On the Almost Two-Year Anniversary of the Germline Disruption Hypothesis of Autism

Nearly two years ago I sent the following email to a large group of researchers and autism leaders (including Tom Insel at NIMH, Autism Speaks, and noted scientists from UCLA to Texas to Hopkins) outlining for the first time the germline disruption hypothesis of autism.

So much has happened since that time to confirm my suspicions — meeting many other autism parents with circa 1960s prenatal drug exposures just like mine, studies of multigenerational effects of endocrine disruptors and hypomethylation in animal models, findings of various multigenerational effects in human cohort studies, recognition of epigenetic involvement in some neurodevelopmental impairments, findings regarding de novo alterations in germline in autism, increasing recognition of gene-environment interaction in autism, to name a few — and as we head into the holiday season let us hope for a 2014 filled with long-overdue public awareness about this most urgent issue.

Jill Escher
-----------------------------------------------------------

Epigenetic Transgenerational Inheritance: How Old Pharmaceuticals May Be Contributing to the Autism Epidemic Today

Summary Document, January 20, 2012

Hypothesis

That the increasing rate of ASDs stems in part from epigenetic disturbance of fetal germ cells caused by in-utero exposures to various pharmaceuticals, including synthetic steroid hormones, used in obstetric practice from roughly the 1950s to 1980s.  In other words, that gestating women given certain pharmaceutical drugs in roughly the first half of pregnancy (likely window is weeks 6-18 of gestation) are more likely to have neurologically impaired grandchildren, through the action of epigenomic alterations of fetal germ cells.

Background

From the 1950s through 1970s, and even before, it was common in obstetric practice to prescribe various pharmaceuticals to pregnant women, even in the first trimester.  The drugs included, among other things, hormone-like compounds, barbiturates, and morning-sickness medications.

In the post-war period chemists began synthesizing various progesterone-like compounds. Based on theory that these compounds could help prevent miscarriage or preterm birth, some obstetricians and early fertility clinics began administering synthetic progestins to pregnant women deemed to have at-risk pregnancies. Commonly, the patients had suffered prior miscarriages or had complications such as bleeding. Use of the compounds was relatively rare in the 1950s, but increased through the 1960s and 70s, particularly in certain locales with a large upper-middle class clientele who could afford the then-expensive fertility treatments, including west Los Angeles and New Jersey, the sites of cohort studies by Dr. June Reinisch.

There was no standard practice for use of these compounds, according to Reinisch, it was more of an art than a science.  Some physicians administered multiple hormones, some used hormones early in pregnancy, some later.  There was no standard dosing.

Little thought was given to the impact of the various drugs, including the progestins, on the exposed fetus.  The progestins were chemically similar to progesterone, except for a small “tail” or “radical” hanging off the end of the molecule. They were not, like DES (diethylstilbestrol), very chemically different from the natural counterparts.  While the idea of impact on fetal germ cells did come up, it was thought at the time that the germ cells would be protected from exogenous chemical exposures.

Dr. Reinisch and a few others did perform research on first-generation impact of prenatal exposure to synthetic steroid hormones.  Impacts on brain development were found, though subtle compared to a neurodevelopmental disorder like autism.  The impacts were primarily on personality traits.  See, eg, Reinisch and Karow,“Prenatal exposure to synthetic progestins and estrogens: Effects on human development.” http://www.springerlink.com/content/k0552583r63776h2/ and “Prenatal exposure to synthetic progestins increases potential for aggression in humans,” http://www.sciencemag.org/content/211/4487/1171.abstract?sid=792459ee-eacf-47d5-8dd3-9e5df26b6b85.

