21-Hydroxylase deficiency, a potentially fatal disease due to deletions or mutations of the cytochrome P450 21-hydroxylase gene (CYP21), causes congenital adrenal hyperplasia (CAH) with low or absent glucocorticoid and mineralocorticoid production. The feasibility of gene therapy for CAH was studied using 21OH-deficient mice (21OH−) and a replication-deficient adenovirus containing the genomic sequence of human CYP21 (hAdCYP21). Intra-adrenal injection of hAdCYP21 in 21OH− mice induced hCYP21 mRNA with the highest expression from 2 to 7 days before a gradual decline. 21OH activity measured in adrenal tissue increased from undetectable to levels found in wild-type mice 2 to 7 days after AdhCYP21 injection. Adrenal morphology of 21OH− mice showed lack of zonation, and hypertrophy and hyperplasia of adrenocortical mitochondria with few tubulovesicular christae. These morphological abnormalities were markedly improved 7 days after hAdCYP21 gene therapy. Plasma corticosterone increased from undetectable levels to values similar in wild-type mice by 7 and 14 days, declining over the next 40 days. This is the first demonstration that a single intra-adrenal injection of an adenoviral vector encoding CYP21 can compensate for the biochemical, endocrine and histological alterations in 21OH-deficient mice, and shows that gene therapy could be a feasible option for treatment of CAH.
Congenital adrenal hyperplasia (CAH) due to 21-hydroxylase (21-OH) deficiency is a relatively common autosomal recessive disease in humans, occurring in one per 14000 live births.1 In this genetic disease, mutations or deletion of the cytochrome P450 21 hydroxylase gene (CYP21) causes glucocorticoid and mineralocorticoid deficiency, leading to an excess of ACTH and androgen production and adrenal hyperplasia. The classical form of the disease which is characterized by severe salt wasting can be fatal if not promptly diagnosed and treated.1,2,3 Current steroid replacement regimens have serious problems with dose adjustment and patient compliance, and development of new therapeutic strategies for treatment of this disease is needed for improving the clinical management of these patients.4,5
A naturally occurring strain of mouse with deletion of the CYP21 and complement 4 component genes has impaired 21-OH activity and glucocorticoid production leading to hyperproduction of ACTH with adrenocortical hyperplasia and accumulation of precursor steroids.6,7 Since the mouse adrenal lacks 17-hydroxylase, unlike the human disease, the accumulated precursor is progesterone and no androgenization occurs in this model. Also, the deletion of the gene for complement component C4 makes the 21-hydroxylase-deficient mouse different from the human disease. However, the presence of severe glucocorticoid and mineralocorticoid deficiency makes this mouse an ideal animal model to explore the possibility of replacing the defective enzyme in the adrenal by gene therapy. Adenoviral vectors have been used to transfer genes to cells in order to correct for genetic deficiencies in animal models and humans.8,9,10,11 While these vectors have several advantages, such as their capacity to infect non-dividing cells, and the possibility of obtaining high virals titers, their use in gene therapy is limited because of severe immune reactions following administration in vivo.12,13 Since steroids are anti-inflammatory,14 we sought the possibility that the high content of corticosteroids in the adrenal gland may prevent the immune reaction to adenovirus injection and make it possible to use these vectors for gene therapy in the adrenal.
To determine the ability of adenoviral vectors to transduce genes in the adrenal we performed preliminary experiments using an adenoviral construct encoding E. coli β-galactosidase under control of the Rous sarcoma virus promoter (AV1.LacZ4). Intra-adrenal injection of AV1.LacZ4 induced β-galactosidase staining in the adrenal cortex and medulla up to 40 days after injection, without any detectable inflammatory reaction. Therefore, we engineered an adenoviral vector encoding human CYP21, and used the 21-OH-deficient mouse to test the feasibility of adenovirus gene therapy for treatment of enzymatic deficiency in the adrenal. The effectiveness of adCYP21 to induce 21-OH activity was first evaluated by analysis of the conversion of 3H-progesterone to 3H-deoxycorticosterone, by primary pituitary cell cultures, which are devoid of endogenous activity. Cells were exposed to 50 plaque-forming units (p.f.u.) per cell adCYP21 or AV1.LacZ4 for 4 h, and maintained from 4 to 30 days before measuring 21-OH activity. Deoxycorticosterone formation was negligible in nontransfected pituitary cells, or in cells transduced with AV1.LacZ4 at any time-point. In contrast, 4, 7 and 14 days after transduction with adCYP21, cell cultures showed marked 21-OH activity with 74.5 ± 4.6, 47.0 ± 5.1 and 26.6 ± 4.1 percent of progesterone being converted to deoxycorticosterone. Activity continued declining with time but it was still present 30 days after transduction.
