Markers associated with genetic resililience to AD in E4/E4's

Insights and discussion from the cutting edge with reference to journal articles and other research papers.
Post Reply
broiler_x
Senior Contributor
Senior Contributor
Posts: 60
Joined: Wed Aug 08, 2018 8:52 am

Markers associated with genetic resililience to AD in E4/E4's

Post by broiler_x »

https://www.ncbi.nlm.nih.gov/pubmed/31506248

These guys did a study to look at just E4/E4's over the age of 75 who have not developed AD and looked at their genome SNP's to see if any genetic markers were associated with the resilience to developing AD. They found 2 SNP's that seemed to have "strong evidence" for resilience (CASP7 rs10553596 and SERPINA3 rs4934-A/A). I just checked my 23andme genome data to see if I have either of those SNP's and saw that I had rs4934-A/A. It's possible that 23andme doesn't even check for the other SNP (not sure how to check for that).

Keep in mind that this is correlation and not causation and all the caveats associated with that. And this is a single study, and genes are not destiny etc., etc.
User avatar
cdamaden
Senior Contributor
Senior Contributor
Posts: 300
Joined: Thu Mar 17, 2016 1:46 pm
Location: Alameda, CA, USA

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by cdamaden »

broiler_x wrote: They found 2 SNP's that seemed to have "strong evidence" for resilience (CASP7 rs10553596 and SERPINA3 rs4934-A/A).
Thanks for posting broiler, I checked on my Promethease results (from my 23&me data), and it didn't have the CASP7 one and for the SERPINA3 it states: "rs4934 increases susceptibility to Alzheimer's disease 2.56 times for (A;A)..." I have A;G. So that begs the question, which variant is protective? Is this one of those wacky ones where A really means G? I can't ever follow those cases.
Chris
Chris
E4/E4
Alameda, CA, USA
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by NF52 »

broiler_x wrote:https://www.ncbi.nlm.nih.gov/pubmed/31506248

These guys did a study to look at just E4/E4's over the age of 75 who have not developed AD and looked at their genome SNP's to see if any genetic markers were associated with the resilience to developing AD. They found 2 SNP's that seemed to have "strong evidence" for resilience (CASP7 rs10553596 and SERPINA3 rs4934-A/A). I just checked my 23andme genome data to see if I have either of those SNP's and saw that I had rs4934-A/A. It's possible that 23andme doesn't even check for the other SNP (not sure how to check for that).

Keep in mind that this is correlation and not causation and all the caveats associated with that. And this is a single study, and genes are not destiny etc., etc.
This is an exciting article, meaning I was willing to pay to read it all! I found a few other SNPs and information: on "potential AD resilience" for ApoE 4/4s, some of which I couldn't find on my 23&me report:

The G/G alleles of rs9472817 "nullified" the risk of ApoE 4 allele (with P value of .06, just above the preferred 5% chance value).

Those with r4934 A/A alleles showed the same risk of AD as those with one ApoE 4, "eliminating the gene dosage effect".

Those with IL-10-1087 (rs1800896) A/A (T/T) had an almost 40% lower risk of AD even in those with ApoE 4.

The A/A allele (or T/T) of rs20119 A/A increases risk of AD, while the C/C (G/G) allele may reduce risk.

COMT rs4680 G/G allele may decrease risk of AD in ApoE 4/4s, but there was a wide confidence interval.

The article included explanations of why many studies listed in their meta-analysis had results that were inconclusive, missing data or unable to be interpreted.

At the end of the analysis, the authors, most of whom are in Australia, with one based at Mt. Sinai in NYC, call for intensive study of ApoE 4/4 in particular, without it being grouped with ApoE 3/4 or 2/4. They note that "resilient family members...homozygous for the APOE e4 allele, aged over 75, excluded for LOAD or mild cognitive impairment using thorough cognitive analysis" could identify "novel" protective variants in LOAD, as well as molecular pathways that support resilience, which could aid in prevention and therapy trials.
They also suggest a technique called "extreme phenotyping" to improve the accuracy of polygenic risk scores, by taking resilient ApoE 4/4s over the age of 75 and comparing their genome to similarly carefully selected and confirmed AD cases of ApoE 4/4s diagnosed under the age of 70, to increase the likely effect size.

The National Institute on Aging already has a multi-year multi-site Alzheimer's genome sequencing project Alzheimer's Disease Sequencing Project (ADSP), so maybe some of this is already happening.
4/4 and still an optimist!
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by NF52 »

cdamaden wrote:
broiler_x wrote: They found 2 SNP's that seemed to have "strong evidence" for resilience (CASP7 rs10553596 and SERPINA3 rs4934-A/A).
Thanks for posting broiler, I checked on my Promethease results (from my 23&me data), and it didn't have the CASP7 one and for the SERPINA3 it states: "rs4934 increases susceptibility to Alzheimer's disease 2.56 times for (A;A)..." I have A;G. So that begs the question, which variant is protective? Is this one of those wacky ones where A really means G? I can't ever follow those cases.
Chris
Hi Chris,

Check out the explanation i found in my post above on SERPINA. It looks like an increase of 2.56x AD risk is the GOOD news for those of us with ApoE 4/4, since the study found that there was not the usual very magnified "dosage" effect of more risk with ApoE 4/4 over ApoE 3/4. So it's resilience based on the expected effect in 4/4s.
4/4 and still an optimist!
TelopeaBlue
Contributor
Contributor
Posts: 12
Joined: Wed Aug 01, 2018 8:36 am

