onapristone/progesterone (in rats) experiment/treatment for 1A

[Guest guest]

From Nature Medicine Nat Med 9(12):1533-1537, 2003.

Therapeutic Administration of Progesterone Antagonist in a Model of

Charcot-Marie-Tooth Disease (CMT-1A) Posted 12/30/2003 (Full Text with

exception of ables and figures)

W Sereda; Gerd Meyer zu Hörste; Ueli Suter; Naureen Uzma;

Klaus-Armin Nave

Abstract and Introduction

Abstract

Charcot-Marie-Tooth disease (CMT) is the most common inherited

neuropathy. The predominant subtype, CMT-1A, accounts for more than 50%

of all cases[1] and is associated with an interstitial chromosomal

duplication of 17p12 (refs. 2,3). We have generated a model of CMT-1A by

introducing extra copies of the responsible disease gene,

Pmp22 (encoding the peripheral myelin protein of 22 kDa), into

transgenic rats.[4] Here, we used this model to test whether

progesterone, a regulator of the myelin genes Pmp22 and myelin protein

zero (Mpz) in cultured Schwann cells, can modulate the progressive

neuropathy caused by moderate overexpression of Pmp22. Male transgenic

rats (n = 84) were randomly assigned into three treatment groups:

progesterone, progesterone antagonist (onapristone) and placebo control.

Daily administration of progesterone elevated the steady-state levels of

Pmp22 and Mpz mRNA in the sciatic nerve, resulting in enhanced Schwann

cell pathology and a more progressive clinical neuropathy. In

contrast, administration of the selective progesterone receptor

antagonist reduced overexpression of Pmp22 and improved the CMT

phenotype, without obvious side effects, in

wild-type or transgenic rats. Taken together, these data provide proof

of principle that the progesterone receptor of myelin-forming Schwann

cells is a promising pharmacological target for therapy of CMT-1A.

Introduction

CMT-1A is associated with overexpression of PMP-22, a tetraspan myelin

protein in Schwann cells.[5] The progressive clinical phenotype shows

variability in humans, even in identical twins.[6,7] We generated a line

of Pmp22-transgenic rats,[4] but noticed marked interindividual

variations in their phenotype (Fig. 1a). When steady-state

levels of Pmp22 mRNA in sciatic nerves were quantified by real-time

RT-PCR, the CMT rats (n = 20) showed a 1.6-fold increase over wild-type

controls (n = 20; Fig. 1b). Although

significant (P = 0.046), the interindividual difference was high

compared with the interindividual difference in ß-actin mRNA, suggesting

additional epigenetic effects. On histological sections, muscles from

CMT rats contained numerous atrophic fibers intermingled with smaller

groups of hypertrophic fibers (Fig. 1c). Muscle atrophy, determined by

measuring the circumference of the hind limb, correlated with motor

performance in the bar test (r = 0.833; P = 0.0028; Fig. 1d).

Steroid hormones are epigenetic regulators of gene expression,[8-10] and

Pmp22 expression in cultured Schwann cells can be stimulated with

progesterone.[11-13] To determine whether circulating steroids can

modulate the CMT phenotype, we treated a cohort of Pmp22-transgenic male

rats and wild-type controls with progesterone. Female

CMT rats, although clinically indistinguishable from male CMT rats, were

excluded to avoid any interference with changing endogenous progesterone

levels. First, we analyzed sciatic nerves of wild-type rats and found

that Pmp22 mRNA was upregulated by progesterone (+25%; n = 8) when

compared with sesame oil vehicle-treated controls (n =

8; P = 0.037; data not shown). Next, we investigated the effect of

onapristone, a selective progesterone receptor antagonist originally

developed to treat primary breast cancer.[14] The genomic effect of

progesterone is mediated by two nuclear progesterone receptor proteins,

PRA and PRB.[15] Onapristone is a pure progesterone receptor

antagonist (type I) that impairs receptor binding to DNA and, in

contrast to mifepristone (type II), is without partial agonistic

activity.[16] Starting at postnatal day five (P5) and continuing for 7

weeks, we gave CMT and wild-type control rats daily subcutaneous

injections of progesterone (20 mg/kg; n = 31), onapristone (20 mg/kg; n

= 18) or sesame oil alone (n = 35). These doses were comparable to those

used in previous in vivo breast cancer models.[14] Animal weight was not

altered by progesterone or onapristone treatment, compared with

placebo-treated controls (data not shown).