Based on these in vitro results, we performed intra-adrenal injection of adCYP21 in 21-OH-deficient mice and evaluated 21-OH expression and the biological effects in the adrenal. In situ hybridization showed human CYP21 mRNA in the adrenal cortex and medulla of adCYP21 injected mice but not in AV1.LacZ4 injected control adrenals. Human CYP21 mRNA was expressed in cell clusters comprising about 30% of the adrenal cortex and medulla, with the highest expression occurring 2 to 7 days after injection, and declining but being clearly detectable after 14 days (Figure 1).
In vitro analysis of 21-OH activity in adrenal quarters of 21-OH-deficient mice showed that the conversion of 3H-progesterone to 3H-deoxycorticosterone or 3H-corticosterone was absent in untreated mice or after intra-adrenal injection of AV1.LacZ4. In contrast, adrenals from mice injected with adCYP21 converted progesterone to deoxycorticosterone and to corticosterone by 2 to 7 days after injection (Figure 2b and c) to levels similar to those observed in adrenals of wild-type mice. Forty days after injection, deoxycorticosterone production had decreased but conversion to corticosterone was still at the levels seen in adrenals from wild-type mice, indicating a decrease in 21-OH activity and that all deoxycorticosterone formed was immediately used as substrate for 11-hydroxylase.
The expression and activity of 21-OH in treated mice was sufficient for production and release of corticosterone to the circulation in vivo. HPLC analysis showed low plasma corticosterone levels (4.7± 0.9 ng/ml) in homozygous 21-OH-deficient mice as compared with 44.4 ± 5.6 ng/ml in age-matched controls. After intra-adrenal injection of adCYP21, plasma corticosterone levels did not change significantly after 4 days, but increased to levels similar to those in wild-type mice by 7 and 14 and 30 days (78 ± 30, 86.0 ± 28 and 36.5 ± 1.0, respectively) (Figure 2a).
We have recently shown marked morphological alterations in adrenal of newborn 21-OH-deficient mice, including lack of normal zonation and ultrastructural abnormalities of the mitochondria.15 Adrenals from untreated adult mice in these studies showed similar alterations. Under light microscopy, the adrenal cortex was enlarged, with no clear zonation, a thin capsule, and fasciculata type cells extending from the capsule to the medulla. Numerous chromaffin cells were observed traversing the cortex as demonstrated by tyrosine hydroxylase staining (Figure 3a). Seven days after gene therapy by intra-adrenal injection of adCYP21, the adrenal showed a thick capsule, differentiated zona glomerulosa, and tyrosine hydroxylase staining corresponding to adrenal medullary cells confined to the center of the organ (Figure 3b). At light microscopy, no evidence of inflammatory reaction was observed at any time-point after intra-adrenal injection of AV1.LacZ4 or adCYP21. At the ultrastructural level, adrenocortical cells in untreated mice had increased smooth endoplasmic reticulum, hypertrophied and hyperplastic mitochondria with protrusion of finger-like projections from one mitochondrion into an invagination in a neighboring mitochondrion. Cells located in the subcapsular region displaying fasciculata-type round hyperplastic mitochondria and few tubulovesicular-christae (Figure 3c). Gene therapy induced dramatic improvement in adrenal ultrastructure in these mice; 7 days after ad CYP injection, mitochondrial size became normal, and subcapsular cells showed elongated tubulolamellar christae, typical of the zona glomerulosa. Fasciculata and reticularis cells showed normal tubulovesicular mitochondria and SER (Figure 3d).
