Natural Killer Cells and the Search for Biomarkers for Autism
January 26, 2008 by Kristina Chew, PhD
Filed under Health
Researchers at the M.I.N.D. Institute at the University of California at Davis are the first to use genomic profiling of blood to note differences in autistic children, the January 25th Health News Digest reports. Their hope is that such “gene expression analyses can provide biological evidence of autism, currently diagnosed only through behavioral assessments, in some children.”
“What we found were 11 specific genes with expression levels that were significantly higher in the blood of children with autism when compared to the blood of typically developing children,” said Frank Sharp, senior author of the study and professor of neurology with the M.I.N.D. Institute. “Those 11 genes are all known to be expressed by natural-killer cells, which are cells in the immune system necessary for mounting a defense against infected cells. We were surprised by our results because we were not looking for these particular genes. And while a number of studies have shown immune system dysregulation to be an important factor in autism, ours is one of the first to implicate these particular cells.”
In conducting the study, Sharp, molecular pathologist Jeff Gregg and their M.I.N.D. Institute colleagues used blood samples from 35 children diagnosed with autism, 14 with development delay but not autism and 12 typically developing children. The samples were subjected to gene expression analysis using microarrays and compared for common patterns. In addition to finding the 11 genes with natural-killer cell connections shared by all of the children with autism, they identified a pattern of 140 genes differentially expressed in children with the early onset form of the disorder and a pattern of 20 genes differentially expressed in children with the regressive form of the disorder. The team is the first to use genomic profiling of blood to observe differences in children with autism.
Might this study point towards the development of a blood test for autism, of a biomarker for diagnosing autism? The study is published in the January issue of Genomics.
Dr. Sharp further notes that the current study also does not “identify whether or not the natural-killer cells are functioning abnormally,” which further work by M.I.N.D. Institute immunologists will consider reveal. He continues:
“If the natural-killer cells are dysfunctional, this might mean that they cannot rid a pregnant mother, fetus or newborn of an infection, which could contribute to autism.”
The image of some “natural killer” cells that ward off tumors and affect cells, and so play a role in the genes of autistic children stands out. One can imagine proponents of theories that something in the environment “triggers” autism making something of this research about something like stealth autism getting into the genes of some children and causing an “infection” in a pregnant mother or fetus.
A life less ordinary by Emily goes into much more detail about this study and about autoimmunity and autism (and I’ll note here that, between Jim’s family and mine, there is a history of autoimmune disorders: multiple sclerosis and allergies on his side, and asthma, which I used to have a severe case of).
It’s a genomics study of groups with small numbers of children, meaning that it’s hitting the surface and that there’s much more investigation to be done. The NK findings are very interesting but shared across both autism “phenotypes.”
There are some other notable findings. (1) More genes altered regulation in the early-form group than in the “regressive form” group compared to the general population. (2) The fold-change was not very great for ANY of the genes. (3) In the early-form vs. regressive-form comparison, while they identified ~500 genes that exhibited putative upregulation (only three were 2-fold or greater, if I remember correctly), there don’t seem to be any that exhibited significant downregulation. That seems odd to me.
There is nothing in this study to imply either way an involvement of any environmental influence. Identification of mechanism or factors involved is a long way down this road. We’ve just made a right turn onto the highway.
To clarify and so as not to unfairly discuss the results, I should say that the way they did their analysis makes the 1.5 cutoff completely defensible. And that ~500 includes a much shorter list of downregulated genes. I just don’t see big old fireworks there yet.
It’s difficult to interpret in any way…and probably should not be attempted…the biological relevance of a particular fold-change for a particular gene. That’s why further characterization–such as the analyses of the NK genes they did–of specific genes is always necessary. They’ll go from this work to targeting the specific genes using molecular biology tools and then assessing their responses using in vitro and in vivo models.
This (and your post) are really helpful—-do you think it’s likely that some sort of test might result from this kind of research?
