| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Medicine, University of Queensland, Greenslopes Private Hospital, Brisbane, Queensland 4120, Australia; Department of Endocrinology, Queen Elizabeth Hospital (S.P.F., P.J.P.), Adelaide, South Australia 5011, Australia; John Curtin School of Medical Research (J.A.W.), Acton, Australian Capital Territory 2601, Australia
Address all correspondence and requests for reprints to: Dr. David Torpy, Department of Medicine, University of Queensland, First Floor Clinical Sciences Building, Greenslopes Private Hospital, Newdegate Street, Brisbane, Queensland 4120, Australia.
Abstract
Corticosteroid-binding globulin is a 383-amino acid glycoprotein
that serves a hormone transport role and may have functions related to
the stress response and inflammation. We describe a 39-member
Italian-Australian family with a novel complete loss of function (null)
mutation of the corticosteroid-binding globulin gene. A second,
previously described, mutation (Lyon) segregated independently in the
same kindred. The novel exon 2 mutation led to a premature termination
codon corresponding to residue -12 of the procorticosteroid-binding
globulin molecule (c.121G
A). Among 32 family members there were 3
null homozygotes, 19 null heterozygotes, 2 compound heterozygotes, 3
Lyon heterozygotes, and 5 individuals without corticosteroid-binding
globulin mutations. Plasma immunoreactive corticosteroid-binding
globulin was undetectable in null homozygotes, and mean
corticosteroid-binding globulin levels were reduced by approximately
50% at 18.7 ± 1.3 µg/ml (reference range, 30–52 µg/ml) in
null heterozygotes. Morning total plasma cortisol levels were less than
1.8 µg/dl in homozygotes and were positively correlated to the plasma
corticosteroid-binding globulin level in heterozygotes. Homozygotes and
heterozygote null mutation subjects had a high prevalence of
hypotension and fatigue. Among 19 adults with the null mutation, the
systolic blood pressure z-score was 12.1 ± 3.5; 11 of 19 subjects
(54%) had a systolic blood pressure below the third percentile. The
mean diastolic blood pressure z-score was 18.1 ± 3.4; 8 of 19
subjects (42%) had a diastolic blood pressure z-score below 10.
Idiopathic chronic fatigue was present in 12 of 14 adult null
heterozygote subjects (86%) and in 2 of 3 null homozygotes. Five cases
met the Centers for Disease Control criteria for chronic fatigue
syndrome. Fatigue questionnaires revealed scores of 25.1 ± 2.5 in
18 adults with the mutation vs. 4.2 ± 1.5 in 23
healthy controls (P < 0.0001). Compound
heterozygosity for both mutations resulted in plasma cortisol levels
comparable to those in null homozygotes. Abnormal
corticosteroid-binding globulin concentrations or binding affinity may
lead to the misdiagnosis of isolated ACTH deficiency. The mechanism of
the association between fatigue and relative hypotension is not
established by these studies. As idiopathic fatigue disorders are
associated with relatively low plasma cortisol, abnormalities of
corticosteroid-binding globulin may be pathogenic.
CORTICOSTEROID-BINDING globulin (CBG) is encoded by a 19-kb gene that is a member of the serine protease inhibitor (SERPIN) family, located at chromosome 14q32.1 (1). The gene encodes a 405-amino acid molecule; a 22-amino acid signal peptide is cleaved before secretion (2). CBG has high cortisol binding affinity (Kd, 3 x 10-7 M), and 90–95% of circulating cortisol is CBG bound at physiological total plasma cortisol levels (3, 4). Apart from a carrier protein role, inhibition of CBG production during stress may exaggerate the cortisol response (5, 6, 7, 8, 9). CBG may play a role in the delivery of cortisol to inflammatory sites (10, 11).
Few abnormalities in human CBG have been described (12, 13, 14, 15, 16, 17, 18). Two inherited functionally important mutations of the CBG gene are reported, each involving single kindreds. The mutant proteins are referred to as transcortin Leuven and CBG Lyon; they arise from single nucleotide/amino acid changes and cause reduced cortisol binding affinity. A lack of substantiated reports of a complete deficiency of CBG has led to the proposition that such a mutation may be lethal in utero (3).
This report describes a 39-member Australian family with the first described complete loss of function (null) mutation of the CBG gene and coinheritance of CBG Lyon. Individuals with the CBG null mutation exhibited a high prevalence of fatigue and hypotension.
Subjects and Methods
Patients
Hypoadrenalism was suspected in the index case because of unexpected low blood pressure (BP) and fatigue. Low plasma cortisol values after ACTH stimulation testing, despite normal basal plasma ACTH levels, were consistent with secondary hypoadrenalism. Paradoxically normal urinary free cortisol levels led to evaluation for CBG deficiency. Adult family members were invited to undergo biochemical, genetic, and clinical evaluations. In children, medical histories were obtained from parents, and a morning blood sample was obtained for biochemical and genetic studies. Family studies were approved by the Greenslopes Private Hospital ethics committee.