This study shows that recombinant adenoviral vectors can be used to correct deficiencies of steroidogenic enzymes by administration directly into the adrenal gland. Adenoviral vectors have the advantage that they can be produced at high titers and can be transduced into nondividing cells.8 However, since these vectors also deliver large amounts of viral sequences they often cause host immune reactions and tissue damage, an affect which can be prevented by immunosuppressants including glucocorticoids.12,13,16 The lack of any detectable inflammation in the adrenal following intra-adrenal injection of the adenoviral constructs suggests that in contrast to other tissues, the high levels of endogenous steroids in this tissue may have a protective effect preventing inflammatory responses. This makes the adrenal an immune privileged organ and an ideal candidate for gene therapy.
Although technically difficult (especially in mice), it is necessary to replace the CYP21 gene or other defective steroidogenic enzymes directly in the adrenal. This will place the transferred gene in close proximity to other enzymes responsible for substrate production and for converting the product (deoxycorticosterone in case of CYP21) to continue the synthesis of the end-point steroids, glucocorticoids and mineralocorticoids. Since the rate-limiting steps of the steroidogenic pathway are the intramitochondrial transport of cholesterol and side chain cleavage enzyme, which are under the control of ACTH, it is unlikely that overexpression of CYP21 would have any adverse effects on plasma glucocorticoid levels. Thus, the increase in circulating corticosterone following restoration of the adrenal capacity to convert progesterone to glucocorticoids by gene therapy, would re-establish the negative feedback of ACTH secretion and the capacity to respond to stress. Because of limited availability of 21-OH-deficient mice, in these experiments it was not possible to measure the responsiveness of the hypothalamic pituitary adrenal axis to stress or circadian variations of plasma glucocorticoids. However, the fact that 21-OH activity in adrenal quarters of treated mice reached the levels of wild-type mice, suggests that the adrenals of these mice have the capacity to respond to stimulation. It should be noted that blood samples for plasma corticosterone measurements were obtained under stress conditions (tail clipping under CO2 sedation), and that the levels were similar in wild-type and treated 21-OH-deficient mice. Although, it is clear that circulating glucocorticoids in 21-OH-deficient mice can reach stress levels after gene therapy, further studies are needed to determine the effects of treatment on the hypothalamic pituitary adrenal axis responsiveness to stress, and sensitivity of the ACTH responses to glucocorticoid feedback. The volume of the plasma samples did not allow measurement of plasma renin activity or aldosterone to evaluate the effects of intra-adrenal injection of adCYP21 on salt and water metabolism. However, the presence of CYP21 mRNA in the subcapsular area as well as the improvement of the adrenocortical structure with differentiation of zona glomerulosa type cells observed after gene therapy suggest that the adrenal has also recovered its capacity to synthesize mineralocorticoids.
In normal conditions, the high intra-adrenal levels of glucocorticoids rather than circulating steroid are involved in adrenal differentiation and regulation of catecholamine biosynthetic enzymes in chromaffin cells.17 Therefore, systemic administration of the gene resulting in little, if any, expression of the enzyme in the adrenal would be unlikely to provide adequate levels of circulating glucocorticoids in case of increased demand, and would be insufficient for local adrenal regulation. The mitochondrial abnormalities in 21-OH-deficient mice resemble those described in humans with CAH,18 and differ from the increase in mitochondrial vesicular membranes that occurs following ACTH or CRH administration.19,20 The remarkable improvement of adrenal structure and function after gene therapy contrasts with the lack of change observed by suppressing excessive ACTH secretion by corticosterone replacement for 7 days (not shown). This suggests that the alterations in adrenal morphology are not solely due to excess ACTH production and supports the view that high intra-adrenal glucocorticoids acting in an autocrine or paracrine manner are necessary for a normal adrenocortical structure and migration of chromaffin cells. It is also possible that accumulation of abnormal steroids contributes to adrenal abnormalities in 21-OH-deficient mice.
