Assembly Modulation: A New Drug Discovery Strategy for CNS Diseases and Beyond

A recap of the CDD Vault webinar featuring Vishwanath R. Lingappa, MD, PhD, CEO and CTO of Prosetta Biosciences and Emeritus Professor of Physiology and Medicine at the University of California, San Francisco

CDD Vault hosted a webinar featuring Vishwanath R. Lingappa, MD, PhD, CEO and CTO of Prosetta Biosciences and Emeritus Professor of Physiology and Medicine at the University of California, San Francisco. Dr. Lingappa presented assembly modulation, a drug discovery strategy that targets how proteins assemble into multi-protein complexes rather than the active sites of individual proteins. The talk traced the approach from antiviral activity through neurodegenerative disease and cancer.

Dr. Lingappa built the strategy on a screening platform that uses cell-free protein synthesis and on a hypothesis that protein assembly is catalyzed, not spontaneous. The sections below summarize the logic, the screen, and the disease data he presented.

The Limits of Active-Site Drugs

Most drugs block disease progression rather than restore health, Dr. Lingappa argued. Restoring health requires returning the cell to homeostasis. Homeostasis is regulated through allosteric sites, the control points on enzymes, not the active sites where catalysis occurs.

Most approved drugs target active sites. Dr. Lingappa listed the liabilities of that approach: side effects, tolerance, limited selectivity for disease, and weak efficacy against complex diseases of homeostasis. The active site is present in every copy of a protein, not the subset involved in a specific function.

Allosteric sites are targeted less often because the right one is hard to find. A large multi-protein complex holds many allosteric sites, each tied to a different function. Most proteins also act inside multi-protein complexes rather than alone. That raised the question at the center of the talk: how are those complexes assembled?

Assembly as a Catalyzed Process

The standard assumption is that complexes assemble spontaneously. Dr. Lingappa challenged this. The cytoplasm is crowded, and spontaneous assembly under those conditions is implausible. His evidence supports a model in which host enzymes catalyze assembly.

He built the case with cell-free protein synthesis, an extract from ground-up cells programmed with a chosen messenger RNA. Programmed with viral capsid messenger RNA, the extract produced capsids through a discrete, stepwise pathway with defined intermediates. Removing components from the extract stopped assembly while protein synthesis continued. Adding the components back restored assembly. Depleting metabolic energy halted the pathway at a fixed point, and restoring nucleotide triphosphates restarted it.

A process that requires host machinery and energy is not spontaneous. It is catalyzed. Dr. Lingappa termed the responsible machinery the enzymes of assembly. Work from a former postdoctoral researcher showed that dominant-negative mutants of this machinery blocked HIV release from infected human T cells, which connected the cell-free observations to live infection.

Viruses as Truffle Hounds

Viruses locate allosteric sites through natural selection over evolutionary time, not through crystal structures or NMR. Dr. Lingappa used that fact as a search tool. He described viruses as "truffle hounds" that point to the allosteric sites on assembly machines, including sites that integrate innate immune defenses. In one example, viruses displace p62, a regulator of autophagy, from the assembly machine. Dr. Lingappa called this sending "the palace guard AWOL," which lets the virus take over host machinery without triggering host defenses.

To screen at scale, his team recreated capsid assembly in 384-well plates. An irrelevant green fluorescent protein messenger RNA served as a control to rule out compounds that blocked protein synthesis. Assembled capsids were captured on antibody-coated plates and then decorated with a biotinylated version of the same antibody, which produced a high signal-to-noise fluorescent readout from the many unoccupied sites on each capsid.

Compounds that blocked early in the pathway drove fluorescence to zero. Compounds that blocked at a later intermediate drove it to a plateau. Across a decade, the team screened libraries of more than 100,000 compounds against capsid proteins from every viral family that causes serious human disease. The result is a collection of roughly 300 structurally unrelated chemotypes, which Dr. Lingappa calls the hitfinder collection. Animal data supported the antiviral activity, including oral treatment of a cat with feline infectious peritonitis and reduced viral load in pigs infected with an arterivirus.