If they characterize this apparent NK expression profile, consistently associate it with the disorder across populations, and elucidate a mechanism, then…maybe it could be a biomarker. With this upregulation situation, it would seem to me that the place to look for a mechanism would be, surprise, the regulatory pathways governing NKs. I think there are a lot of those, so it may be awhile. Also, there needs to be an investigation of cause vs. effect, etc. (i.e., does this follow on autism, or is it causative? It’s present in both “phenotypes” they examined, so doesn’t seem “phenotype” specific).
Another thing I was wondering about was if they had looked at any sex-based association. They age and gender-matched the groups, so I assume both sexes are represented. Given the putative differences in “phenotype” between the sexes, I’d wonder if that would show up in gene expression profiles. I don’t see a mention of such an analysis in the paper; probably, the groups for comparison would be very small. Nonetheless, given their argument for phenotypes, it would be an interesting avenue to pursue.
Emily
Your conclusions about possible environmental implications arising from this study seem at odds with those actually involved with the study. This is what the senior researcher Frank Sharp said about possible environmental factors:
“What we are seeing can reflect something in the environment that is triggering the activation of these genes or something genetic that the children have from the time they were conceived,” Sharp explained. “Such an immune response could be caused by exposure to a virus, another infectious agent or even a toxin.
Another possibility is that these changes represent a genetic susceptibility factor that predisposes children to autism when they are exposed to some environmental factor.”
Dr. David Amaral the UC Davis MIND Institute’s research director stated:
“Things are moving really, really fast now,” Amaral said, with scientists around the country working to understand the relationship of genetic and environmental factors that may underlie autism.”
Hi, Harold:
Yes, I read the paper and the article.
“What we are seeing can reflect something in the environment that is triggering the activation of these genes or something genetic that the children have from the time they were conceived,” Sharp explained. “Such an immune response could be caused by exposure to a virus, another infectious agent or even a toxin.”
He’s doing the “speculation” part of science there. He said “can” not “does,” and that’s significant. It means that what he says is their developing hypothesis about what they’ve seen here. That doesn’t mean that the story is now told.
“Another possibility is that these changes represent a genetic susceptibility factor that predisposes children to autism when they are exposed to some environmental factor.”
Again, this is the “speculation” part of science, the stuff you put in the discussion that involves YOUR ideas about explanations for what you are seeing. that’s why he used the word “possibility.” That’s why, in the quotes you provide, he offers two *different* possible overall mechanisms. I’ve written or edited literally hundreds of scientific papers, and I can tell you from experience that’s exactly what’s behind these statements. It’s the fun part of writing a paper, the part where you put out your kind-of-supported ideas about what the explanations for your findings might be. You can’t go to far afield with it or the reviewers will come after you, but as long as you keep it to a dull roar and make sense, it’s cool.
Again, this is a GENOMICS study. It does not generally tell us a single thing about mechanism or what influences the genes. That’s why characterization studies of the genes and what causes them to upregulate are still required. These genes are upregulated in children with EXISTING autism. Thus, we do not know whether we have a causative outcome here or one involving effect. Genomics results generally do not tell us that. Just because someone *wants* a result to have a certain explanation does not mean that explanation is the only one or even the likeliest one.
You’ll note that even in your first quote, he offers two possibilities: Something environmental triggering activation of the genes, OR something the children have from the moment the egg met the sperm, i.e., straight-up genetic.
“Things are moving really, really fast now,” Amaral said, with scientists around the country working to understand the relationship of genetic and environmental factors that may underlie autism.”
I agree…relative to where we were even a decade ago, things ARE moving fast. It’s that way across the biomedical sciences. That doesn’t mean we’re going to have a biomarker test for NK genes tomorrow. The researchers will take these genes through a pretty standard process of characterization and response in the context of their hypotheses. That does take time. It won’t happen overnight, and that’s what I mean when I say, “it may be awhile.” And I actually know from experience what I’m talking about.