It is noteworthy that the striking effects of gene therapy on adrenal structure and function occur with only about 30% of adrenocortical cells expressing CYP21. This suggests that functional recovery of only a proportion of the adrenal cells can lead to production of sufficient glucocorticoids to improve adrenal zonation and differentiation, and that it would not be necessary to express the enzyme in the whole adrenal cortex to obtain therapeutic benefits.
Although the adenoviral vector effectively transduced the missing CYP21 gene into the adrenal and restored normal steroidogenic capacity and anatomic structure of the adrenal, expression of the transferred gene started to decline by 14 days. Typically, the duration of expression with these vectors is limited because of lack of integration of the transferred gene to the cell genome.8 Studies using different viral vectors and stronger promoters to transfer the CYP21 gene into the adrenal of CYP21[−] mice are currently being performed to increase the magnitude and duration of expression.
In summary, we have shown that a single intra-adrenal injection of an adenoviral vector encoding CYP21 induces compensation of biochemical, endocrine and histological alterations in 21-OH-deficient mice. While development of alternative viral vectors can improve the efficiency and duration of gene transfer in the adrenal, the present data show that gene therapy could be a feasible option for treatment of CAH.
Materials and methods
The construction of AVC2.null has been described elsewhere.21 Briefly, the virus consists of a ‘left viral arm’ containing 400 bp of the left terminus of Ad5, the immmediate–early promoter of cytomegalovirus, and a multiple cloning site. The ‘right viral arm’ contains the SV40 early intron and polyadenylation site connecting to viral sequences derived from pJM17, nt 3328 to the right viral terminus.22,23 A 3.1 kb DNA fragment containing the full genomic sequence of the human CYP21-OH was obtained by PCR and cloned into pBluescript SK+ (Stratagene, La Jolla, CA, USA). The DNA fragment extends from 45 bases upstream of the genomic cyp21 transciptional start site and extends 2700 bases downstream, including all nine exons and introns with the exclusion of the last 250 bases of 3′ untranslated region (including the Cyp 21 polyadenylation site). The CYP21 gene was excised from pBluescript by digestion with XbaI and ClaI and ligated directly into AVC2.null using the DNA–protein complex directional ligation technique of Okada et al.21 The DNA–protein complex was introduced into 293 cells by calcium phosphate cotransfection, and viral isolates screened by PCR and selected clones amplified by standard techniques.24 AV1.lacZ4, a first-generation adenoviral vector expressing E. coli β-galactosidase modified by addition of a nuclear localizing signal, has been described previously25 and was kindly provided by Dr Bruce Trapnell (Genetic Therapy, Gaithersburg, MD, USA).
CYP21 transduction in cultured cells
Rat pituitary cells were isolated by trypsin digestion and cultured in 24-well plates as previously described.26 Cells were exposed for 4 h to adCYP21, 50 p.f.u., washed and cultured for 2 to 30 days before measurement of 21-OH activity by their ability to convert 3H-progesterone to 3-deoxycorticosterone. Parallel wells were transduced with 50 p.f.u. of adLac-Z in the same experimental conditions to evaluate gene transfer efficiency by measurement of the number of cells showing β-galactosidase staining.
In vivo procedures
Female 21-OH-deficient mice kindly provided by Dr T Shiroishi, Institute of Genetics, Mishima, Japan) were mated with male C57BL10J mice (Jackson Labs, Bar Harbor, ME, USA). All dams received injections of 5 μg dexamethasone from gestational day 20 until delivery to prevent death at birth. Pups were treated with corticosterone (5 μg per day) fludrocortisone (0.025 μg per day) until day 14, followed by corticosterone in the drinking water until day 21. After this age mice were maintained in standard conditions with regular rodent diet pellets and water ad libitum. Homozygous 21-OH-deficient mice were identified by Southern blot analysis of genomic DNA and the presence of high plasma progesterone levels. Bilateral intra-adrenal microinjections of adCYP21 or ad-LacZ were performed in 2- to 3-month old mice, after surgical exposure of the adrenals via a dorsal approach under ketamine/xylazine anesthesia. Two μl containing 108 p.f.u. of adCYP21were injected into each adrenal at the rate of 0.4 μl/min using a Hamilton syringe and a PE-10 polyethylene tubing attached to a 32-gauge stainless steel needle. Plasma corticosterone levels were measured by high-performance liquid chromatography (HPLC) at days 4, 7, 14 and 30 after injection. Mice were killed by decapitation from 2 to 40 days after injection and the adrenals removed for determination of human CYP21-OH expression, 21-OH activity or histological analysis. All animal protocols were approved by the Animal Care Users Committee, NICHD.
