20 Apr 2017

Tau pathology is linked with neurodegeneration in multiple tauopathies, and efforts are underway to vanquish this tangle-forming protein with anti-tau antibodies. Besides perhaps having a therapeutic effect, these antibodies may open a window to soluble tau in the brain, according to a study published April 19 in Science Translational Medicine. Researchers led by David Holtzman at Washington University in St. Louis reported not only that an anti-tau antibody stabilized tau in the blood in tau-transgenic mice and in people with progressive supranuclear palsy (PSP), but also that levels of plasma tau bound by the antibody reflected the amount of soluble, extracellular tau in the brain.

“The antibody catches tau protein as it comes out of the brain and into the blood,” Holtzman told Alzforum. This suggests the antibody could serve as a blood-based marker to track tau levels in people with neurodegenerative diseases or following neuronal injury, he added.

As part of their goal of stopping the spread of tau aggregates between neurons in the brain, Holtzman and colleagues previously developed tau antibodies, including HJ8.5 used in this study. HJ8.5 is specific for tau’s N-terminus, and reduces tau pathology and cognitive deficits in P301S mice (see Sep 2013 news). Subsequent work revealed that besides striking down levels of hyperphosphorylated and insoluble tau in the brain, intraperitoneal injections of HJ8.5 elevated plasma tau (see Yanamandra et al., 2015). C2N Diagnostics and AbbVie subsequently humanized the antibody, dubbed C2N-8E12, which is being tested in clinical trials in people with AD and PSP.

The current study, by co-first authors Kiran Yanamandra and Tirth Patel and colleagues, is an attempt to get a clearer vision of the antibody’s modus operandi and to understand the relationship between tau in the plasma and brain. In agreement with their previous study, the researchers observed a sizable, dose-dependent increase in plasma tau two days after injecting three-month-old P301S animals with 10 to 200 mg/kg of HJ8.5.

This plasma increase happened in PSP patients injected with C2N-8E12 as well. Following intravenous infusion of either 7.5 mg/kg or 15 mg/kg of antibody in four people, their plasma tau jumped from being undetectable, i.e., below 10 pg/mL, up to 150-300 pg/mL within several days. This is an estimated 50- to 100-fold increase over baseline plasma tau.

The researchers wondered whether the antibody lifted plasma tau by prolonging its half-life. To test this, they injected P301S mice intravenously with recombinant human tau (hTau), and then measured how the presence of the antibody altered its clearance. Injecting HJ8.5 one hour prior to hTau boosted hTau’s plasma half-life from 8.5 minutes to 3.4 hours, while co-incubating hTau with the antibody prior to injection upped it to more than 13 hours. When the researchers injected hTau directly into the CNS via the cisterna magna, they found an uptick of it in plasma tau within minutes. Intravenous antibody injection boosted the half-life of this CNS-derived hTau from 1.6 hours to almost 24 hours. Together, these findings suggested that the antibody stabilized tau in the blood, including tau transported there from the CNS.

Because the researchers observed CNS-derived tau in the blood, they next asked whether plasma tau levels directly correlated with levels of tau in the brain. In support of this idea, they first found that levels of soluble tau in the brains of three different animal models, as measured in cortical brain homogenates, correlated with plasma tau levels observed after antibody injection. To investigate this further, the researchers tracked levels of soluble and insoluble tau in the brains of P301S mice, finding that as the mice aged, insoluble tau increased, while soluble tau went down. They also tracked levels of extracellular, soluble tau in the interstitial fluid (ISF) of living P301S mice. As these mice aged from three to nine months, their soluble ISF tau dropped by 60 percent. Following intraperitoneal injection of the HJ8.5 antibody, plasma tau paralleled soluble ISF tau, measuring 60 percent higher in younger animals.

The relationship between soluble ISF and plasma tau also held true when the researchers stimulated release of soluble tau from neurons by injuring them. An intrastriatal injection of the neuronal toxin 3-NP made soluble tau in the striatum rise by 10- to 20-fold. This acute release of tau showed up as elevated plasma tau levels detected two days following antibody treatment. Together, the findings suggested that, following antibody treatment, plasma tau levels correlate with soluble tau in the brain.

Holtzman suggested that the stabilization of tau in the blood, where it normally undergoes rapid clearance and is difficult to detect, may offer researchers a tool to track changes in soluble tau in the brain. At the same time, it’s less clear what those changes in soluble tau say about ongoing neurodegeneration, the extent of tau pathology, or therapeutic benefit in response to tau antibodies or other drugs, he said. Soluble tau, as measured in the CSF, is known to be elevated in people with AD, but reduced in people with PSP (see Wagshal et al., 2015). 