I’ve stated here and elsewhere repeatedly that my personal opinion as a scientist and a parent is that we will find that autism is an interaction of genes, epigenetic regulation, and environmental factors. The authors of this study appear to speculate exactly that, and autism-specific upregulation of a specific suite of genes fits with it, too. But that doesn’t mean I’m going to start yelling from the rooftops that I’m right. The data are not all in by any stretch.
Emily’s statements aren’t at odds at all with those, from what I saw. She noted that it didn’t implicate the trigger either way, which isn’t at odds with the possibility of the environmental triggers (not fact, look at “can” and “could” in Sharp’s statement), and the last one is actually kind of irrelevant (”with scientists…working to understand the relationship of genetic and environmental factors that may underlie autism” isn’t much of a statement of that relationship, but of the work).
Cliff
Emily and Cliff
I thank both of you for clarifying that environmental factors can not be ruled out. I think though that you understated the implications of the quotes I posted by failing to consider them in the overall context in which they were made – specifically the role of the 11 specific genes found in the study in protecting the body from external (environmental in origin) threats:
“What we found were 11 specific genes with expression levels that were significantly higher in the blood of children with autism when compared to the blood of typically developing children,” said Frank Sharp, senior author of the study and professor of neurology with the M.I.N.D. Institute.
“Those 11 genes are all known to be expressed by natural-killer cells, which are cells in the immune system necessary for mounting a defense against infected cells. We were surprised by our results because we were not looking for these particular genes. And while a number of studies have shown immune system dysregulation to be an important factor in autism, ours is one of the first to implicate these particular cells.”
Obviously no specific threats were identified with any certainty but the implication of environmental or external factors seems clear to my reading of the authors comments. That implication, based on the research findings seems to be more than simple speculation. No doubt more research, a lot more research, needs to be done in that direction, but it is direction that clearly should be pursued with more research.
I’m very thankful to have Emily clarifying and explaining the complexities of the study, with careful reference back to the original article.
I look forward to further research and balanced, informed, commentary about all possible factors, including possible environmental causal factors pertaining to autism.
One of the researchers who has often been cited for her work on autism and the environment is Martha Herbert, who is now also talking about the benefits of psychoanalysis for autistic children. I think her interest in psychoanalysis for autistic children raises some interesting, and troubling, questions about Herbert’s views about autism.
“Obviously no specific threats were identified with any certainty but the implication of environmental or external factors seems clear to my reading of the authors comments. That implication, based on the research findings seems to be more than simple speculation. No doubt more research, a lot more research, needs to be done in that direction, but it is direction that clearly should be pursued with more research.”
I never said it was “simple speculation.” I described it as their informed speculation and an expression of their hypothesis for what they’re seeing, which they will no doubt now pursue testing. You appear to be constructing an argument not of my own making and attempting to rebut it. I’m not sure why, but given that I cannot see what “we’re” arguing about here, I’ll just concede to you whatever point it is you’re trying to make.
As a scientist, I will say it once again: One cannot interpret the results of a genomics analysis such as this as anything but what it is: an indication of specific expression changes of a suite of genes in association with a specific condition. This is *not* a mechanistic study and does *not* reflect any data about environmental involvement. My saying that is neither an attempt to diminish the probable role of environmental factors in autism nor to make a statement in relation to any agenda. It is simply what the study tells us.
I personally think that it would be sheer idiocy not to assume environmental influence on the function and expression of *most* genes, not just those involved in autism. Genes do not operate in a vaccuum and the regulatory systems are finely tuned to environmental cues. Thus, I would *never* state that environment is not a causative element in expression of a gene. Even in many autosomal dominant disorders, environment modulates the penetrance, even though possessing the single disease-causing allele produces the disease.
@Emily (and thanks for your patience—I’m learning a lot from your careful explanations),
Quoting from your response to Harold L. Doherty:
“You appear to be constructing an argument not of my own making and attempting to rebut it.”