A complete medical history, physical examination, blood counts, serum transferrin saturation and thyroid function tests, and plasma electrolytes and creatinine determinations were performed. BP was recorded with an aneuroid sphygmomanometer validated against a mercury column. Measurements were taken seated, from the left arm, then repeated 5 min later. The mean of the two BP readings was used for analysis. To allow for age and gender effects on BP, measurements were related to normative data from the Australian National Nutrition Survey 1995 of 16,400 random subjects and expressed as a z-score (19). Body weight and height were measured, and body mass index (BMI; weight in kilograms/height in meters squared) was calculated.
There were frequent spontaneous complaints of fatigue from family
members. Idiopathic chronic fatigue and chronic fatigue syndrome were
diagnosed with Centers for Disease Control and Prevention criteria
(20). Idiopathic chronic fatigue (ICF) was defined as
fatigue of greater than 6 months duration with no apparent medical
explanation, or abnormality of thyroid function, transferrin saturation
and/or routine electrolytes, renal function, and urinalysis. In
addition, adult family members were asked to complete the Fibromyalgia
Impact Questionnaire, aimed at evaluating fatigue, that had been
modified to delete questions directed at pain except for a single item
(21, 22). Two family members with major disabling
concomitant illnesses were not assessed with the fatigue questionnaire
due to the likely effects of their associated illnesses. The excluded
subjects included the proband, a homozygote with severe cardiac disease
(no. 22; Fig. 1
), and a subject who does
not carry the mutation, but has traumatic quadriplegia, onset at age 15
yr. Fatigue scores were compared with those obtained in 20 volunteers
(age range, 22–55 yr; mean, 37.8 ± 2.2; 10 males and 10 females)
without major debilitating illness. These subjects were of similar mean
age to the adults of generations 2–3 from the kindred (age range,
22–57; mean, 36.7 ± 2.5; 8 males and 11 females), which
comprised the kindred members being studied for fatigue. Confounding
depression was excluded by clinical interview and the Beck Depression
Inventory (23). Some subjects could not be studied,
including one who died before evaluation (no. 2; Fig. 1) and her spouse
who was unavailable (no. 1).
|
Venous blood was collected in EDTA-containing tubes between
0800–0900 h. Plasma was aliquoted for measurement of total and free
cortisol and CBG. In addition, total and free cortisol were measured at
1 min before and 30 and 60 min after ACTH-(1–24) (250 µg) was given
in the midmorning to the proband (Fig. 1
, no. 22). Free cortisol was
separated from the bound fraction by a method validated in our
laboratory. One milliliter of plasma was added to each
ultracentrifugation column (Centrifree catalogue no. 4104, Amicon,
Danvers, MA; YM membranes, 30,000 mol wt cut-off). After equilibration,
ultrafiltration chambers are centrifuged at 2,000 x g
for 10 min. The ultrafiltrate and the noncentrifuged plasma were
assayed for free and total cortisol, respectively, using a commercial
RIA (Amerlex IM2021B, catalogue no. 8758401; Amersham, Castle
Hill, NSW, Australia), modified to increase assay sensitivity
(24). Inter- and intraassay coefficients of variation for
the cortisol assay using plasma or ultrafiltrate are below 5%. The
free cortisol fraction is the plasma free cortisol divided by the total
plasma cortisol concentration (reference range, <5%).
Plasma CBG assay
Plasma CBG was measured using a commercial RIA (Biosource Technologies, Inc., Nivelles, Belgium). The reference range (41 ± 5.5 µg/ml; mean ± 2 SD) was validated in our laboratory. The detection limit for this assay is 0.25 ± 0.02 µg/ml (mean ± SD), and intra- and interassay coefficients of variation were less than 5%.
Plasma 
1-antitrypsin levels and phenotype
Plasma 1-antitrypsin levels were
measured using a commercial immunoassay. Phenotypes were established by
isoelectric focusing.
