Psychoneuroendrocrinology, Vol. 22 (Supplement 1) pp.

S109-S 113, 1997

(~) 1997 Elsevier Science Ltd. All rights reserved.
Printed in Great Britain
PII: S0306-4530(97)00014-0
0306-4530/97 $17.(XI + 0.00

Pergamon

P A T H O G E N E T I C R O L E , IN H U M A N A N D M U R I N E
TUBERCULOSIS, OF CHANGES IN THE PERIPHERAL
METABOLISM OF GLUCOCORTICOIDS AND
ANTIGLUCOCORTICOIDS
G. A. W. Rook t., and R. Hernandez-Pando ~
I Department of Bacteriology, UCL Medical School, Windeyer Building, 46 Cleveland St..
London W I P 6DB, UK
Department of Pathology, lnstituto Nacional de la Nutricion, Salvador Zubiran, Calle
Vasco de Quiroga 15, Delegacion Tlalpan, 14000 MEXICO DE Mexico

SUMMARY
Immunity to tuberculosis (TB) requires a Thl pattern of cytokine release, dominated by interleukin-2
(IL-2) and interferon gamma (IFNF). In experimental models even a minor Th2 component (characterized
by production of IL-4) abrogates immunity, and leads to an immunopathology that mimics the human
disease. Increased exposure of T cells to glucocorticoids drives them towards a Th2 cytokine profile and
could therefore help to explain the presence of an inappropriate Th2 component in TB.
Analysis of adrenal steroid metabolites in 24-h urine collections revealed a striking increase in metabolites
of active cortisol relative to metabolites of inactive cortisone. This indicates a change in the balance
of I lfl-hydroxysteroid dehydrogenase to l l/Lketosteroid reductase. The site of this disease-associated
alteration in reductase/dehydrogenase balance may be the lung. The lung contains IlflHSD-1, (a reversible
oxido-reductase) which in the liver works as a reductase. In the normal lung it functions paradoxically as
a reductase, but it can alter its function in the presence of cytokines.
TB patients (like other ill individuals) also show reduced 24-h urinary secretion of dehydroepiandrosterone
(DHEA) derivatives. Since these have antiglucocorticoid functions in vivo, this fall may exacerbate the
effects of the reduced inactivation of cortisol and loss of diurnal rhythm, and contribute to immunological
dysfunction. Recent studies of TB in mice, and in children during infancy, adrenarcbe and puberty, suggest
that the ratio of cortisol to DHEA may be crucial both to susceptibility and to the pathology of the disease
that develops. @ 1997 Elsevier Science Ltd

Keywords--Tuberculosis;Glucocorticoid; Antiglucocorticoid; 16tx-hydroxylation; l l/~-hydroxysteroid

dehydrogenase; Adrenarcbe

INTRODUCTION
Glucocorticoids and tuberculosis

The immune response mechanisms that control M. tuberculosis are sensitive to suppression by
glucocorticoids (McCune et al., 1966). This can be seen after exposure of tuberculous mice to
stress, and the effects are at least partly attributable to direct inactivation of anti-mycobacterial
* Address correspondence to: Dr G. A. W. Rook, Department of Bacteriology, UCL Medical School, Windeyer Building,
46 Cleveland St., London W l P 6DB, UK (Tel: 44 171 380 9489; Fax: 44 636 8175; E-mail: g.rook@ucl.ac.uk).
S109

Sll0

G . A . W . Rook & R. Hernandez-Pando

mechanisms within macrophages (Brown et al., 1995; Rook et al., 1987). Glucocorticoids also
tend to bias the T lymphocyte response towards a Th2 cytokine profile (IL-4, IL-5) (Brinkmann
& Kristofic, 1995; Daynes et al., 1995; Mason, 1991).
Immunity to mycobacteria requires an exclusively Thl (IFN and IL-2) pattern of cytokine
release (Rook & Bloom, 1994). Not only is the Thl pattern required for immunity, but even a
minor Th2 component can abrogate this immunity. For instance, if mice are preimmunized so
that they have a mixed Thl+Th2 pattern of response to mycobacterial antigens, they become
more susceptible to the infection than are unimmunized control animals (Rook & HernandezPando, 1996). Similarly, the presence of a Th2 component can lead to increased tumour necrosis
factor alpha (TNFo0-mediated tissue damage (Hernandez-Pando & Rook, 1994), mimicking the
classical immunopathology of tuberculosis described by Robert Koch (1891).
It is clear that tuberculosis patients have an inappropriate Th2 component in their responses to
M. tuberculosis because they have IgE antibody (Yong et al., 1989), abnormally high expression
of the IL-4 gene in peripheral blood lymphocytes (Schauf et al., 1993), and demonstrable Th2
responses to mycobacterial antigen in their peripheral blood lymphocyte populations (Sanchez et
al., 1994).
We have therefore been investigating the possibility that changes in cortisol effects or levels
contribute to the shift towards Th2 in TB. It was already known that the diurnal rhythm of cortisol
production is lost in TB (Sarma et al., 1990), so that the immune system is exposed to cortisol
levels sufficient to occupy glucocorticoid receptors in T cells throughout the 24-h cycle, rather
than only in the morning as in normal individuals. We now report other endocrine changes that
may contribute to the development of an inappropriate Th2 component in human TB.

