Urinary DENND1A.V2 as a Predictor of Pubertal Hyperandrogenemia

To date, the mechanisms underlying excess androgen production from steroid producing tissues
have been unclear, but recent Genome Wide Association Studies (GWAS) have provided additional
data in this regard. Initial GWAS data in a Han Chinese population identified some 12 loci as
potential PCOS susceptibility gene candidates. One of these loci, the DENND1A locus at
9q22.32, has been replicated in GWAS studies in both Asian and European populations. Strauss
and colleagues recently demonstrated DENND1A expression in testes, ovarian theca cells, and
H295 adrenal carcinoma cells — all androgen producing tissues/cells. DENND1A is therefore an
exceedingly strong PCOS susceptibility gene candidate. However, it remains unclear how
DENND1A may contribute to the PCOS phenotype.

Importantly, there are two transcriptional forms of the DENND1A gene — a consequence of
alternate splicing. The larger transcript (DENND1A Variant 1, or DENND1A.V1) encodes a 1009
amino acid protein, while the smaller transcript DENND1A.V2 encodes a truncated 559 amino
acid. The product of DENND1A.V2 contains the DENN domain and a clathrin-binding domain, but
differs from the product of DENND1A.V1 in two ways: (1) it does not contain the proline-rich
C-terminal domain present in Variant 1, and (2) DENND1A.V2 results from differential splicing
and contains a unique 33 amino acid C-terminal sequence. Of significant interest, published
studies by Drs. McAllister and Strauss (McAllister JM, et al. Proc Natl Acad Sci USA.
2014;111:E1519-27) strongly implicate the DENND1A.V2 splice variant as a contributor to
excessive theca cell androgen production in PCOS:

- Expression of DENND1A.V2 protein in cultured theca cells isolated from women with PCOS
was over 3-fold elevated compared to normal ovarian theca cells. Similarly, DENND1A.V2
mRNA abundance was elevated in PCOS theca and correlated with increased theca cell
androgen (dehydroepiandrosterone [DHEA]) production.

- Forced expression of DENND1A.V2 in normal theca cells increased expression of CYP17A1
and CYP11A1 (genes for key steroidogenic enzymes) along with androgen/androgen precursor
(e.g., 17OHP4, DHEA, testosterone) and progesterone production.

- Knockdown of DENND1A.V2 in cultured PCOS theca cells reduced CYP17A1 and CYP11A1
expression in addition to 17OHP4 and DHEA production.

- Treatment of cultured PCOS theca cells with anti-DENND1A.V2 IgG antibodies reduced
expression of CYP17A1 and CYP11A1 mRNA expression as well as DHEA and 17OHP4 synthesis.

These data provide strong support to the contention that the DENND1A.V2 splice variant is a
major factor underlying the phenotype of cultured PCOS theca cells and, thus, the etiology of
ovarian hyperandrogenemia in PCOS. By extension, the DENND1A.V2 splice variant may be a
significant contributor to peripubertal hyperandrogenemia, and it may explain much of the
variable hyperandrogenemia observed in obese peripubertal girls. The investigators propose
that, in some peripubertal girls, increased expression of DENND1A.V2 in ovarian and/or
adrenal cells promotes excess androgen secretion with the advent of puberty (when ovarian
stimulation by LH increases) and/or in the presence of obesity (e.g., enhanced insulin
secretion, augmenting LH-stimulated ovarian androgen synthesis and/or ACTH-stimulated adrenal
androgen synthesis).

Of great interest, preliminary data from 5 normal women and 6 women with PCOS suggested that
urinary exosomal DENND1A.V2 mRNA was approximately 3-fold elevated in women with PCOS
compared to normally-cycling controls (McAllister JM, et al. Proc Natl Acad Sci USA.
2014;111:E1519-27). Exosomes are small vesicles (40-100 nm) that contain nucleic acids shed
by cells into blood and urine, and they are known to be a stable source of RNA. Thus, urinary
exosomal DENND1A.V2 quantity may serve as a marker of DENND1A.V2 activity in the
ovary/adrenal.