Some of the pharmaceuticals used

According to Prenatal Exposure to Synthetic Progestins and Estrogens: Effects on Human Development, Archives of Sexual Behavior, Vol 6, No. 4, 1977, p. 267, the most common synthetic steroid hormones used in pregnancies deemed at-risk appear to have been (progestins) Colprosterone, Delalutin, Deluteval, Norlutin Acetate, Provera, Provest, and (additional estrogens) Stilbestrol.  Full list from that study:

Progestins
Broxorone (Squibb)
Colprosterone (e)
Deladroxate 110 (Squibb)
Deladroxate 130 (Squibb)
Deladroxate 150 (Squibb)
Delalutin 142 (Squibb)
Delaxadrone (Squibb)
Deluteval (b) (Squibb)
Depo-Provera (Upjohn)
MK665(c) (Merck) ethymerone
19NET, Norlutin (Syntex)
Norlutin Acetate (Parke Davis)
Norethynodrel (b) (Searle)
Pranone (Schering)
Provera (Upjohn)
Provest (b) (Upjohn)
RS1280 (c, e) (progestogen)
SC4641 (c) (Searle) 19 NET
SC4642 (c) (Searle) norethynodrel
SC9022 (c) (Searle) methylnortestosterone
SC10230 (c,g) (Searle)
SC11800 (c) (Searle) ethyndiol
1 142.53 (e) aqueous progesterone
Additional estrogens (d)
Allyl Estranol (Organon)
Amestrogen (Squibb)
Delestrogen (Squibb)
Estrone (a)
Hexestrol (e)
Mestranol (e)
Stilbestrol (e)
Other drugs
Thyroid (Armour)
Cytomel (Smith, Klein& French)
Methergine (Sandoz)
Prednisone (McKesson)
Proloid (Wamer-Chilcott)
Sterane (Pfizer)
Synthroid (Flint)

Notes:
a Many mothers received more than one drug.
b Compounds that have estrogens included.
C Experimental compounds.
d Not including estrogen found in combination with progestin.
e Company not identified.
f Number of pregnancies in which medication was administered (omitted in this list, see article)
g 21-Fluoro-17-hydr oxy-6-methylpregna-4,6-diene-

3,20-dione acetate.

Other pharmaceuticals, such as the barbiturates and morning sickness medications, were also prevalent, however I do not have a list of those medications at this time.

The “Hidden History”

Reinisch calls the past use of these hormone-like pharmaceuticals, as well as the barbiturates and morning sickness drugs, the “hidden” history for several reasons:
--The records have largely been destroyed
--The women who received the medications often did not know what they were getting
--The women have largely forgotten taking the medications
--The exposed children were never told of their exposure
--The practice was very little studied, and has been forgotten by contemporary researchers and medical practitioners

Case Study

By the end of 1964 a well-to-do 27 year-old mother of one son in Los Angeles had suffered two miscarriages and badly wanted another child.  She sought treatment at a pioneering private fertility clinic in west Los Angeles which was associated with UCLA.  She conceived after being given clomiphene and Pergonal to induce ovulation.  Through at least the first two trimesters, she was then given regular doses of a synthetic progestin, Deluteval 2x, manufactured by Squibb, for the ostensible prevention of miscarriage (these compounds were never actually proven to promote gestation).  It appears Prednisone was also administered at some point.  She delivered a healthy baby girl in September 1965. 

This child now has three children of her own (“grandchildren”), two of whom have markedly abnormal neurological development (label used is autism).  All genetic tests have been negative, including tests for cnv’s and microdeletions.  All pregnancies and deliveries were normal, and Apgar scores were high.  The families of the grandchildren have no history of neurological problems or mental illness.  There were no unusual exposures during the pregnancies.  In spite of the very evident neurological dysfunction, the children are robustly healthy and normal-looking.

The child born in 1965 also has a younger brother, born in 1968.  He, too, was exposed to exogenous synthetic steroid hormones.  He is the father of two children, one of whom has a mysterious neurological disorder that is similar to NF1, though apparently not genetic in origin.  The older brother, unexposed to any hormones, has two normal children.

Several other autism parents (both mothers and fathers) are also aware of their own in-utero exposures to synthetic steroid hormones:  symptoms reported in the offspring include Aspergers, autism, sensory processing disorder, and learning disorders.  Many of these second-generation offspring also have no apparent, or very mild, symptoms.

However, the vast majority of the parents who were exposed have no idea about their prenatal histories--nearly all records have long ago been destroyed, and it appears to have been uncommon for a mother to think of the medication as worth mentioning to her offspring.