ANALYSIS OF 24-H OUTPUT OF STEROID METABOLITES IN THE URINE OF

TUBERCULOSIS PATIENTS

Glucocorticoid metabolites
To investigate the endocrine changes in tuberculosis, gas chromatography and mass spectrometry were used to quantify and identify adrenal steroid metabolites in 24-h urine collections. Full
details of the patients and controls are published elsewhere (Rook et al., 1996). The total output
of glucocorticoid derivatives was reduced by as a much as 50% in many patients as shown in Table 1 (Rook et al., 1996). Pardoxically, the 0800h plasma cortisol is normal or marginally raised
in TB patients (Sarma et al., 1990). This paradox is at least partly resolved by the observation
that there is also a striking increase in metabolites of cortisol relative to metabolites of cortisone
as reported in Table 1 (Baker et al., 1996; Rook et al., 1996). This may tend to maintain plasma
cortisoi, and indicates an increase in the activity of 11/3-reductase (cortisone ~cortisol) relative
to dehydrogenase (cortisol ~cortisone), which is discussed below.

Metabolites of dehydroepiandrosterone (DHEA)

In addition to the reduced total 24-h output of DHEA derivates, shown in Table 1 (Rook et al.,
1996), and a low ratio of DHEA/glucocorticoid derivatives, there was also a decreased conversion
of dehydroepiandrosterone (DHEA) to reduced forms (aetiocholanone and androsterone) with
increased conversion to 16o-hydroxylated forms, also indicated in Table 1 (Rook et ai., 1996).
This pattern of DHEA metabolism is not normally seen except in premature neonates (Kitada
et al., 1987). Further studies in a murine model (thymus protection assay; Blauer et al., 1991)
suggest that these 16x-hydroxylated derivatives are not active as antiglucocorticoids in an in vivo
system (Alnakhli S and Rook G, unpublished observations). We speculate therefore that both in

Steroid Metabolism and the Immune Response

Sl I I

Table 1. Steroid metabolites in 24-h urine samples from 7 male tuberculosis patients before
treatment, compared to 9 normal male donors
Mean:
DHEA

/.Ag/24 h
80.0

Androstentriol
322.9
16or(OH) DHEA
287.1
Androsterone
411.4
Aetiocbolanolone
432.8
Pregnanediol
132.9
Pregnanetriol
157. l
Tetrahydrocortisoi, THF 1080.0
Tetrahydrocortisone, THE
474.3
Ot-cortolone
245.7
fl-cortol and /3-cortolone
235.7
allo THF
518.6
11,6(OH) androsterone
305.7
1I/3(OH) aetiocholanolone
155.7
Total androgen derivatives 1534.3
Total cortisol derivatives
3015.7
Androgen/cortisol ratio
0.62

TB (n=7)
SD
100.8

465.7
338.7
309.9
179. l
175.8
157.7
928.6

377.8
212.8
115.0
665.2
167.5
93.8
1094.2
1939.9
0.48

U-TEST
0.029

Controls (n=9)
/.tg/24 h
SD
320.0
343.1

0. 15
413.3
0.46
255.6
0.00087 1216.7
0.0013
1037.7
0.63
94.4
0.4
233.3
0.46
1023.0
0.00086 2188.8
0.013
615.6
0.00005
854.4
0.017
727.8
0.0018
641. I
0.79
176.7
0.007
3243.4
0.0t)7
6227.8
0.7
0.53

116.5
85.9
190.9
369.5
56.4
86.5
261.5
595.5
277. I
231.7
148.1
95.7
99.9
525.5
1301.5
0.09

(Adapted from Rook et al. (1996))

tuberculosis and in the foetus, conversion of D H E A to 16ct-hydroxylated derivatives may serve
to reduce antiglucocorticoid function, and so divert the immune response away from Thl mode
(Wegmann et al., 1993).