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by TelopeaBlue »

cdamaden wrote:
broiler_x wrote: They found 2 SNP's that seemed to have "strong evidence" for resilience (CASP7 rs10553596 and SERPINA3 rs4934-A/A).
Thanks for posting broiler, I checked on my Promethease results (from my 23&me data), and it didn't have the CASP7 one and for the SERPINA3 it states: "rs4934 increases susceptibility to Alzheimer's disease 2.56 times for (A;A)..." I have A;G. So that begs the question, which variant is protective? Is this one of those wacky ones where A really means G? I can't ever follow those cases.
Chris
Hi cdamaden, for SERPINA3 the report is saying the A allele is the increased risk allele. It says you have one copy of that allele, and one copy of the alternative version, the G allele. That is, you are heterozygous at that locus. I have no insight into relative risks with one copy of each.

And for your little mix up issue - for what it's worth A can never really mean G. An easy way to remember is the pointy letters match (A and T) and the round letters, C and G, match. It's all about the biochemistry, which I won't go into here. Just easier to think about pointy and round ;-)
broiler_x
Senior Contributor
Senior Contributor
Posts: 60
Joined: Wed Aug 08, 2018 8:52 am

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by broiler_x »

Thanks guys for the great responses. So it looks like 23andme may not report the CASP7 status for this particular SNP (or none of us have seen it so far), but they do for the other one. It also looks like rs4934-A/A may possibly override the apoE4/E4 phenotype and bring it down to possibly E3/E4 (so many caveats, so many questions), but for non-E4's it could be slightly worse news to be rs4934-A/A. Yes, this can be very confusing and complex. As is all biology!

What I really liked about this study is that they focused on 4/4's. Having worked in AD research with 2 different biotech companies I can tell you that we generally assumed that 4/4's were a lost cause (sorry, but we excluded them from any consideration - this was a practical decision to increase the chances for success). I totally went along with this until I found out that I'm a 4/4 myself!

I do want to point out that genes are not our destiny! However, the reason we are all here is because of our genes (you wouldn't be here unless you found out you had at least one ApoE4 allele). We know this a multi-factorial disease, and I'm sure over time there will be many other genetic and non-genetic causes, and many ways to avoid a negative outcome.
antimatter37
Contributor
Contributor
Posts: 56
Joined: Fri Feb 02, 2018 11:52 am

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by antimatter37 »

I am also a bit confused here with regard to SERPINA3. Looking at the full text of the two references used in the study under discussion, both references use "ACT" instead of SERPINA3. Not sure why, but I have copied the exact results from both papers below.

From the paper by DeKosky et al;

"Of the three ACT genotypes, the ACT A/T genotype did not affect the risk of AD beyond that which is conferred by the APOE €4 allele. However, the ACT A/A genotype greatly increased the risk associated with APOE €4 homozygosity. On the other hand, the ACT T/T genotype suppressed the gene dosage effect of the APOE €4 allele; the risk of AD in APOE ~4k4, ACT T/T cases was the same as that of APOE E4/x"

From the paper by Kamboh et al (includes DeKosky as author as well, so basically same primary authors on both papers);

The ACT/TT genotype had the same risk with either one or twoAPOE*4 allele copies; thusACT/TT could be construed as protective from the added gene dosage effect of APOE*4. However, in the presence of ACT/ AAgenotype, an individual's risk increased to almost
two fold with one APOE*4 copy (3.5 versus 6.4) and to threefold with two APOE*4 copies (11.1 versus 34.0) above that which is associated with the APOE*4 allele alone.
DoubleBond
Contributor
Contributor
Posts: 24
Joined: Wed Feb 06, 2019 4:21 pm

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by DoubleBond »

antimatter37 wrote:I am also a bit confused here with regard to SERPINA3. Looking at the full text of the two references used in the study under discussion, both references use "ACT" instead of SERPINA3.
I think ACT stands for the alpha-1-antichymotrypsin protein, SERPINA3 is the gene coding for the ACT protein. As far as I understand (experts please correct), the original paper is in terms of the amino acid polymorphism in ACT (alanine A vs threonine T, A being the ancestral case, T mutated), while the underlying codon in the SERPINA3 mutation rs4934 is G or A (G is ancestral, A mutated). The double mutation (i.e. AA in rs4934 or TT in terms of the ACT protein) was considered protective for E4 homozygotes in the original DeKosky and Kamboh 1995 and 1996 papers:
https://www.ncbi.nlm.nih.gov/pubmed/7670501
https://www.ncbi.nlm.nih.gov/pubmed/8993481

But I would not ascribe too much importance to these findings. First of all, confidence intervals are really wide; eg the odds ratio for AD for the supposedly protective TT in the ACT were in the confidence interval 1-20.
Even more importantly, these findings were later contradicted; at least one paper found no interaction of APOE and ACT for the risk of Alzheimer's.
https://www.ncbi.nlm.nih.gov/pubmed/8617509
Also, an Italian study found TT in the ACT protein was increasing odds for early-onset AD
https://www.ncbi.nlm.nih.gov/pubmed/15653173
I also find it curious that this supposed effect was not found in any more recent studies.
antimatter37
Contributor
Contributor
Posts: 56
Joined: Fri Feb 02, 2018 11:52 am

Re: Markers associated with genetic resililience to AD in E4/E4's

Post by antimatter37 »

Thanks for the reply, this does clear up the SERPINA results questions I had. I wonder how the CASP7 rs10553596 result will stand the test of time.
Post Reply