“These are carefully conducted and informative measurements by a stellar group of investigators,” commented Einar Sigurdsson of New York University School of Medicine. “It will be interesting to see how the tau antibodies in clinical trials may influence tau levels in CSF and plasma, which may differ to some extent between AD and PSP subjects” (see full comment below).

The present findings bear a striking resemblance to those from solanezumab, Eli Lilly’s failed anti-Aβ antibody. Solanezumab also stabilized its target—Aβ—in the plasma. Some researchers suspect that this may have bungled efforts to assess target engagement during early trials, or that the antibody could have unwittingly interfered with natural clearance of Aβ (see Jan 2017 news).

The fact that tau antibodies also stabilize their targets should raise a red flag about their therapeutic use, commented Henrik Zetterberg of University of Gothenburg, Sweden. “To me, that is not necessarily a good thing to do if you want to get rid of a protein,” he said. However, how (or if) accumulation of antibody-bound tau in the plasma relates to tau pathology in the brain is unclear, both he and Holtzman acknowledged. Preclinical studies suggest that treatment with the antibodies reduces tau pathology, despite this plasma tau elevation. Zetterberg added that tau’s protection by the antibody may destroy its usefulness as a marker for neuronal injury in clinical trials of tau antibodies. To measure any therapeutic effects on neurodegeneration in such trials, researchers would need to turn to other markers, such as neurofilament light (NfL) or neurogranin, he said. Alternatively, tau PET tracers are being studied to measure fibrillary tau (see Apr 2017 conference news).—Jessica Shugart

Comments

1. • Einar Sigurdsson
     New York University School of Medicine
   • Posted: 20 Apr 2017

   These are carefully conducted and informative measurements by a stellar group of investigators. It will be interesting to see how the tau antibodies in clinical trials may influence tau levels in CSF and plasma, which may differ to some extent between AD and PSP subjects. Since CSF tau levels increase in AD but not in PSP, extracellular tau may be less important for propagation of tau pathology in PSP than in AD. Thereby, antibodies that only work extracellularly may be less effective in PSP than in AD.

   In addition, since recent mass spectrometry studies indicate that most CSF tau has ragged N- and C-termini and consists mostly of tau fragments around amino acids 150-250 of the protein, it will be interesting to see how antibodies that bind within or outside this main region compare. Again, if the antibody works mainly by an extracellular mechanism, one would think it should preferably bind to this most prominent extracellular region.

   Regardless of the tauopathy or the binding site, the most effective antibodies will likely be those that can access tau in all compartments, including inside the neurons, where most of pathological tau resides.

   View all comments by Einar Sigurdsson

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Antibodies Stop Toxic Tau in Its Extracellular Tracks 26 Sep 2013
2. Solanezumab: Did Aβ ‘Reflux’ From Blood Confound Target Engagement in CSF? 31 Jan 2017
3. Next-Generation Tau PET Tracers Strut Their Stuff 14 Apr 2017

Therapeutics Citations

1. Tilavonemab
2. Solanezumab

Research Models Citations

1. hTau.P301S

Paper Citations

1. Yanamandra K, Jiang H, Mahan TE, Maloney SE, Wozniak DF, Diamond MI, Holtzman DM. Anti-tau antibody reduces insoluble tau and decreases brain atrophy. Ann Clin Transl Neurol. 2015 Mar;2(3):278-88. Epub 2015 Jan 23 PubMed.
2. Wagshal D, Sankaranarayanan S, Guss V, Hall T, Berisha F, Lobach I, Karydas A, Voltarelli L, Scherling C, Heuer H, Tartaglia MC, Miller Z, Coppola G, Ahlijanian M, Soares H, Kramer JH, Rabinovici GD, Rosen HJ, Miller BL, Meredith J, Boxer AL. Divergent CSF τ alterations in two common tauopathies: Alzheimer's disease and progressive supranuclear palsy. J Neurol Neurosurg Psychiatry. 2015 Mar;86(3):244-50. Epub 2014 Jun 4 PubMed.

Further Reading

News

1. Tau Immunotherapy for Aβ? 1 Apr 2017
2. Therapies Take Aim at Tau 15 Aug 2014
3. In First Phase 3 Trial, the Tau Drug LMTM Did Not Work. Period. 29 Jul 2016
4. Sink or Swim?—New Take on Aβ Antibody’s Modus Operandi 16 Sep 2009