Background Research

It has been postulated that common pharmaceuticals could have epigenetic side-effects that appear only years or decades after the exposure.  See, eg, Szyf, M “Epigenetic side-effects of common pharmaceuticals: A potential new field in medicine and pharmacology.”  http://www.medical-hypotheses.com/article/S0306-9877%2809%2900291-6/abstract

Researchers such as Michael Skinner, Andrea Gore, Moshe Szyf, and Joseph Nadeau have looked into transgenerational epigenetic impact of chemical exposures and have found early evidence for the phenomenon.  See, for example, Nadeau “Transgenerational genetic effects on phenotypic variation and disease risk,” http://m.hmg.oxfordjournals.org/content/18/R2/R202.abstract, Skinner et al, “Endocrine Disruptor Vinclozolin Induced Epigenetic Transgenerational Adult-Onset Disease,” http://endo.endojournals.org/content/147/12/5515.short, and Skinner, “Role of Epigenetics in Developmental Biology and Transgenerational Inheritance,” http://onlinelibrary.wiley.com/doi/10.1002/bdrc.20199/full.

New studies on endocrine disruptors, some yet unpublished, point to differing epigenetic impacts on the F2 and F3 generations, likely due to impact on developing germ cells. (Interviews with Szyf and Gore, December 2011, and Skinner, January 2012.)

There is broad agreement, even among traditional geneticists, that it is biochemically plausible that the synthetic hormones, or their metabolites, or other prenatal pharmaceuticals, could have had a deleterious impact on germ cell reprogramming. Various theories were espoused, including altered methylation patterns and impact on histone tails during reprogramming. It is agreed that at a minimum, synthetic steroid hormones would cross the placenta, enter fetal tissues, and likely bind to receptors in ways that vary from natural, endogenous steroids. The reasons for epigenetic alterations of germ cells to lead specifically to a phenotype of abnormal neurodevelopment are unknown.

Professor Skinner summarizes the core biochemical issue in his article cited above, “Role of Epigenetics in Developmental Biology and Transgenerational Inheritance” as follows (emphasis added):

“[T]he basic mechanism of epigenetic transgenerational inheritance involves the actions of an environmental factor (e.g., chemical or nutrition) during germ line remethylation at gonadal sex determination to permanently alter the germ line epigenome (Anway et al., 2005; Guerrero-Bosagna C et al., 2010; Skinner et al., 2010) to then transmit this altered germ line epigenome to subsequent generations (Anway et al., 2005; Anway et al., 2006a,b; Anway and Skinner, 2008; Guerrero-Bosagna C et al., 2010; Skinner et al., 2010). As the embryonic stem cell epigenome is altered due to this germ line transmission, all cell populations and tissues will have an altered epigenome and corresponding transcriptome (Anway et al., 2008; Skinner et al., 2010). The germ line generated by the next generation will also have this altered epigenome and transmit it to the subsequent generation (Guerrero-Bosagna C et al., 2010; Skinner et al., 2010). Exposure to the endocrine disruptors at other times of development do not appear to have the capacity to permanently alter the germ line epigenome (Anway et al., 2005; Anway and Skinner, 2008; Skinner et al., 2010). Of course, the vast majority of exposures will alter the somatic cells at critical periods of development to modify later cellular development and potential adult onset disease, Figure 1, but this does not have the capacity to become transgenerational as the germ line is not involved (Skinner et al., 2010). Epigenetic transgenerational inheritance through a permanently altered epigenome of the germ line has the capacity to have a dramatic influence on developmental biology, as well as other areas of biology such as evolution.

“In the event the base-line epigenome is altered, then the cascade of epigenetic and genetic steps during development will be altered and a modified differentiated or developmental state achieved, Figure 1. Therefore, epigenetic transgenerational inheritance has a dramatic effect on the developmental biology of all cells and tissues derived from the germ line transmitting this modified baseline epigenome. Although not all cell types or tissues will develop a disease state, those tissues that have a sufficiently altered transcriptome will have a greater susceptibility to develop disease (Skinner et al., 2010). As all development and differentiation processes involve a cascade of epigenetic and genetic steps, alteration of the baseline epigenome, similar to alteration in the genetic baseline, will have the capacity to promote abnormal development which may lead to disease later in life. For this reason, environmentally induced epigenetic transgenerational inheritance through the germ line will have a significant impact on developmental biology. This mechanism and consideration of the cascade of integrated epigenetic and genetic events during development, Figure 1, will be an important factor in disease etiology not previously considered (Skinner et al., 2010).”