Molecular studies
Genomic DNA from 32 family members was extracted, using a commercial method (Nucleon, BACC2, Amersham Pharmacia Biotech, Castle Hill, Australia), from peripheral blood leukocytes isolated by Ficoll-Histopaque centrifugation. The four CBG gene-coding exons (no. 2–5) were amplified using previously published intron-specific primers (14). PCR was performed in a 100-µl reaction containing sterile water, PCR buffer (10x), Q-solution (5x), 2.5 U Taq DNA polymerase (QIAGEN, Clifton Hill, Australia), 100 pmol each of forward and reverse primers, 25 nmol each of deoxy-NTPs (Promega Corp., Annandale, Australia), and 1 µg genomic DNA. The reaction mixture was amplified for 30 cycles in a thermal cycler (Perkin-Elmer Corp., Norwalk, CT). Cycles for exons 2, 3, and 4 were 95 C for 2 min, followed by 30 cycles of denaturation at 94 C for 1 min, annealing at 57 C for 2 min, and extension at 72 C for 3 min and a final 10-min extension cycle at 72 C. Amplification of exon 5 required an annealing temperature of 54 C. The size of the PCR products that contained the exons and some intronic sequence were: exon 2, 750 bp; exon 3, 400 bp; exon 4, 300 bp; and exon 5, 280 bp. The PCR products were purified using the QIAquick PCR purification kit (QIAGEN). Exons 2 and 3 were purified directly from PCR products, but exons 4 and 5 were purified from agarose gel slices to separate the product from nonspecific bands.
Before sequencing we examined segregation of the CBG deficiency trait and severity with several microsatellite markers close to the CBG allele, including a (CA)n repeat near the 3'-end of the CBG gene (25). However, the microsatellites tested were not sufficiently polymorphic in this family (data not shown).
Automated direct sequencing
PCR-amplified genomic DNA fragments from seven family members
were sequenced (Fig. 1; pedigree nos. 1, 2, 8, 16, 18, 22, and 25).
Purified DNA was sequenced using an ABI 366 DNA sequencer (PE Applied Biosystems, Foster City, CA) and a dideoxy cycle
sequencing protocol in the presence of specific primers and
fluorescent-labeled dideoxy terminators (26). Exon
sequences were compared with published cDNA sequences. Exon sizes were:
exon 2, 615 bp; exon 3, 267 bp; exon 4, 150 bp; and exon 5, 183 bp.
Restriction enzyme analyses
CBG null mutation. To facilitate rapid detection of the null
mutation, the restriction enzyme BfaI
(5'-C
TAG-3' restriction site; New England Biolabs, Inc., Beverly, MA) was used to digest PCR-amplified
products of exon 2 in all 32 genotyped family members. The null
mutation introduces a second exon 2 restriction site.
CBG Lyon. Sequencing of exon 5 detected the reported CBG
Lyon mutation in one subject (no. 1). The Lyon mutation was confirmed
in this subject as well as his descendants (pedigrees 29, 31, 33, and
34) using the restriction enzyme TaqI (New England Biolabs, Inc.). The TaqI enzyme has the restriction
recognition site 5'-TCGA-3'. TaqI digestion
results in two bands (180 and 100 bp) after 2% agarose gel
electrophoresis. Three bands (280, 180, and 100 bp) were seen in
subjects heterozygous for the Lyon mutation. This study allowed
detection of CBG Lyon in all family members who may have inherited this
allele, except for two children (nos. 36 and 37; Fig. 1).
Statistical analyses
Pearson product-moment correlations were used to evaluate the relations described. Fatigue scores in CBG mutation subjects and controls were compared with independent t tests. Descriptive statistics and all analyses were performed using the Statistica computer application (Statsoft, Tulsa, OK). Results are expressed as the mean ± SE.
Results
Patients
The index case, a 48-yr-old woman, was referred for possible hypoadrenalism. She had been diagnosed in her 30s with a ventricular septal defect and consequent Eisenmenger’s complex. In 1996 low BP (100/70 mm Hg) and chronic fatigue were noted. After iv synthetic ACTH-(1–24) (250 µg) administration, plasma cortisol levels were 1.5, 7.3, and 9.5 µg/dl at -1, 30, and 60 min, respectively. A basal ACTH level was 15 ng/liter (reference range, 10–60). Free T4 and TSH levels were normal, and the patient had a normal menstrual pattern. Pituitary magnetic resonance imaging was normal. Isolated ACTH deficiency was suspected, and dexamethasone (0.25 mg daily) was prescribed, but dexamethasone either was not taken or was discontinued within 1 month. In 1999, a random 0900 h plasma cortisol measurement was 2.8 µg/dl, and ACTH was 33 ng/liter (reference range, 10–60). A 24-h urinary free cortisol measurement was 96 µg/24 h (reference range, 35–120), with concomitant creatinine 10 nmol/24 h (reference range, 7–13) and volume of 1.61 liters. A repeat ACTH stimulation test showed total/free cortisol levels at -1, 30, and 60 min of 19/0.3, 2.7/1.4, and 4.1/1.7 µg/dl, respectively. These values correspond to free cortisol fractions of 14.3% (-1 min), 51.5% (30 min), and 42.6% (60 min).