In these proposed studies, the investigators (including collaborators Drs. McAllister and
Strauss) will obtain urine for exosomal DENND1A.V2 mRNA quantitation in addition to blood
samples to measure androgen levels, LH and insulin (the latter two previously-demonstrated to
predict free testosterone). The investigators will test the hypothesis that the quantity of
urinary exosomal DENND1A.V2 mRNA is a strong and independent predictor of androgen levels in
peripubertal girls.

To the investigators' knowledge, no data exist that tie a PCOS susceptibility gene candidate
to pubertal hyperandrogenemia. The investigators propose to evaluate a novel biomarker
(urinary exosomal DENND1A.V2 mRNA) as a predictor of free testosterone in pubertal girls.
Evaluation of this biomarker in adolescents is highly innovative.

The present proposal will build on Aim 3 of Project 1 (NIH P50 HD28934), in which the
investigators carefully assess the determinants of hyperandrogenemia in girls with
peripubertal obesity. Specifically, Aim 3 is an intensive protocol involving (a) overnight
frequent blood sampling for LH and insulin along with (b) ovarian and adrenal stimulation
protocols over the subsequent 2 days. Some of the subjects participating in this proposed
pilot study will also participate in Aim 3. Correlation between urinary exosomal DENND1A.V2
mRNA and ovarian/adrenal stimulation protocols will allow a preliminary assessment of urinary
exosomal DENND1A.V2 mRNA as a non-invasive measure of ovarian/adrenal steroidogenesis — a
potentially important clinical tool.

Peripubertal girls with hyperandrogenemia are believed to be at high risk for developing
adolescent/adult PCOS, but since the manifestations of PCOS can overlap with the findings of
normal puberty, the diagnosis of PCOS may be unreliable until later adolescence. If urinary
exosomal DENND1A.V2 mRNA is a significant predictor of hyperandrogenemia in this cohort,
future studies will determine the utility of urinary exosomal DENND1A.V2 as a diagnostic test
to predict the development of PCOS across puberty. This is an innovative and exceedingly
exciting prospect that may have significant potential implications for at-risk girls.

List of Inclusion Criteria

• Peripubertal girls, Tanner breast stages 1-5

List of Exclusion Criteria

- Age < 8 or > 17 y

- Men and boys are excluded

- Inability to obtain proper consent/assent

- Atypical obesity

- Underweight: Underweight is defined as a BMI-for-age percentile < 5

- Positive pregnancy test or lactation

- Assessment during the luteal phase as suggested by a serum progesterone ≥ 1.5 ng/ml

- Virilization or a total testosterone > 150 ng/dl

- Excessively elevated DHEA-S: This will be defined as a DHEA-S > 1.5 times the
age-appropriate upper limit of normal

- Congenital adrenal hyperplasia (CAH)

- Cortisol deficiency/excess

- Inadequately-treated or unstable thyroid dysfunction

- Significant hyperprolactinemia: Since mild elevations may be seen in girls with
hyperandrogenemia or PCOS, elevations up to 30 (i.e., 1.5 times the upper limit of
normal) will be accepted in such girls

- Significant chronic medical history: This includes a significant history of cardiac or
pulmonary dysfunction (e.g., known or suspected congestive heart failure; asthma
requiring intermittent systemic corticosteroids; etc.); history of renal insufficiency
or durable electrolyte abnormalities; or a history of substantial liver disease. A
history of liver test abnormalities will be allowed in two circumstances: (1) mild
bilirubin elevations will be accepted in the setting of known Gilbert's syndrome; (2)
mild transaminase (ALT, AST) elevations may be seen in obese girls, so stable
elevations < 1.5 times the upper limit of normal will be accepted in this group.

- Uncontrolled type 2 diabetes mellitus: This will be reflected by a hemoglobin A1c >
7.0%. Subjects with impaired glucose tolerance or a diagnosis of type 2 diabetes that
is well-controlled with lifestyle management alone will be allowed to participate.

- Type 1 diabetes mellitus: Since subjects with type 1 diabetes invariably require
exogenous insulin, they will not be allowed to participate.