Plasma steroid levels
Blood was collected into non-heparinized capillary tubes after tail clipping under CO2 sedation. Serum corticosterone levels were measured quantitatively by HPLC as previously described.27 Serum progesterone concentrations were determined using commercial kit reagents from Diagnostic Systems Laboratories (Webster, TX, USA).
21-OH expression and activity
The presence of human CYP21-OH was determined by in situ hybridization using 12 μm cryostat sections and a specific 35S-labeled oligonucleotide complementary to base pairs 350 to 386 of exon 2 of the human 21-OH gene, according to previously described procedures.28 21-OH activity was determined by the ability of cell cultures or adrenal quarters to convert 3H-progesterone to 3H-deoxycorticosterone and corticosterone. Tissue was incubated in medium 199 with Earle salts (Biofluids, Gaithersburg, MD, USA) containing 0.1% BSA, 1 μM progesterone and 1 μCi of 3H-progesterone (NEN, Boston, MA, USA), under 95%O2/5% CO2 at 37°C for 2 h. After extraction with CH3Cl, steroids in the supernatant were separated by thin layer chromatography using silica plates and 20% chloroform/80%methanol. Steroids were visualized under UV light, the spot removed into a vial, and radioactivity counted in a liquid scintillation counter. The results are expressed as percent conversion from progesterone to deoxycorticosterone and corticosterone.
Adrenal glands were fixed and embedded in epoxy resin as previously described.15,18 Semithin sections (0.5 μm) were stained with toluidine blue for light microscopy, and ultrathin sections (70 nm) were stained with uranyl acetate and lead citrate for electron microscopy examination at 80 kV in a Phillips EM 301 (Phillips Electronics, Mahway, NJ, USA). For specific staining of chromaffin cells, paraffin sections of mice adrenals were immunostained with anti-tyrosine-hydroxylase antibodies.
New MI . Diagnosis and management of congenital adrenal hyperplasia Annu Rev Med 1998 49: 311–328
Pang S . Congenital adrenal hyperplasia Endocrinol Metab Clin N Am 1997 26: 853–889
Cutler GB Jr, Laue L . Congenital adrenal hyperplasia due to 21-hydroxylase deficiency New Engl J Med 1990 223: 1806–1813
Cutler GB Jr . Treatment of congenital adrenal hyperplasia: the case for new medical approaches J Clin Endocrinol Metab 1996 81: 3185–3186
Migeon CJ . Can the long range results of the treatment of congenital adrenal hyperplasia be improved? J Clin Endocr Metab 1996 81: 3187–3189
Shiroishi T, Sagai T, Natsume-Sakai S, Moriwaki K . Lethal deletion of the complement component C4 and steroid 21-hydroxylase genes in the mouse H-2 class III region caused by meiotic recombination Proc Natl Acad Sci USA 1987 84: 2819–2823
Gotoh H et al. Steroid 21-hydroxylase deficiency in mice Endocrinology 1988 123: 1923–1927
Robbins PD, Ghivizzani SC . Viral vectors for gene therapy Pharmacol Ther 1998 80: 35–47
Vincent N et al. Long term correction of mouse dystrophic degeneration by adenovirus-mediated transfer of a minidistrophin gene Nat Genet 1993 5: 130–134
Jaffe AH et al. Adenovirus-mediated in vivo gene transfer and expression in normal rat liver Nat Genet 1992 1: 372–378
Ye X et al. Prolonged metabolic correction in adult ornithine transcarbamylase-deficient mice with adenoviral vectors J Biol Chem 1996 271: 3639–3646
Ginsberg HS . The ups and downs of adenovirus vectors Bull NY Acad Med 1996 72: 53–58
Paltekian E et al. Adenovirus mediated gene transfer into the brain: methodological assessment J Neurosci Meth 1997 71: 77–84