Plasma concentrations of progesterone were determined using blood taken

from the left ventricle 24 h after the last injection. Progesterone

levels increased about threefold (from 8.9 ± 1.9 to 27.7 ± 4.9 ng/ml; P

< 0.016 compared with untreated controls). These values are below the

peak value of progesterone during the estrous cycle of untreated

female rats,[17] and are therefore within a nontoxic, 'physiological'

range (Fig. 2a). Although treated at a young age, we observed no obvious

developmental abnormalities in CMT rats and wild-type controls as a

result of progesterone or onapristone treatment (data not shown).

After 7 weeks, we dissected sciatic nerves and quantified Pmp22 mRNA.

PCR primer pairs were intron-spanning (Fig. 2b), and all PCR products

had the expected size (data not

shown). Primer combinations distinguished between total Pmp22 mRNA and

the exon 1B-containing form, the latter being non-steroid-responsive[11,

18] and serving as an

internal standard. We compared sciatic nerves from CMT rats treated with

onapristone (n = 15), progesterone (n = 16) and sesame oil (n = 16).

Pmp22 mRNA increased by ~30% after

progesterone treatment (P = 0.0015). In contrast, CMT rats treated with

onapristone showed a 15% decrease in Pmp22 mRNA compared with placebo (P

= 0.013; Fig. 2c). After progesterone treatment, we also observed a

1.5-fold (P = 0.022) increase in Mpz mRNA, which encodes the major

structural myelin protein of the peripheral nervous system

(Fig. 2d), in agreement with previous cell culture data.[11,12] There

was no significant (P = 0.36) downregulation of Mpz expression by

onapristone. It is possible that

the baseline expression of Mpz in Schwann cells is secured by mechanisms

that compensate for reduced progesterone stimulation.

The histological hallmarks of CMT-1A, demyelination and onion-bulb

formation, are present in sciatic nerves of the transgenic model.[4]

Progesterone-treated animals showed a higher fraction of unmyelinated

axons (9.0 ± 2.8%; n = 4) when compared with the onapristone (5.0 ±

1.5%; n = 5) or vehicle (5.9 ± 2.7%; n = 6; Fig. 3a) groups, but

interindividual variability was too high to detect a significant effect

of onapristone (P = 0.766). Likewise, myelin thickness was reduced in

CMT rats (g-ratio of 0.66 ± 0.008; n = 5), not obviously altered by

onapristone (g = 0.65 ± 0.008; n = 5), and further reduced by

progesterone (g = 0.72 ± 0.003;

n = 4; data not shown).

Onapristone's therapeutic effect was more obvious when scoring the

absolute number of sciatic nerve axons. CMT rats treated with

onapristone maintained a significantly larger

number of axons (7,509 ± 507; n = 5) than placebo-treated rats (5,803 ±

518; n = 6; P = 0.045; Fig. 3b). Progesterone-treated animals did not

differ significantly from controls (P = 0.53).

Next, we determined that Pmp22 mRNA levels (which were variable in all

treatment groups) are related to the percentage of unmyelinated axons of

the same nerve. This yielded a

strong correlation (P < 0.001) and established an important link between

modulation of Pmp22 gene expression and degree of histopathology (Fig.

3c).

Hormonal treatment did not affect the animals' weight and had no toxic

effects in transgenic or wild-type rats (data not shown), but modulation

of the CMT clinical phenotype was obvious. Although onapristone did not

'rescue' the disease, it improved motor performance. In the standardized

bar test,[4] there was no effect after 3 weeks, but 5 weeks of

onapristone treatment (Fig. 4a) increased the score (holding time) by

65% (P = 0.034). The therapeutic effect was still significant after 7

weeks (Fig. 4b).