The 39-member 4-generation kindred is shown in Fig. 1. The proband (no.
22) died several months after the above investigations as a result of
pulmonary hemorrhage. Another subject (no. 2) died of metastatic
gastric adenocarcinoma in her 80s. The early generations of this family
are Italian immigrants to Australia.
Plasma CBG RIA
The plasma immunoreactive CBG concentration was measured in 32
family members using plasma taken between 0800–0900 h. The results are
shown in Fig. 2. Plasma CBG was
undetectable in the 3 homozygotes for the null mutation (no. 8, 16, and
22). CBG levels in 19 null heterozygotes were 20.3 ± 2.0 (range,
9.0–49.9 µg/ml). All null heterozygotes had subnormal CBG levels
(<30 µg/ml), except one subject who was taking an oral contraceptive
preparation (49.9 µg/ml). The CBG level in this woman was subnormal
when related to the reference range for estrogen-treated women (67–116
µg/ml) (27). The 3 null homozygotes had undetectable CBG
levels. Compound heterozygotes, who have both the null and Lyon
mutations, had low CBG levels compared with null heterozygotes (8.7 and
11.9 µg/ml). The 3 Lyon heterozygotes had normal CBG levels (32.1,
42.4, and 46.4 µg/ml).
|
Plasma cortisol results are shown in Fig. 3. Three null homozygotes had very low
total plasma cortisol levels (no. 8, 0.6 µg/dl; no. 16, 1.3 µg/dl;
no. 22, 1.9 µg/dl). Corresponding free cortisol levels were 0.1, 0.3,
and 0.5 µg/dl, respectively. Hence, total cortisol levels would often
be undetectable in routine clinical assays. The free cortisol levels,
however, were comparable to those in normal subjects. Free cortisol
fractions were 25%, 22.9%, and 28.3%, respectively. These values are
much higher than those in normal subjects (range, <5%).
|
Plasma 1-antitrypsin levels
Plasma 1-antitrypsin levels were normal,
and the phenotypes were of common types in seven individuals, including
three null homozygotes and four null heterozygotes (no. 1, 1.1 g/liter,
phenotype M1 M2; no. 2, 1.8 g/liter, phenotype M1 M2; no. 8, 1.1
g/liter, phenotype M2; no. 16, phenotype M2; no. 22, 1.7 g/liter,
phenotype M2; no. 24, 1.1 g/liter, phenotype M2 M3; no. 25, 1.3
g/liter, phenotype M1 M2).
Sequencing of the CBG gene
A novel point mutation (c.121GA) in exon 2 was found in some
individuals, which would lead to a premature termination codon
(Trp-12x) at residue -12 of the pro-CBG molecule. The sequence trace
is shown in Fig. 4. Sequencing of exon 5
detected the CBG Lyon mutation in one subject (no. 1). The Lyon is a
point mutation (c.1254GA) at the codon for residue 367
(Asp367Asn). The Lyon mutation was confirmed in
subject 1 and four relatives (no. 29, 31, 33, and 34) by restriction
enzyme analysis. Inheritance of the CBG Lyon allele confirms that the
mutation involves the other CBG allele to the null mutation,
i.e. the alternative chromosome 14.
|
CBG null mutation detection using the restriction enzyme,
BfaI, was performed in the 32 genotyped family members (Fig. 1). The patterns in normal subjects, homozygotes, and heterozygotes are
shown in Fig. 5. The Lyon mutation was
detected by sequencing in one subject (no. 1), and subjects who may
have inherited the Lyon mutation were screened by TaqI
restriction enzyme analysis. This led to the detection of four further
Lyon mutation subjects (pedigrees 29, 31, 33, and 34; Fig. 1).
|
Systolic and diastolic BPs were significantly lower in CBG null
mutation subjects than expected based on population norms
(19). Systolic and diastolic BPs, expressed as z-scores
are shown in Fig. 6. Among null
heterozygotes, the mean systolic BP z-score was 10.2 ± 4.0; the
diastolic BP z-score was 17.1 ± 4.0 (n = 15). The null
homozygote mean systolic BP z-score was 17.0 ± 8.7; the diastolic
BP z-score was 15.3 ± 4.8. Only three adult subjects without any
CBG mutation were available for BP measurement; systolic BP z-scores
varied from 5–51, and diastolic BP z-scores ranged from 19–88. Only a
single Lyon heterozygote was available for BP measurement; the z-scores
were 70–80. A compound heterozygote had systolic/diastolic BP z-scores
of 26/40, respectively.
|
Fatigue
Persistent unexplained fatigue was a frequent complaint among family members with the CBG null mutation. ICF was present in 12 of 14 adult null heterozygote subjects (86%). Of the 3 null homozygote subjects ICF was present in 2 cases, but the proband had known hypoxia and hence could not be classified as having idiopathic fatigue. The third null homozygote subject had widespread pain and relatively mild fatigue, but had fibromyalgia. The 3 adult subjects without any CBG mutation did not report significant fatigue. The single available compound heterozygote had marked ICF. The Lyon heterozygote adult did not have ICF. The timing of the maximum fatigue symptoms was skewed toward the early morning in 10 of 12 null heterozygotes, 2 of 2 null homozygotes, and the only compound heterozygote subject.
