Target-engagement markers for LRRK2 drugs

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Dzamko, Deak, Alessi, Nichols and colleagues, University of Dundee (Biochemical Journal, 2010). Fuji, Flagella, Baca, Watts, Estrada and colleagues at Genentech, with the Michael J. Fox Foundation (Science Translational Medicine, 2015). Alcalay (Columbia University), Merchant and colleagues, with Michael J. Fox Foundation staff (Movement Disorders, 2020; the abstract lists an NINDS grant to Alcalay). Jennings and colleagues at Denali, with Biogen and two Dutch trial units (Movement Disorders, 2023); the studies were Denali's. Fan, Alessi, Sammler and colleagues (Biochemical Journal, 2018; Medical Research Council, Academy of Medical Sciences and Parkinson's UK funding listed).

What happened

A drug trial needs a measurement that shows the drug is doing its job, long before any change in symptoms. For LRRK2 inhibitors, three came together.

• pS935. In 2010, Dzamko et al. showed that two different LRRK2 inhibitors made the enzyme lose the phosphate group (phosphorylation, a chemical tag) at two sites, Ser910 and Ser935, in Swiss 3T3 cells and in white-blood-cell-derived (lymphoblastoid) cells from a patient with two copies of the G2019S mutation (inhibitors H-1152 and sunitinib; a drug-resistant LRRK2 mutant did not respond). The authors proposed this as a test of how active an inhibitor is. Blood pS935 later became the standard "target is hit" marker in the LRRK2 trials (for example DNL201, Jennings et al., Sci Transl Med 2022).
• BMP. Fuji et al. (2015) found that a lysosome lipid, di-22:6-BMP (bis(monoacylglycerol)phosphate), fell in the urine of LRRK2 knockout mice and of monkeys given LRRK2 inhibitors (landmark 199). Alcalay et al. (2020) found four BMP forms 2.5 to 4.3 times higher in people with the LRRK2 G2019S mutation than in non-carriers, in two groups (80 and 116 people).
• pRab10. LRRK2 tags Rab proteins, including Rab10 (landmark 37). Fan et al. (2018) showed the Rab10 tag (phospho-Thr73) can be measured in human blood neutrophils, with a small feasibility test in a few G2019S, sporadic and control donors, and suggested it for trials.

In the DNL151 (BIIB122) studies (landmark 229; 186 healthy people and 36 patients), median whole-blood pS935 fell by up to 98%, pRab10 in blood mononuclear cells by up to 93%, total LRRK2 in spinal fluid by up to 50%, and urine BMP by up to 74%. DNL201 also measured pS935 and pRab10 (landmark 200).

Why it moved us toward a cure

These markers let a small Phase 1 show target engagement, and they guided dose choice. They also exposed a limit. In the Phase 2b LUMA (landmark 145; sponsor figures, no paper yet), blood pS935 fell over 90%, yet a spinal-fluid LRRK2 activity marker (pRab10) fell by only about 30% at best, and the disease did not slow. A blood marker of target hit is not proof of effect in the brain.

Outcome

Advanced as a tool: all three were taken into industry trials. They show a drug is active; none has been linked to clinical benefit. No correction or retraction was found for any of the papers; the only notice is an editorial commenting on Jennings 2023.

Sources

• Dzamko et al., Biochem J 2010, DOI 10.1042/BJ20100784
• Fuji et al., Sci Transl Med 2015, DOI 10.1126/scitranslmed.aaa3634
• Alcalay et al., Mov Disord 2020, DOI 10.1002/mds.27818
• Fan et al., Biochem J 2018;475:23-44, DOI 10.1042/BCJ20170803
• Jennings et al., Sci Transl Med 2022 (DNL201), DOI 10.1126/scitranslmed.abj2658
• Jennings et al., Mov Disord 2023, DOI 10.1002/mds.29297
• Not yet verified: the Wang or Ysselstein papers named in the project's candidate list (not read); whether Fan 2018 was the first pRab10 measurement in blood cells; how well BMP tracks dose across different drugs; the LUMA figures (sponsor releases only); PubPeer not checked.