Contribution from Tony Michael - May 15th 1997

First things first - I see Malcolm's sense of humour hasn't improved. Even
worse, I actually GOT his joke (and laughed)!!

On to Martin's questions re: ovarian cortisol metabolism.

The simple answer to both questions is "I don't know" (so turn off now
unless you're really interested).

>"Do ovarian cells synthesize corticosteroids?"

Well, certainly luteinized cells synthesize micromolar concentrations of
progesterone and its theoretically possible to synthesize ALL steroids
(including cortisol and corticosterone) from this precursor. Indeed, even if
corticosteroid synthesis were to account for 0.1% of the progesterone
synthesized, this could still yield up to 100nM cortisol.

Practically, two enzymes are needed to turn progesterone into a
glucocorticoid: 21-hydroxylase and 11=DF-hydroxylase. Last time that I
checked, nobody was certain if both of these are expressed in the ovary
(unless anyone knows otherwise) and so the potential for local
glucocorticoid synthesis was thought to be limited. However, worth bearing
in mind that the ovary DOES contain a reasonable amount of
11=DF-hydroxyprogesterone, and with this as a precursor, it would only take
21-hydroxylase to make corticosterone (plus 17a-hydroxylase for cortisol)
which seems more feasible. Certainly there has been talk in the past of
"adrenal cell islands" within the ovary that could express the enzymes
necessary for local glucocorticoid synthesis.

After this rambling, and not wishing to commit myself to an answer (!), I
would just add that there is no arterio-venous difference in glucocorticoid
concentrations across the ovary (circa 500nM in the ovarian vein and artery)
and yet the concentration of cortisol in follicular fluid ranges between 100
and 400nM. I can only take this to mean that either cortisol is synthesized
within the ovary OR that glucocorticoids are somehow concentrated within the
follicle at a rate that does not lead to an obvious depletion of cortisol
from the ovarian circulation.


>"Is ovarian 11=DFHSD geared up for plasma or local concentrations of
glucocorticoids?"

Hard to say. Superficially the concentrations of cortisol in systemic plasma
and ovarian follicular fluid are comparable, but whereas plasma cortisol is
bound predominantly to cortisol binding globulin (CBG), follicular cortisol
is almost exclusively free, having been liberated from CBG by the binding of
progesterone.

Before signing off, I would like to quickly clear up the kinetic arguments
for the two cloned isoforms of 11=DFHSD (the enzyme that inactivates
cortisol). Type 1 11=DFHSD has a high =B5M Km (i.e. a low affinity for=
 cortisol) whereas type 2 11=DFHSD has a low nM Km (i.e. a high substrate 
affinity). Conventional wisdom has it that only enzymes with a high affinity for
substrate can have a physiological relevance. After all, how relevant can an
enzyme be if it needs pharmacological concentrations of steroid to even
achieve half its maximum velocity? HOWEVER, I propose that this argument is
flawed as follows. An enzyme with a high affinity/low Km is easily saturated
and would be unable to cope with any increase in plasma concentrations of
hormone. Specifically, in the case of type 2 11=DFHSD, this enzyme is
saturated at 120nM cortisol. Since this is the nadir value for plasma
cortisol concentrations, one would expect type 2 11=DFHSD to be working flat
out throughout the day, and as the steroid concentration rises from 150nM to
600nM (following a circadian rhythm), all cortisol over and above 120nM
would have free access to corticosteroid receptors, making a mockery of the
guardian role attributed to type 2 11=DFHSD. On the other hand, although=
type 1 11=DFHSD has a low affinity/high Km, this means that as the concentration=
of cortisol increases 4-fold from 150 to 600nM, the enzyme merely works 4 times
faster, presenting a relatively effective barrier to changing hormone concentrations.

I have been puzzling over the above for some years now, and would be
grateful for feedback on this logic, even if you're not interested in cortisol 
metabolism.

In closing, I'd add that in our studies of human granulosa-lutein cell
cultures, we've identified both a low affinity (NADP-dependent) 11=DFHSD
isoform (apparently type 1 11=DFHSD) AND a novel high affinity=
NADP-dependent isoform. So in short, human ovarian cells should be able to 
inactivate cortisol across a wide range of concentrations.

Thanks for the questions Martin.


Tony Michael

Dr.Tony Michael
Lecturer in Biochemistry & Molecular Biology,
Dept.Biochemistry & Molecular Biology,                   Tel 0171-794-0500 x.4999 (Office)
Royal Free Hospital School of Medicine,                 Tel 0171-794-0500 x.4988 (Lab)
Rowland Hill Street,
Fax : 0171-794-9645
London NW3 2PF