Chronic fatigue syndrome (CFS) was diagnosed in five null heterozygote subjects on the basis of CDC criteria (20). These subjects met the criteria because of the presence of disturbed concentration or memory and pain, especially joint pains and headaches. No subjects exhibited the flu-like symptoms (sore throat, tender lymph nodes) that occur in some CFS patients.
The results of questionnaire-based fatigue quantitation are shown in
Fig. 7. The mean fatigue score was
elevated in 18 adults bearing CBG mutations relative to the scores of
age-matched control subjects (25.1 ± 2.6 vs. 4.2
± 1.5; P < 0.0001). The fatigue scores of the 2
surviving homozygotes were not noticeably greater than those of the
heterozygotes (no. 6, 10.3; no. 16, 16.5). Severe fatigue was present
in the proband, but her data were excluded from analysis because of the
presence of another disorder that could explain her fatigue (heart
disease with hypoxia). Individuals with the Lyon mutation alone or the
compound heterozygote appeared to have significant fatigue. Among
children, fatigue questionnaires were not applied, but parents reported
markedly lower activity levels in subject 32 compared with his sisters
(no. 31 and 33) before knowledge of their mutation status.
|
There was no significant correlation between plasma CBG level and fatigue score. No significant correlation was noted between systolic or diastolic BP and fatigue score among those individuals with the mutation.
Other clinical features
Obesity was evident among the null homozygotes relative to null heterozygotes and unaffected subjects. The mean BMI of the null homozygotes was 32.7 ± 3.4 vs. 25.0 ± 1.4 in null heterozygotes (P = 0.038). Two of the null homozygote subjects were obese (no. 8 BMI, 30.8; no. 16 BMI, 39.4), and a third homozygote was overweight (no. 22 BMI, 28.0). The overweight null homozygote was ill with hypoxia and had lost some weight with anorexia. A lower proportion of null heterozygote subjects were obese (4 of 14).
Seven of the 16 examined adult subjects with the mutation have cardiac disease including ventricular septal defect/Eisenmenger’s complex, mitral valve replacement for regurgitation, recurrent symptomatic supraventricular tachycardia requiring catheter ablation of an aberrant electrical pathway, mitral valve disease attributed to childhood rheumatic fever, and 3 cases of mitral valve prolapse (systolic click). Thyroid enlargement, focal or diffuse, was noted in 8 of 19 subjects. One patient had been subjected to thyroidectomy as a child, and another had a fine needle biopsy for a benign 2.5 cm lesion. Some asymptomatic goiters and an undiagnosed 3- to 4-cm thyroid nodule were noted.
Pain syndromes were observed in six adult subjects; four of these were null heterozygotes, and two were homozygotes. These individuals reported single region or multiregion chronic unexplained pain, such as migraines, nonspecific headaches, lumbar back pain, cervical neck pain, pelvic pain, or a picture consistent of typical fibromyalgia. One pain-affected subject had CFS.
One null heterozygote subject had biopsy-proven unexplained hepatic steatosis, and a null homozygote had hepatic steatosis diagnosed on the basis of liver function tests, negative screens for viral hepatitis, and ultrasound. Finally, a single homozygote subject has been treated with warfarin postmitral valve replacement and has exhibited warfarin resistance. International normalized ratios were 0.9 while taking warfarin (30 mg daily), and at this level the patient sustained a transient ischemic attack. Baseline coagulation parameters (international normalized ratio and activated partial thromboplastin time) were normal in a surviving homozygote.
Discussion
This report describes a novel complete loss of function mutation
of the human CBG gene. The mutation is a single base substitution
(c.121GA) leading to a premature termination codon at residue -12
(replacing tryptophan) of the pro-CBG molecule (Fig. 8). The kindred includes 3 homozygotes,
21 heterozygotes, and 8 individuals without the null mutation. Null
homozygosity leads to undetectable CBG levels with very low total
plasma cortisol concentrations. Null heterozygosity leads to 50%
reduced CBG levels and low total plasma cortisol. The correlation
between the number of mutant alleles and the CBG level excludes the
possibility that a CBG pseudogene was sequenced.
|
Heterozygosity and homozygosity for the null allele results in relatively low blood pressure. Sixty-eight percent of null mutation adults had systolic BP z-scores below 3, and 42% had diastolic BP z-scores below 10 compared with age/gender-related population norms. ICF, based on Centers for Disease Control criteria, was present in 82% of those bearing the CBG null mutation. A minority of these also met the additional required criteria for CFS. Fatigue could not be explained by abnormalities of thyroid function, iron deficiency, or other disorder. The presence of these clinical features in both null heterozygotes and homozygotes suggests that these effects are dominantly inherited.