This is the usual “comment behavior” of the commenter in question; another blogger noted that Harold L. Doherty likes to build a “straw man army (and here’s a parody by Harold L. Doherty of Autism Vox).
Thank you Emily.
I appreciate the perspective and explanation from someone who has the relevant education and professional experience to tell us what the study tells us, and what it doesn’t.
Thanks for taking the time to do so.
Regan
I agree with your point Regan. That is why I quoted the researchers who were actually involved with the study in question.
Ms Chew
Personal smearing does not become you.
So good to hear from you, Harold L. Doherty, and thank you too for this post.
Hi, Harold–
That’s a sweet little attempt at a swipe (even someone like me can see that! Woohoo), but those of us who actually live in the world of science know that our results are out there for interpretation and that many of the people who interpret them will know plenty about what they’re doing, sometimes as much as those of us who did the work. I haven’t said anything on here that’s even a fundamental criticism of the study.
You seem willfully intent on ascribing to me some kind of agenda that is counter to your own, whatever that is. I don’t do agendas because they’re limiting and boring and constrained, and I don’t engage in manufactured controversies or arguments because that’s a complete waste of time. People who pit creationism against evolution do the same thing you appear to be trying to do now–manufacture a debate out of nothing–and it results in an extraordinary amount of nonsensical blather that we could all live without. I can’t even tell what you’re trying to argue here; all I can see is that you are trying to say something about me and my intentions in the process. Don’t. You wouldn’t be able to figure me out even if you knew me personally, and it certainly isn’t going to happen in the virtual world.
What you say here reads like repetitive incomprehensible gibberish to me–I can’t help that interpretation because when an agenda sticks out as much as yours does, I can’t understand the words any more because of the distraction.
Everything I’ve said here about this study is completely defensible under normally accepted standards of scientific discourse among scientists. As for relying only on what the researchers involved in the study have to say about their own work–that’s certainly not something that we in the scientific community would do. We poke and prod and evaluate and expect others to do so, too. If memory serves, some folks in the UK found themselves neck deep in a years’-long sham of a “link” between vaccines and autism because too many people relied on what the researchers who did the study had to say.
I thought this study interesting. At this time, I am leaning of the opinion that autism is somehow linked to autoimmune disorders. I have a family history of allergies, diabetes, MS, Chronic Fatigue Syndrome, etc. What is really interesting is that there are 3 cases of CFS (2 on the maternal side – 1 on fraternal) which is not all that common. However, I have no idea if anybody could come up with a family history of autoimmune disorders if one looks far enough.
Interesting—-allergies on my husband’s side and also on mine, plus the asthma I had as a child (and no longer do). Both my father-in-law and sister-in-law have MS (though she does not, as far as I know, have symptoms right now).
The answers you are searching for will never be discovered until you focus on central control and command. That is the central nervous system. I have been observing something I discovered over a decade ago. It’s a virus. The virus is a lenti-retrovirus. It uses a special enzyme(reversetranscriptese) to operate with. It is also an intracellular virus. Our brain is where it resides. It can be found in all bodily fluids. They call it SIV(simian immune virus). I call it CNSV(central nervous system virus). It also has been called the master of disguise. It has eluded 99.9% of all humans on this planet. I could continue to educate you about this virus,but ten years of facts cannot be explained in this e-mail.I have always felt a moral and ethical obligation to tell someone about the facts,symptoms, and microscopic details that I discoverd about this virus. Sincerely Kristal
Jill – my family have CFS, Aspergers and MS.
It would be interesting to compare this gene expression study on autism with Jonathan Kerr’s gene expression studies on CFS (Myalgic Encephalomyelitis).
I know that this study cannot say anything about cause of autism, but I have been wondering whether it is possible that if the symptoms of autism are caused partly by excessive brain plasticity, if the excessive brain plasticity has been triggered as the brain’s attempt to repair some kind of damage.
George Ebers on genetic susceptibility to MS is interesting stuff.