Some studies have suggested that Alzheimer’s disease biomarker levels may differ by race. Now, researchers led by John Morris and Suzanne Schindler, Washington University, St. Louis, offer a genetic explanation—at least for soluble TREM2, the extracellular domain of the microglial receptor that ends up in the cerebrospinal fluid. In the March 4 Neurology Genetics, they reported that African Americans are more likely to carry TREM2 variants that dampen the protein’s expression than are non-Hispanic Caucasians. They are also less likely to have a variant in the nearby MS4A4A locus that has been linked to higher CSF sTREM2. These inheritance patterns translate to African Americans having less sTREM2 in their CSF. This could be important when deciding sTREM2 cutoffs for use as a diagnostic biomarker or an enrollment criterion in clinical trials.
- African Americans more likely carry genetic variants that lower TREM2 expression and processing.
- They produce less soluble TREM2 in their cerebrospinal fluid than do Caucasians.
- This difference should be considered if sTREM2 is to be a biomarker.
“This article is very encouraging because it demonstrates that carefully designed studies can provide valuable insight into [the genetics of] disease pathophysiology despite having relatively smaller sample sizes,” Minerva Carrasquillo, Mayo Clinic, Jacksonville, wrote to Alzforum (full comment below). “This will be the first of many studies focusing on genetics and biomarkers in more diverse participants,” Carlos Cruchaga, a co-author also at WashU, told Alzforum.
African Americans are twice as likely as Caucasians to get dementia (Dec 2020 news). Yet despite being 13 percent of the U.S. population, they account for 5 percent or less of people enrolled in AD clinical trials and study cohorts. To address this underrepresentation, Morris, Schindler, and colleagues have recruited people from diverse backgrounds to the Knight Alzheimer’s Disease Research Center cohort, which now comprises 19 percent African Americans, and more recently have led efforts to boost minority participation elsewhere (Oct 2018 news). “The racial diversity of the Knight ADRC cohort can serve as an excellent model for other observational cohorts,” Alberto Lleó, Hospital of Sant Pau, Barcelona, Spain, and Marc Suárez-Calvet, BarcelonaBeta Brain Research Center, Spain, wrote to Alzforum (full comment below).
The researchers are starting to accrue enough data to be able to tease out biomarker differences by race. So far, Schindler and colleagues have found less total tau and phosphotau-181 in the CSF of cognitively normal or cognitively impaired African American study participants than in their Caucasian counterparts (Jan 2019 news). Ihab Hajjar, Emory University, Atlanta, found the same in a cohort of people with mild cognitive impairment (Garrett et al., 2019).
What about microglial markers? Previously, researchers reported that sTREM2 creeps up in the CSF of people with AD and that higher levels correlate with slower cognitive decline (Aug 2019 news; Jan 2016 news). Schindler and colleagues wondered if CSF sTREM2 levels differed by race.
To find out, they collected CSF biomarker and genetic information from 91 older African Americans and 868 non-Hispanic Caucasians in the Knight cohort. The majority of participants were cognitively normal—13 percent of African Americans and 25 percent of Caucasians had MCI. There were some notable baseline differences. Compared to the Caucasian participants, African Americans were three years younger on average, had about half a year less education, were less likely to report a family history of dementia, and were less likely to have MCI.
Schindler and colleagues compared participants by self-reported race and by race as determined by the presence of genetic variants traced to African countries or regions. Self-identified race matched genetic ancestry for all but one participant, who was African American; therefore the scientists saw the same trends when grouping participants by either method.
The researchers compared CSF levels of sTREM2, Aβ42, total tau (t-tau), phosphorylated tau 181 (p-tau181), and neurofilament light (NfL) among participants. While amyloid did not differ, sTREM2, t-tau, p-tau181, and NfL were lower in the African Americans. These differences held true after correcting for a plethora of other variables, including age, years of education, family history of dementia, ApoE4 status, and dementia status determined by the clinical dementia rating score.