What this hypothesis might explain

It goes without saying that there are many roads to the various phenotypes we lump together under the label of autism.  Among them: genetic disorders, prenatal and perinatal complications, prematurity and multiple birth, prenatal exposure to certain pathogens or chemicals, and seizure disorders.  However, it is worth noting that the pharmaceutical/epigenetic hypothesis is consistent with a great many research findings, as follows:

--Abnormally high rates of autism in west Los Angeles (the site of early, aggressive fertility and prenatal treatments beginning in the 50s)
--Abnormally high rates of autism in New Jersey (same)
--Rising rates of autism diagnoses after about 1990 (the time when the early crop of grandchildren of the growing number of women treated in the 1960s were entering their preschool years)
--Findings of certain epigenetic signatures in autism brains
--The lack of evidence for a genetic root of most cases of ASD
--Findings implicating very early neurological development in ASDs
--Certain demographic patterns, eg, why certain religious or ethnic populations that had little access to or interest in prenatal care have lower autism rates
--The vast heterogeneity of the ASD syndrome

Some of the researchers with whom I spoke agreed that pharmaceuticals could be playing a role, but that other low-level environmental exposures such as to plastics, endocrine disruptors, stress and hydrocarbons could also be contributing.  I feel the need for an editorial comment here:  I cannot fathom how ubiquitous and low-level exposures like those can possibly compare to the sometimes huge direct doses of potent (though forgotten) pharmaceuticals injected directly into our mothers during the first half of pregnancy.  The study of the pharmaceutical impacts appears to be of much greater urgency.

Early Initiatives

Various early initiatives have been proposed:

Epidemiology.  There is broad agreement that epidemiological work should be done.  There exist at least four distinct cohorts of individuals known to have been exposed to exogenous hormones in utero.  It would be important to do prospective studies (starting with obstetric records of the grandmothers, and working downward), and not retrospective studies (starting with autistic kids, and working up the family tree).  This is due to the absence of reliable information of past pharmaceutical use of the grandmothers. To do meaningful epidemiological work, one would need not only the basic knowledge of “exposure” but also the identities of the drugs used, the dosages, and the timing.

--Denmark (Dr. Reinisch is looking into the possibility of a second-generation study on this cohort, which is part of the Prenatal Development Project at Kinsey)
--Finland (Dr. Brown is looking into the possibility of a second-generation study on this cohort, which had been identified by E Hemminki as part of a study into the fertility rates in this cohort)
--Los Angeles (Dr. Reinisch’s detailed files on individuals exposed in utero to synthetic steroid hormones, in the 1960s and 70s, still exist, in storage at the Kinsey Institute)
--New Jersey (same as above)

Quite possibly, with some sleuthing, cohorts exposed in utero to various pharmaceuticals could be found within initiatives at Kaiser and at Johns Hopkins as well.

Broader epidemiological work, looking at data from several countries at once, was also suggested.  However, the integrity of the data regarding the F1 (grandmother) pharmaceutical use is in question.

History.  A few researchers suggested doing a basic history of past obstetric pharmaceutical use around the world, since it is so poorly understood, and attempting to make connections to the development of various diseases, including ASDs, increasing in prevalence today.

Animal models.  A great many researchers suggested three-generation animal modeling.  This way, different compounds and different timing of exposure could be tested for resulting changes in behavior, brain morphology, and epigenetic changes.

Case studies.  A few researchers suggested doing direct epigenetic studies on affected families. However, others cautioned that such studies would be difficult owing to the complexity involved in studying the epigenome.

In summary

Long-forgotten pharmaceutical use may be playing a significant role in the autism epidemic today.  If this is the case, there is urgent work to be done in terms of prevention, family planning, changing fertility and obstetric medical practice, reducing pharmaceutical use by women of childbearing age by shifting emphasis to natural treatments of chronic conditions, ensuring all prenatal medical records remain available and never discarded, and possibly, finding targeted therapies for the many disabled individuals suffering from artificially altered epigenomes.  Therefore, expediting research into this hypothesis is critical.