In contrast, treatment of CMT rats with progesterone reduced motor

performance (Fig. 4a,. After 5 weeks, holding time was reduced by

nearly 60%. After 7 weeks, progesterone-treated animals showed motor

defects as severe as those in nontreated controls, suggesting a

saturation effect, in agreement with the axonal counts (Fig. 3b). To

confirm that the clinical phenotype was steroid-modulated, we also

assigned a

subjective score between '1' (fully normal) and '5' (paraplegic) when

handling and examining each animal. CMT rats treated with onapristone

for 5 weeks scored better (2.7 ± 0.2) than untreated CMT rats (3.4 ±

0.2; P = 0.022; Fig. 4c), and the result was even more obvious at 7

weeks (3.1 ± 0.2 versus 3.8 ± 0.2; P = 0.003; Fig. 4d). In contrast,

progesterone treatment worsened this clinical score at 5 weeks (3.4 ±

0.2 versus 4.1 ± 0.1; P = 0.026), but not at 7 weeks of age (Fig. 4c,d).

Thus, progesterone enhanced disease progression without altering the

clinical endpoint.

Finally, we correlated the fraction of unmyelinated sciatic nerve axons

in individual rats (of all treatment groups) with their individual motor

performance (Fig. 4e). This yielded an inverse correlation that was

highly significant (r = -0.78; P = 0.001), supporting the causal role of

histopathological changes in the development of motor deficits. That

clinical severity is a function of Pmp22 mRNA expression is suggested by

the nature of the disorder, and by the correlation (r = -0.42, P = 0.04)

between motor performance of CMT rats (n = 32) and individual Pmp22 mRNA

levels (Fig. 4f).

Taken together, these data show that onapristone improves the CMT

phenotype in Pmp22-transgenic rats by slowing disease progression. The

rats with the highest expression of Pmp22 mRNA seemed to respond

relatively poorly to onapristone, but at present we have no reliable way

of quantifying gene expression before the start of treatment.

Several transgenic models[4,19-21] have shown that moderate

overexpression of Pmp22 is sufficient to cause disease, and that this

phenotype is, in principle, reversible.[22]

Based on gene duplication, human Pmp22 overexpression is 50% in CMT-1A.

Theoretically, a small reduction (< 30%) in Pmp22 transcription may be

sufficient to alter the course of the disease. Ligand-gated

transcription factors are ideal pharmacological targets, as specific

antagonistic drugs have been developed that easily reach their targets

in vivo. In the present study, we used a transgenic model of CMT-1A to

show that onapristone reduces Pmp22 mRNA by 15%. Whether the gene is a

direct target of the progesterone receptor or regulated through the

progesterone-responsive transcription factor Egr-2 (ref. 23) remains to

be determined. It has been reported that progesterone stimulates Pmp22

expression in Schwann cells, suggesting a GABAA receptor-mediated

mechanism.[12] It would be interesting to compare the outcome of

anti-GABAergic treatment of CMT rats with our present data, and to

explore possible additive effects with the progesterone receptor

antagonist.

There are no obvious clinical differences between male and female human

CMT-1A patients, but case reports document a connection between

pregnancy (with up to tenfold increase

in progesterone plasma level) and the worsening of clinical symptoms of

CMT-1A.[24, 25] Our data suggest that progesterone should be taken with

caution by patients with PMP22 gene duplications. Future studies with

onapristone and more recently developed

antiprogestins (lacking the side effects reported for onapristone in

humans[26]) will include female CMT rats as a separate treatment arm,

and will extend into long-term observations (>1 year). This also

requires that the minimal effective dosage be defined, and that the

clinical rating of aged CMT rats be improved. Any clinical trial will

have to await the comparison of available antiprogestins, long-term

applications in males and females, and safety studies that address

possible side effects.[26,27] We hope that optimization of the

anti-progesterone strategy for CMT-1A treatment will benefit from the

available transgenic disease models.

The generation of Pmp22-transgenic CMT rats has been described.[4]

Routine genotyping was done by PCR, using genomic DNA from tail biopsies

and mouse transgene-specific primers under standard conditions, as

described.[4] All experiments were conducted according to the Lower

Saxony State regulations for animal experimentation in Germany.

Steroid Treatment

Randomly chosen male transgenic rats were treated with either

onapristone (ZK

98299; Schering) or progesterone (Sigma) at 0.02 mg per g body weight.

Steroids were suspended in 0.2 ml sesame oil (Sigma) at 20 mg/kg,

adapted to increasing body weight, and injected subcutaneously daily

between P5 and P49. Control animals received injections of sesame oil

without steroids. Age-matched wild-type males received an equal

treatment to determine possible toxic effects. All experiments were

carried out in a fully blinded fashion by the same investigator.