Three Properties That Hid the Targets

Dr. Lingappa argued that assembly machines escaped detection because of three properties:

• They are transient. The complexes form and dissolve rather than persisting, which makes them hard to capture.
• They are energy-dependent, both for the reaction they catalyze and for their own formation. Supplying metabolic energy was the step that let his team detect them.
• They draw on a minuscule subset of each component protein. Less than 5 percent of total cellular 14-3-3 zeta sits in the target complex, which leaves the other functions of that protein untouched.

From Antivirals to Neurodegeneration

Dr. Lingappa proposed that the same host machinery viruses exploit also fails in non-viral disease. Viruses do not cause neurodegeneration, but they associate with it. Endogenous retroviruses activate in ALS. Influenza links to alpha-synuclein aggregation in Parkinson's disease. Herpesviruses appear in Alzheimer's disease.

He reframed the standard model of neurodegeneration. The conventional view holds that misfolded proteins escape degradation, aggregate, and cause disease. Dr. Lingappa argued instead that a disease trigger produces an aberrant assembly machine, which builds a pathological signaling complex. The aggregates that most research targets are, in his words, "the wreckage at the roadside when the traffic light is broken." The treatment goal is to fix the assembly machine, not clear the aggregates.

In ALS patient cells, the nuclear protein TDP-43 leaks into the cytoplasm. Assembly modulators active against HIV normalized that mislocalization within 72 hours and blocked stress-induced TDP-43 aggregation. The compounds worked in familial and sporadic cellular models and in every transgenic animal model tested. Protein disulfide isomerase is the direct drug-binding protein for the advanced ALS compound, with 14-3-3 family members present in the same complex.

A Blood-Based Biomarker for ALS

Dr. Lingappa described a diagnostic application built on the same chemistry. Using energy-dependent drug resin affinity chromatography (EDRA), his team examined peripheral blood mononuclear cells. Healthy individuals showed abundant p62 and no fragments of Ran GTPase. Every ALS patient examined showed reduced p62 and a Ran GTPase fragment, and both signals tracked with disease progression.

The signature appears in blood before severe disability, which points to early detection, trial enrollment selection, and treatment monitoring. Dr. Lingappa offered a hypothesis for why a motor neuron defect shows up in blood: blood circulates through every tissue and reports on what it finds. He noted the hypothesis remains untested.

Parkinson's and Alzheimer's

In Parkinson's models, the pesticide rotenone produces alpha-synuclein aggregates and destroys dopaminergic neurites in culture. Assembly modulators active against influenza eliminated the aggregates or redirected alpha-synuclein into ring structures, and they protected the neurite network. The direct drug-binding protein is a subset of mTOR, again a small fraction of the total. The compound reaches rodent brain, and animal studies await funding.

In Alzheimer's work with collaborators, an assembly modulator active against herpesviruses bound an allosteric site and prevented tau hyperphosphorylation in an Alzheimer's mouse model. It also detected its target in autopsy tissue from sporadic Alzheimer's disease.

Selectivity and the Allosteric Punch Code

An audience question raised selectivity directly: would long-term treatment that targets a physiological function prove toxic? Dr. Lingappa pointed to the respiratory program. The early compound bound both the normal assembly machine and the aberrant one and carried toxicity. Structure-activity work produced compounds selective for the disease state.

Both the early and advanced compounds bind the same allosteric site on 14-3-3 zeta, not the active site, but at slightly different positions. Dr. Lingappa compared the site to a "punch code" in which different entry sequences open different gates. The advanced compound detects only the virally repurposed machine, which gives selectivity for disease and low toxicity. He also argued that correcting an assembly machine could require short-term treatment rather than chronic dosing, though he noted this remains a hypothesis.