Surprisingly, five subjects were heterozygous for the CBG Lyon allele.
The Lyon mutation (c.1254GA; Asp367Asn) has
only been described in one previous publication describing five related
individuals (15). Two subjects were compound heterozygotes
for the two mutations, three subjects were heterozygous only for the
Lyon mutation. This mutation is associated with a 4-fold reduction in
CBG-cortisol binding.
Compound heterozygotes (null/Lyon) exhibited very low total plasma cortisol concentrations. Lyon heterozygotes had morning cortisol levels similar to levels in unaffected subjects. Only one compound heterozygote was available for fatigue and BP studies. Marked ICF was noted in this subject, and BP z-scores were approximately 30. A single Lyon heterozygote adult was noted to have significant fatigue but normal blood pressure (z-score, 70–80). Among three siblings marked inactivity was noted in a null heterozygote, but not in his two Lyon heterozygote sisters. The smaller number of occurrences of the Lyon mutation does not allow definitive conclusions about its clinical effect. Nevertheless, the effects of the Lyon mutation appear less pronounced than that of the more severe null mutation.
These observations accord with a previous report, in which only one subject homozygous for the Lyon mutation was noted to have fatigue and hypotension, not four heterozygote daughters (15). The first described Lyon family was of northwest African origin (15). The family reported here is from Italy, suggesting that the Lyon mutation may be widespread.
CFS is a syndromal diagnosis applied to a subset of patients with ICF. There is substantial overlap with fibromyalgia, a disorder characterized by chronic pain (>3 months) and greater than 11 of 18 tender points. ICF is defined as unexplained fatigue of greater than 6-month duration (20). In addition to chronic fatigue, CFS patients exhibit 4 or more of a defined group of symptoms, which include impaired memory or concentration, sore throat and tender lymph nodes, pains such as headaches and muscle or joint pain, and unrefreshing sleep (20). CFS has a prevalence of 0.2%, whereas ICF affects about 2% of the U.S. adult population (28). The clinical course of CFS is varied, and some features that are prominent early in the disease, such as sore throat and muscle pains, tend to diminish with time. ICF and CFS are probably manifestations of diverse underlying processes. Patients with idiopathic fatigue syndromes as a group tend to have lower plasma cortisol levels than healthy controls (29). Abnormal autonomic regulation of blood pressure with hypotension has also been noted (30).
There were six subjects with the null mutation who had chronic pain, one of whom met the criteria for fibromyalgia syndrome. This is interesting in view of the reports of reduced neuroendocrine function in fibromyalgia, an idiopathic pain syndrome, and the association of pain with CFS (31).
Recent studies have shown that a proportion of patients with CFS respond with complete resolution of symptoms to low dose (5 or 10 mg) hydrocortisone (32). It is possible that CBG mutations may contribute to these disorders, and their detection may allow glucocorticoid therapy to target those individuals most likely to benefit. A study of CBG gene mutations and polymorphisms in patient groups in which their frequency may be enriched may be warranted.
A convincing correlation between BP z-score and fatigue was not demonstrated, and it is possible that each feature relates to a function of CBG beyond plasma transport of cortisol. In this regard animal studies have shown tissue-specific expression of CBG in a range of tissues apart from hepatocytes, and a developmental role for CBG has been suspected based on gestation date-dependent expression of CBG in the fetal rat (33, 34, 35, 36).
In this kindred and two others (15, 16), low or inactive CBG seems to result in low BP and fatigue. These features are seen in glucocorticoid deficiency. However, CBG-bound cortisol is not thought to contribute to circulating glucocorticoid activity. The free hormone hypothesis suggests that it is the free, or unbound, hormone that contributes to circulating hormone activity as the free fraction can readily cross cell membranes and interact with the GR.
The finding of normal ACTH and free cortisol levels suggests normal circulating glucocorticoid activity. It is of interest, however, that CBG expression has been detected in the guinea pig corticotrope, but not other pituicytes (34). If CBG is expressed in the human corticotrope and attenuates glucocorticoid feedback, a lack of corticotrope CBG in these patients may lead to increased glucocorticoid sensitivity at the GR and hence normal ACTH levels despite reduced circulating glucocorticoid activity. Alternatively, the clinical features may relate to a physiological role for CBG, independent of cortisol (37).