Radioimmunoassay

Twenty-four hours after the last injection of progesterone, rats were

killed by CO2 narcosis. Within 1 min, ~1 ml blood was obtained from the

left ventricle. Serum samples were analyzed for progesterone by

radioimmunoassay in a commercial laboratory.

Motor Tests and Clinical Scores

All experiments were conducted by the same investigator, who was blinded

toward the genotype and treatment groups. Motor performance was assessed

in a standardized bar test, in which rats were placed on a round metal

tube with a rough surface (length 60 cm), held horizontally 70 cm above

a soft pad. The rod diameter was 1.5 cm at 3

weeks of age, 2.5 cm at 5 weeks, and 3.0 cm at 7 weeks. Rats were placed

by hand onto the bar, and the time (in s) that the animals were able to

hold on was scored. In a series of six trials per rat, a maximum of 300

s per trial was allowed, and the average was calculated. Disease

severity was independently assessed using a clinical ranking

scale from 1 to 5, with the following criteria: 1, normal phenotype; 2,

unsteady gait but good grip and good movement on bar; 3, impaired gait,

good grip on bar but falling on movement; 4, impaired gait, no grip on

bar; 5, paraplegic.

Sciatic Nerve Histology

After 7 weeks of treatment, randomly chosen CMT rats were deeply

anesthetized and perfused with HBSS, followed by fixation with 5%

glutaraldehyde and 4% paraformaldehyde in cacodylate buffer (pH 7.2).

After perfusion, sciatic nerves were dissected and embedded in Epon.

Histological sections (1 µm) of the sciatic nerve were obtained 10 mm

distal to the sciatic notch. Sections were stained with methylene blue

and analyzed under a Zeiss Axiophot microscope, and images were

photodocumented (Improvision, Macintosh). Total axon numbers were

obtained (physiologically unmyelinated axons were not counted), and the

percentage of unmyelinated axons was calculated from one quadrant of the

section by a blinded observer. The g-ratio was defined as the numerical

ratio of

unmyelinated axon diameter to myelinated axon diameter, and was derived

from the images of at least 100 fibers per animal.

Muscle Pathology

Randomly chosen rats from all treatment groups were deeply narcotized

and perfused with HBSS, followed by 4% paraformaldehyde in HBSS. After

skinning, the hind-limb circumference was calculated by repeatedly

winding a nylon thread around the limb (10 mm from the knee joint).

Histological sections (5 µm) from the quadriceps muscle were

stained with H & E and analyzed under a Zeiss Axiophot microscope, and

images were photodocumented. Atrophic muscle fibers were counted and

their percentage out of all muscle fibers was determined.

Statistical Analysis

All values are expressed as mean ± s.e.m. For progesterone levels and

RNA analysis, the overall significance was determined using the Student

t-test for unrelated groups. Correlation analyses were performed using

the Spearman rank correlation test (StatSoft). The Wilcoxon-Mann-Whitney

U-test was used to analyze the motor tests and clinical

scores.

Note: Supplementary information is available on the Nature Medicine

website.

Acknowledgements

We thank E. Nicksch, U. Bode and C. Stünkel for technical help; M.R.

Schneider (Schering) for providing us with onapristone; C. Scheidt-Nave

for statistical advice; J.R. Lupski for helpful comments on the

manuscript; and members of the Nave lab for discussion.

Funding Information

This work was supported by the Max-Planck Society and by grants from the

European Union (to K.A.N.). U.S. was supported by the Swiss National

Science Foundation and by the

National Center for Competence in Research " Neural Plasticity and

Repair " . M.W.S. was supported in part by the Departments of Clinical

Neurophysiology and Neurology at

the University of Göttingen.

W Sereda,1,2,4 Gerd Meyer zu Hörste,1,4 Ueli Suter,3 Naureen

Uzma,1 Klaus-Armin Nave1

1Max-Planck Institute of Experimental Medicine, Department of

Neurogenetics, Hermann-Rein-Str. 3, D-37075 Göttingen, Germany

2Departments of Neurology and Clinical Neurophysiology, University of

Göttingen, -Koch Str. 40, D-37075 Göttingen, Germany

3Institute of Cell Biology, Department of Biology, Swiss Federal

Institute of Technology, ETH Hoenggerberg, 8093 Zürich, Switzerland

4These authors contributed equally to this work.