Were these differences truly linked to race, or could the baseline differences still be the cause, even after correction? To double-check, Schindler and colleagues also used a computer algorithm to match 86 of the 91 African Americans with a Caucasian counterpart who was closest to them in age, years of education, CDR score, and family history of dementia. African Americans still had less CSF sTREM2, t-tau, and p-tau181.
To see if these findings would hold in other cohorts, the researchers collected data on 23 older African Americans and 917 non-Hispanic Caucasians volunteers from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). In this group, African Americans had about a year less education than Caucasians, but were the same on all other demographics. CSF Aβ42, t-tau, p-tau181, and NfL did not differ by race. The researchers attributed this to the cohort size, which may be too small to tease out such subtle differences. Even so, in this small ADNI sample, sTREM2 was lower in African Americans as determined by both self-reported race and genetic ancestry. This difference held true after adjusting for confounders.
Could genetics explain the TREM2 dichotomy? The researchers searched for differences in TREM2 coding variants among the volunteers. In both cohorts, African Americans were five times more likely to have any of these variants than were Caucasians. Schindler also looked at the effect of two single-nucleotide polymorphisms—rs1582763 and rs6591561—around the nearby MS4A4A locus. These SNPs previously had been linked to higher and lower CSF sTREM2, respectively, likely through TREM2 processing differences (Jul 2018 news). Lo and behold, African Americans were 4.6 and 2.3 times less likely to have the sTREM2-boosting rs1582763 in the ADRC and the ADNI cohorts, respectively. The sTREM2-lowering rs6591561 did not differ between groups in either cohort.
TREM2 Trends. On average, African Americans (red dots) had less sTREM2 (red line) in their CSF than non-Hispanic Caucasians (black dots and line). Left two panels show carriers and noncarriers, respectively, of a TREM2-lowering coding variant. Right two panels show the same, but this time among people who also carried the sTREM2-boosting rs1582763 SNP at the MS4A4A locus. African Americans were more likely to carry the TREM2-suppressing allele and less likely to carry the sTREM2-boosting SNP. [Courtesy of Schindler et al., Neurology Genetics, 2021.]
Did these variants directly relate to CSF sTREM2? Yes. People with a TREM2 coding variant had less sTREM2 in their CSF, while people with the rs1582763 SNP had more (see image above). These relationships remained after correcting for the previously mentioned confounds.
Schindler said the lower CSF sTREM2 levels in African Americans should be considered during diagnosis and treatment. Researchers believe that production of soluble TREM2 reflects microglial activity, and that less sTREM2 means fewer amyloid-clearing microglia. This could make African Americans more susceptible to amyloid pathology. In contrast, less CSF t-tau, p-tau181, and NfL, all markers of neuronal damage, could mean African Americans have less brain damage despite having similar amyloid levels as do Caucasians. This needs to be investigated, the scientists agree.
Researchers believe racial differences in TREM2 biology should also be considered when deciding diagnostic biomarker cut-offs, which are largely based on people of European ancestry. “Genetic variants tied to ancestry influence baseline biomarkers and, ultimately, whether someone is above the diagnostic threshold or not,” Cruchaga said. Lleó agreed. “The heterogeneity of TREM2 function and microglial activity, irrespective of self-identified race, may also contribute to clinical heterogeneity in AD, similar to what we see with tau,” he wrote to Alzforum.
Understanding racial diversity in the underlying biology of Alzheimer's is key to developing safe and effective drugs for everyone, researchers agreed. “If we do not know about biological differences as we design drugs, we may end up with drugs that do not work in African Americans,” Schindler said. This was the case with high blood-pressure medications, which work to varying degrees and with different side effects in African Americans than in Caucasians (Clemmer et al., 2020; Chekka et al., 2021; Wu et al., 2005).
