbinding regions to the TfRbinding Fc domain, producing an IgG1like construct that possesses the biophysical properties and desirable pharmacokinetics of typical monoclonal antibodies, which the authors called an antibody transport vehicle (ATV). Alternatively, the Fc domain can be fused to a therapeutic protein such as an enzyme to produce an enzyme transport vehicle (ETV). The first target chosen for proof of concept was βsecretase 1 (BACE1). BACE1 is involved in the cleavage of amyloid precursor protein to amyloidβ 40 (Aβ40), aggregates of which contribute to the pathology of Alzheimer disease. The authors fused the TfRbinding Fc fragment clones to antiBACE1 antibody Fab arms to produce a construct named ATV:BACE1. In vitro studies showed ATV:BACE1 variants were only internalized by cells expressing TfR — indicating TfRmediated uptake of the TV. To explore the in vivo activity of ATV:BACE1, the authors used a mouse model expressing a chimeric murine/human TfR containing the hTfR apical domain (TfRmu/hu mice). TfRmu/hu mice given intravenous ATV:BACE1 had levels of human IgG1 in the brain 40fold higher than those given antiBACE1 antibodies. In addition, ATV:BACE1 treatment reduced soluble brain Aβ40 levels 57% more than antiBACE1, without reducing TfR expression in the brain. Studies in cynomolgus monkeys similarly showed that treatment with ATV:BACE1 resulted in the accumulation of human IgG1 in the brain and reduced levels of Aβ40 in the cerebrospinal fluid (CSF) and in various brain regions. These data suggest ATV:BACE1 successfully crosses the BBB in these species and can exert therapeutic effects. To demonstrate the modular aspect of the ATV system, the authors constructed a bispecific ATV structure with two distinct Fab fragments — one specific for BACE1, the other for Tau protein. Addition of the ATV to cultures of cortical neurons expressing TfR and BACE1 resulted in a 50% reduction in Aβ40 concentrations. The ATV also reduced Tau aggregation in brain lysates from Alzheimer disease patients, suggesting the ATV platform could allow for modulation of two distinct therapeutic targets. In a concurrent study, Ullman et al. examined the use of a TV for treating a lysosomal storage disease known as mucopolysaccharidosis type II (MPS II), or Hunter syndrome. This disease is caused by a deficiency in the lysosomal enzyme iduronate 2sulfatase (IDS), which is needed to break down glycosaminoglycans (GAGs). The authors used a TV composed of the TfRbinding IgG1 Fc domain tethered to a molecule of IDS (ETV:IDS). In IDSknockout (KO) HEK293 cells and MPS II patient fibroblasts, administration of an ETV:IDS construct was as effective at reducing GAG levels as an existing IDS enzyme replacement therapy, idursulfase, which is not able to address the CNS manifestations of MPS II because of its lack of BBB penetration. Importantly, administration of ETV:IDS to IDSKO; TfRmu/hu mice — which have greatly elevated GAG levels — reduced GAG levels in the brain and CSF. Although GAG levels are a marker of IDS activity, the accumulation of lysosomal lipids and prolonged activation of glial cells ultimately drive the symptoms of lysosomal storage diseases such as MPS II. The authors found intravenous ETV:IDS reduced the levels of a range of lysosomal lipids in IDSKO; TfRmu/hu mice, whereas idursulfase treatment did not. The expression of CD68 and TREM2 — markers of glial cell activation — were elevated in IDSKO; TfRmu/hu mice, and four weekly doses of ETV:IDS reduced levels of CD68 and TREM2. These data show ETV:IDS is able to reverse the MPS IIlike phenotype seen in IDSKO mice. These studies show that TVs are modular structures that can deliver antibodies or other proteins to the brain in order to modulate therapeutic targets in the CNS. ETV:IDS is entering clinical trials for proof of concept in humans, and the authors are working to apply the TV platform to a broad range of biotherapeutics.