LRRK2 G2385R and R1628P

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

Alessio Di Fonzo, Vincenzo Bonifati and colleagues (G2385R, Taiwan, Neurogenetics 2006); Matthew Farrer, Ruey-Meei Wu and colleagues (G2385R, Taiwan, 2007); Manabu Funayama, Nobutaka Hattori and colleagues (G2385R, Japan, 2007); Owen Ross, Ruey-Meei Wu, Eng-King Tan and colleagues (R1628P, Taiwan and Singapore, 2008). The Farrer and Ross papers list US NIH extramural support; other funders were not checked.

What happened

The LRRK2 gene was found in 2004 (landmark 7). The best-known mutation, G2019S, is rare in East Asia. Two other changes were found to be common there:

• G2385R. Di Fonzo 2006 first reported it as a risk variant: in 608 Taiwanese patients and 373 controls the odds ratio was 2.24. Farrer 2007 confirmed it in ethnic Chinese Taiwanese and traced carriers to a single ancestor about 4,800 years ago. Funayama 2007 tested 448 Japanese patients and 457 controls and found an odds ratio of 2.63 (95% CI 1.56–4.35). A 2014 meta-analysis of 23 case-control studies (12,915 cases, 12,451 controls) gave an odds ratio of 2.4 (95% CI 1.97–2.92) (Xie 2014).
• R1628P. Ross 2008 studied 1,986 people across three centres in Taiwan and Singapore: odds ratio 1.84 (95% CI 1.20–2.83, p=0.006), with a founder about 2,500 years ago. A letter by Tan and colleagues (2008) looked at non-Chinese Asian groups; its abstract is not available. A 2016 meta-analysis of 14 studies (5,736 patients, 4,786 controls) gave an odds ratio of about 1.97 (95% CI 1.63–2.39) for carriers, and similar in Chinese and non-Chinese Asians (Wang 2016).

Unlike G2019S, these are common, low-penetrance variants, so most carriers never get Parkinson's.

Why it moved us toward a cure

It showed that LRRK2 contributes to ordinary-looking Parkinson's in a large population, not only to rare families (carrier odds ratios of about 1.8 to 2.6, so most carriers stay healthy). That supports the idea that LRRK2 biology might matter in people without family history, which later underpinned the broad-population trial of a LRRK2 inhibitor (landmark 145, via landmark 198). It also showed that a drug target can be population-specific.

Outcome

Advanced. Meta-analyses of many studies confirm both variants as risk factors in East Asians. A 2026 study of 242 people in Malaysia found raised LRRK2 kinase activity (pRab10, about 1.2-fold) in G2385R carriers but not in R1628P carriers, and in some people without either variant (Toh 2026, Mov Disord, funded by the Michael J. Fox Foundation). That suggests carrier status alone does not say who has overactive LRRK2. No correction or retraction was found for the 2006–08 papers (Crossref, Europe PMC).

Sources

• Di Fonzo et al., Neurogenetics 2006;7:133–138, DOI 10.1007/s10048-006-0041-5 (PubMed 16633828)
• Farrer et al., Parkinsonism Relat Disord 2007;13:89–92, DOI 10.1016/j.parkreldis.2006.12.001 · Funayama et al., NeuroReport 2007;18:273–275, DOI 10.1097/WNR.0b013e32801254b6
• Ross et al., Ann Neurol 2008;64:88–92, DOI 10.1002/ana.21405 · Tan et al., Ann Neurol 2008, DOI 10.1002/ana.21467
• Xie et al., Neurol Sci 2014;35:1495–1504, DOI 10.1007/s10072-014-1878-2 (G2385R meta-analysis) · Wang et al., Neurosci Lett 2016;623:22–27, DOI 10.1016/j.neulet.2016.04.056 (R1628P meta-analysis)
• Toh et al., Mov Disord 2026;41:1750–1761, DOI 10.1002/mds.70314 (abstract read)
• Not yet verified: the Tan 2008 letter's findings (no abstract); carrier frequencies in controls; Farrer 2007 odds ratio (not in abstract).