One immediate example is Alector’s anti-TREM2 antibody, AL002. It is currently in Phase 2 for AD. In Phase 1, the antibody lowered sTREM2 in the CSF of people with mild to moderate AD (June 2020 news). “It will be important to investigate whether TREM2 therapies have a different effect in participants with lower TREM2 function,” Lleó and Suárez-Calvet wrote in an accompanying editorial.
Arnon Rosenthal at Alector LLC, South San Francisco, plans to do just that. For the Phase 2 study, the company will retrospectively screen for TREM2 mutations and look at drug efficacy through the lens of mutation carriers and baseline sTREM2 concentration. He said Alector also aims to broaden participant diversity. “We are opening additional clinical sites in places like Atlanta and Baltimore to make accessing the trial easier,” Rosenthal told Alzforum.
Mark Gluck and his group are building a research cohort of older African Americans at Rutgers University in Newark, New Jersey. They have not yet looked at TREM2 in the more than 400 people they have genotyped so far. “Based on this paper, we will add TREM2 to all our future analyses, especially as we begin new studies of immune health and COVID-19 history,” Gluck wrote to Alzforum (comment below).—Chelsea Weidman Burke
- In the United States, Racial Disparities in Dementia Risk Persist
- Alzheimer’s Researchers Seek Advice on How to Include African-Americans
- Do Alzheimer’s Biomarkers Vary by Race?
- In Alzheimer’s, More TREM2 Is Good for You
- TREM2 Goes Up in Spinal Fluid in Early Alzheimer’s
- MS4A Alzheimer’s Risk Gene Linked to TREM2 Signaling
- In Mice, Activating TREM2 Tempers Plaque Toxicity, not Load
- Garrett SL, McDaniel D, Obideen M, Trammell AR, Shaw LM, Goldstein FC, Hajjar I. Racial Disparity in Cerebrospinal Fluid Amyloid and Tau Biomarkers and Associated Cutoffs for Mild Cognitive Impairment. JAMA Netw Open. 2019 Dec 2;2(12):e1917363. PubMed.
- Clemmer JS, Pruett WA, Lirette ST. Racial and Sex Differences in the Response to First-Line Antihypertensive Therapy. Front Cardiovasc Med. 2020;7:608037. Epub 2020 Dec 17 PubMed.
- Chekka LM, Chapman AB, Gums JG, Cooper-DeHoff RM, Johnson JA. Race-Specific Comparisons of Antihypertensive and Metabolic Effects of Hydrochlorothiazide and Chlorthalidone. Am J Med. 2021 Jan 9; PubMed.
- Wu J, Kraja AT, Oberman A, Lewis CE, Ellison RC, Arnett DK, Heiss G, Lalouel JM, Turner ST, Hunt SC, Province MA, Rao DC. A summary of the effects of antihypertensive medications on measured blood pressure. Am J Hypertens. 2005 Jul;18(7):935-42. PubMed.
- Jin SC, Carrasquillo MM, Benitez BA, Skorupa T, Carrell D, Patel D, Lincoln S, Krishnan S, Kachadoorian M, Reitz C, Mayeux R, Wingo TS, Lah JJ, Levey AI, Murrell J, Hendrie H, Foroud T, Graff-Radford NR, Goate AM, Cruchaga C, Ertekin-Taner N. TREM2 is associated with increased risk for Alzheimer's disease in African Americans. Mol Neurodegener. 2015 Apr 10;10:19. PubMed.
- Schindler SE, Cruchaga C, Joseph A, McCue L, Farias FH, Wilkins CH, Deming Y, Henson RL, Mikesell RJ, Piccio L, Llibre-Guerra JJ, Moulder KL, Fagan AM, Ances BM, Benzinger TL, Xiong C, Holtzman DM, Morris JC. African Americans Have Differences in CSF Soluble TREM2 and Associated Genetic Variants. Neurology Genetics, March 4, 2021
- Lleó A, Suárez-Calvet M. Race and Alzheimer Disease Biomarkers: A Neglected Race. Neurol Genet. 2021 Apr;7(2):e574. Epub 2021 Mar 4 PubMed.