The null homozygote subjects were significantly more obese than heterozygotes. Two of three null homozygotes were obese (BMI, >30), and a third overweight subject lost significant weight due to intercurrent illness at the time of analysis. As CBG is expressed in rat adipocytes (38), lack of CBG may accentuate the known effects of cortisol on adipocytes by increasing the availability of cortisol at the GR.
The SERPIN (serine protease inhibitor) gene cluster is located at
14q32.1 and comprises centromere-telomere, CBG, a highly
homologous pseudogene, 1-antitrypsin,
kallistatin, protein C inhibitor, and
1-chymotrypsin (39). The CBG gene
has been thought to arise from an ancient gene duplication event
involving the 1-antitrypsin gene
(40). The reported CBG mutation appears to have arisen
since this event, as 1antitrypsin levels and
phenotype were normal.
The SERPIN structure of CBG has been proposed to relate to a role for CBG in inflammation, where neutrophil elastase cleaves CBG, releasing cortisol at inflammatory sites (10, 11). However, the lack of any infection or inflammatory phenotype in this large kindred suggests that this proposed antiinflammatory role for CBG may not be critical, or there are compensatory changes in other inflammatory mechanisms in individuals with an inherited loss of function CBG mutation.
CBG has significant binding affinity for other steroids, such as the progesterone and adrenal biosynthetic intermediaries, 11-deoxycortisol, 17-hydroxyprogesterone, and deoxycorticosterone (41). With respect to progesterone, menstrual irregularity was not noted in reproductive age women in the kindred.
Recently, stress-induced, probably cytokine (including IL-6)-mediated, reductions in CBG levels have been demonstrated (42, 43). Decreases in CBG of approximately 50% have been noted in sepsis, burns, and cardiac surgery (5, 6, 7, 8, 9). Individuals with only a single functioning CBG allele may be more sensitive to cytokine-mediated inhibition of CBG synthesis, extending the stress response. These could lead to exhaustion of the stress system (hypothalamic-pituitaryadrenal axis and sympathetic nervous system), leading to a syndrome similar to that seen in chronically stressed animal models.
Several clinical conditions were seen that may have arisen by non-CBG mechanisms. These include cardiac valvular and rhythm disturbances, thyroid nodules or goiter, and idiopathic pain conditions. With respect to the thyroid, rat thyroid follicle cells contain CBG on immunocytochemistry (4). A single subject exhibited warfarin resistance, and there may be a role for CBG in the coagulation pathway, similar to the highly homologous SERPIN molecule, protein C inhibitor. A null heterozygote individual died of gastric adenocarcinoma, as did a patient from the CBG Lyon kindred, although this association may be coincidental.
Some limitations of these studies arise from the number of observations of clinical manifestations of certain genotypes possible within a single family. The kindred has three null homozygotes. Hence, an apparent increase in obesity in these subjects may have arisen by chance. On the other hand, comparisons between heterozygote and homozygote subjects may have underestimated the effect of the mutation on phenotype when the heterozygotes are affected. As there were only three adults without CBG mutations, we lacked an intrafamily mutation-free control population and had to use population controls for blood pressure and fatigue levels.
The assessment of possible ACTH deficiency with concomitant hypocortisolism is often difficult, and the biochemical strategy that best predicts a need for chronic glucocorticoid replacement is controversial. ACTH deficiency is often a late manifestation of a structural pituitary lesion (tumor, surgery, or radiation), but may occur alone as isolated ACTH deficiency. Low plasma cortisol levels in the proband with normal ACTH led to an initial diagnosis of isolated ACTH deficiency. Peak plasma cortisol after ACTH stimulation is often recommended as the criterion for diagnosis of hypocortisolism. In the proband the marked increment in cortisol was not typical of ACTH deficiency. Hence, individuals with deficient CBG or impaired CBG-cortisol binding may be misdiagnosed as isolated ACTH deficiency. The likelihood of this is unclear, as the frequency of abnormal CBG binding or concentration is unknown.
In conclusion, we report a novel loss of function mutation of the CBG gene in association with low blood pressure and fatigue. Confirmation of this association in other kindreds, the mechanism of these findings, and their applicability to related idiopathic syndromes require further study.
Acknowledgments
We thank the Department of Veterans’ Affairs and Ramsay Health Care for their ongoing and generous support. We also thank Prof. Colin Chesterman (SEALS Pathology, Sydney, Australia), Dr. David Kanowski (Sullivan & Nicolaides Pathology, Brisbane, Australia), and Dr. George Phillipov (Queen Elizabeth Hospital) for the routine laboratory assays. We are grateful to Dr. Geoffrey Hammond (London Regional Cancer Center, London, Canada) for helpful discussions.