Munch A, Guyre PM, Holbrook NJ . Physiological functions of glucocorticoids in stress and their relation to pharmacological action Endocrine Rev 1984 5: 25–44
Tajima T, Ma X-M, Bornstein SR, Aguilera G . Prenatal treatment does not prevent alterations of the hypothalamic pituitary adrenal axis in steroid 21-hydroxylase deficient mice Endocrinology 1999 140: 3354–3362
Geddes BJ et al. Persistent transgene expression in the hypothalamus following stereotaxic delivery of a recombinant adenovirus: suppression of the immune response with cyclosporin Endocrinology 1996 137: 5166–5169
Bornstein S et al. Adrenomedullary function is severely impaired in 21-hydroxylase deficient mice FASEB J 1999 13: 1185–1194
Bornstein SR et al. Ultrastructural dynamics of mitochondrial morphology in varying functional forms of human adrenal cortical carcinoma Horm Metab Res 1996 28: 177–182
Nussdorfer GG et al. Investigations on the turnover of adrenocortical mitochondria. XI A stereological study of the effects of chronic treatment with ACTH on the size and number of mitochondria from adult human adrenocortical cells cultured in vitro Cell Tissue Res 1977 182: 145–150
Bornstein SR, Ehrhart-Bornstein M, Guse-Behling H, Scherbaum WA . Structure and dymnamics of adrenal mitochondria following stimulation with corticotropin releasing hormone Anat Record 1997 234: 255–262
Okada T et al. Efficient directional cloning of recombinant adenovirus vectors using DNA–protein complex Nucleic Acids Res 1998 26: 1947–1950
Graham FL, Prevec L . Adenovirus based expression vectors and recombinant vaccines Biotechnology 1992 20: 363–390
McGrory WJ, Bautista DS, Graham FL . A simple technique for the rescue of early region I mutations into infectious human adenovirus type 5 Virology 1988 163: 614–617
Jones N, Shenk T . Isolation of adenovirus type 5 host range deletion mutants defective for transformation of rat embryo cells Cell 1979 17: 683–689
Smith TA et al. Adenovirus mediated expression of therapeutic plasma levels of human factor IX in mice Nat Genet 1993 5: 397–402
Stojilkovic SS, Chang JP, Ngo D, Catt KJ . Evidence for a role of protein kinase C in luteinizing hormone synthesis and secretion: impared responses to gonadotropin releasing hormone in protein kinase C depleted cells J Biol Chem 1988 263: 17307–17311
Fukushi M et al. A method for quantitative analysis of adrenal steroids with high-performance-liquid-chromatography for diagnosis of congenital adrenal hyperplasia. In: Therrel BL (ed) . Advances in Neonatal Screening Elsevier Science Publishers BV: New York 1987 303–308
Aguilera G, Lightman SL, Kiss A . Regulation of the hypothalamic-pituitary-adrenal axis during water deprivation Endocrinology 1993 132: 241–248
About this article
Cite this article
Tajima, T., Okada, T., Ma, X. et al. Restoration of adrenal steroidogenesis by adenovirus-mediated transfer of human cytochromeP450 21-hydroxylase into the adrenal gland of21-hydroxylase-deficient mice. Gene Ther 6, 1898–1903 (1999) doi:10.1038/sj.gt.3301018
- gene therapy
Letter to the Editor: “Congenital Adrenal Hyperplasia Due to Steroid 21-Hydroxylase Deficiency: An Endocrine Society Clinical Practice Guideline”
The Journal of Clinical Endocrinology & Metabolism (2019)
Congenital Adrenal Hyperplasia Due to Steroid 21-Hydroxylase Deficiency: An Endocrine Society* Clinical Practice Guideline
The Journal of Clinical Endocrinology & Metabolism (2018)
Endocrinology and Metabolism Clinics of North America (2017)