Footnotes
This work was supported by the Sylvia and Charles Viertel Charitable Foundation, a University of Queensland Early Career Grant, and the Ramaciotti Foundation (to D.T.). Presented in part at the International Congress of Endocrinology, Oct. 29–Nov. 2, Sydney, Australia.
Abbreviations: BMI, body mass index; BP, blood pressure; CBG, corticosteroid-binding globulin; CFS, chronic fatigue syndrome; ICF, idiopathic chronic fatigue; SERPIN, serine protease inhibitor.
Received October 17, 2000.
Accepted April 25, 2001.
References
transactivation cascade regulates gene activity and
chromatin structure of the human serine protease inhibitor gene cluster
at 14q32.1. Proc Natl Acad Sci USA 96:10308–10313This article has been cited by other articles:
 |
|
 |
|
  K. Wingenfeld, C. Heim, I. Schmidt, D. Wagner, G. Meinlschmidt, and D. H. Hellhammer HPA Axis Reactivity and Lymphocyte Glucocorticoid Sensitivity in Fibromyalgia Syndrome and Chronic Pelvic Pain Psychosom Med, January 1, 2008; 70(1): 65 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Crawley and G. Davey Smith Is chronic fatigue syndrome (CFS/ME) heritable in children, and if so, why does it matter? Arch. Dis. Child., December 1, 2007; 92(12): 1058 - 1061. [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Cooper and P. M. Stewart Adrenal Insufficiency in Critical Illness J Intensive Care Med, November 1, 2007; 22(6): 348 - 362. [Abstract] [PDF]
|
|
|
|
 |
|
 D. Qi and B. Rodrigues Glucocorticoids produce whole body insulin resistance with changes in cardiac metabolism Am J Physiol Endocrinol Metab, March 1, 2007; 292(3): E654 - E667. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H. Petersen, T. K. Andreassen, T. Breiderhoff, J. H. Brasen, H. Schulz, V. Gross, H.-J. Grone, A. Nykjaer, and T. E. Willnow Hyporesponsiveness to glucocorticoids in mice genetically deficient for the corticosteroid binding globulin. Mol. Cell. Biol., October 1, 2006; 26(19): 7236 - 7245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Guyonnet-Duperat, N. Geverink, G. S. Plastow, G. Evans, O. Ousova, C. Croisetiere, A. Foury, E. Richard, P. Mormede, and M.-P. Moisan Functional Implication of an Arg307Gly Substitution in Corticosteroid-Binding Globulin, a Candidate Gene for a Quantitative Trait Locus Associated With Cortisol Variability and Obesity in Pig Genetics, August 1, 2006; 173(4): 2143 - 2149. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Ousova, V. Guyonnet-Duperat, N. Iannuccelli, J.-P. Bidanel, D. Milan, C. Genet, B. Llamas, M. Yerle, J. Gellin, P. Chardon, et al. Corticosteroid Binding Globulin: A New Target for Cortisol-Driven Obesity Mol. Endocrinol., July 1, 2004; 18(7): 1687 - 1696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Joyner, L. J. Hutley, A. W. Bachmann, D. J. Torpy, and J. B. Prins Greater replication and differentiation of preadipocytes in inherited corticosteroid-binding globulin deficiency Am J Physiol Endocrinol Metab, May 1, 2003; 284(5): E1049 - E1054. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Cleare The Neuroendocrinology of Chronic Fatigue Syndrome Endocr. Rev., April 1, 2003; 24(2): 236 - 252. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Afari and D. Buchwald Chronic Fatigue Syndrome: A Review Am J Psychiatry, February 1, 2003; 160(2): 221 - 236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-M. Fernandez-Real, M. Pugeat, M. Grasa, M. Broch, J. Vendrell, J. Brun, and W. Ricart Serum Corticosteroid-Binding Globulin Concentration and Insulin Resistance Syndrome: A Population Study J. Clin. Endocrinol. Metab., October 1, 2002; 87(10): 4686 - 4690. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
), 3 null homozygotes (
), 5 unaffected
related individuals (
), 3 Lyon heterozygotes, and 2 heterozygotes
for both null and Lyon mutations had IR-CBG measured in plasma from an
0800–0900 h morning blood sample. The reference range mean ± 2
SD for IR-CBG concentrations is 30–52 µg/ml, as shown by
the dotted lines. Null homozygotes had undetectable CBG
concentrations (<0.25 µg/ml). Null heterozygotes had variable levels
(mean, 18.7 ± 1.3 µg/ml, or 
60% of the lower limit of the
reference range), excluding the 1 individual (*) with the high normal
CBG level who was taking an oral contraceptive preparation.