D I S C U S S I O N : R E G U L A T I O N O F C O R T I S O L A C T I V I T Y IN T H E L U N G

The cortisol/cortisone shuttle

Both the concentration and efficacy of cortisol are regulated within target organs, and therefore
do not depend only on plasma concentrations. This regulation is achieved mainly by conversion
of cortisol to inactive cortisone within the tissue, and also by generation locally of steroids with
'antiglucocorticoid' activity. The latter are derived from another adrenal product, D H E A sulphate
(Blauer et al., 1991; Daynes et al., 1995; Rook et al., 1994). Such mechanisms are essential
since each peripheral organ needs to adjust the concentration and activity of cortisol to its own
requirements. This principle has been known for the kidney for some time. In normal kidneys
1 lfl-hydroxysteroid dehydrogenase type 2 (1 1/~HSD-2, an NAD-dependent enzyme) converts
cortisol to cortisone and so prevents cortisol from reaching the aldosterone receptor, which would
otherwise be swamped by it, leading to salt retention and hypertension (Walker, 1993). The same
mechanism is now known to occur in lymphoid tissue, where an unidentified form of 1 l f l H S D
is found in stromal cells (Berliner & Dougherty, 1961; Dougherty et al., 1960), and exerts a
potent effect on the Th 1/Th2 balance of the developing immune response within the lymph nodes
(Daynes et al., 1995).
In normal individuals the cortisone generated by these enzymes tends to be converted back
to cortisol by 11 flHSD type 1, (11/3HSD- 1, an NADPH-dependent oxido-reductase) in the liver,
where this enzyme functions as a reductase. Interestingly, a rise in the ratio of cortisol to cortisone
derivatives also occurs in patients with severe liver disease (Stewart et al., 1993), but excessive
activity of the liver enzyme in tuberculosis seems an unlikely explanation, and this enzyme is
thought to be functioning at near maximal rate in normal individuals.

s112

G . A . W . Rook & R. Hernandez-Pando

A more probable explanation, at least in pulmonary TB, is a change in enzyme activity in the
lung. Recently it has become apparent that normal lung parenchyma and bronchial epithelium
also contain l l/~HSD-1 (the enzyme found in the liver), but paradoxically in the normal lung
this enzyme appears to function as a dehydrogenase (or oxidase), converting cortisol to inactive
cortisone (Hubbard et al., 1994; Schleimer, 1991 ). However, since 11/~HSD- 1 is in fact a reversible
oxido-reductase, and affected by cytokines (Porter & Svec, 1995), it seems probable that the lung
is the site of the overall change in the balance of reductase to dehydrogenase activity that we have
detected. The next step is to discover whether in the tuberculous lung 11/~HSD- 1 has changed its
function and is converting cortisone to active cortisol (as this enzyme always does in the liver),
resulting in the striking dominance of cortisoi metabolites in the urine of TB patients.
The cortisol/DHEA ratio

Another finding from the analysis of the 24-h urine samples is the fall in DHEA/cortisol ratio.
Since DHEA opposes several effects of gluocorticoids in vivo (Blauer et al., 1991; Daynes et
al., 1995; Wright et al., 1992), the fall in DHEA/cortisol ratio must exacerbate the effects of the
decreased overall conversion of cortisol to inactive cortisone discussed above, and the loss of the
evening trough in cortisol levels (Sarma et al., 1990).
The importance of the ratio of glucocorticoid to DHEA has been studied in a murine model
of pulmonary tuberculosis, where it has emerged that if there is very little DHEA (as in normal
mice) the T lymphocyte response shifts progressively towards a Th2 cytokine profile, and the
animals die from pulmonary consolidation and pneumonia. However, if there is too much DHEA
relative to corticosterone, the animals also die, but from a tissue-destructive pathology that is not
characteristic of the disease in mice. There is, however, an optimal ratio of the two steroids that
is protective when given as supplements in late disease (Hernandez-Pando et al., manuscript in
preparation). The possibility that a critical appropriate ratio of glucocorticoid to antiglucorticoid
can protect against TB can also be deduced from the age-related changes in susceptibility and
pathogenesis of tuberculosis seen in human children. Tuberculosis is common in very small infants,
but is characterized by consolidation and pneumonia and is quite unlike the tissue-destructive
cavitatory adult type of disease (Donald et al., 1995). (In fact it resembles the disease seen in
mice not given DHEA supplements.) In infants of this age group DHEA levels are very low (dePeretti & Forest, 1978), as they are in mice. In contrast, TB is rare in the 5-10 year age group
when DHEA levels, which start rising at adrenarche, are about 50% of adult values, in spite of
evidence from skin-test surveys of continuing exposure to infection (Donald et al., 1995). Finally
the disease is common, and of adult type, in adolescents at puberty when adult DHEA levels are
achieved (Donald et al., 1995).
Clearly there are many other physiological changes during the same decades, but in view of
the correlation with the murine model, this hypothesis deserves further investigation.

REFERENCES
Baker, R, Zumla, A and Rook, GAW (1996) Tuberculosis, steroid metabolism and immunity. Q J Med 89,
387-394.
Berliner, DL and Dougherty, TF (1961) Hepatic and extrahepatic regulation of corticosteroids. Pharmacol
Rev 13, 329-359.
Blauer, KL, Poth, M, Rogers, WM and Bernton, EW (1991) Dehydroepiandrosterone antagonizes the
suppressive effects of dexamethasone on lymphocyte proliferation. Endocrinology 129, 3174-3179.
Brinkmann, V and Kristofic, C (1995) Regulation by corticosteroids of Thl and Th2 cytokine production
in human CD4+ effector T cells generated from CD45RO- and CD45RO+ subsets. J lmmunol 155,

Steroid Metabolism and the Immune Response

S113

3322-3328.
Brown, DH, LaFuse, W and Zwilling, BS (1995) Cytokine-mediated activation of macrophages from Mycobacterium bovis BCG-resistant and -susceptible mice: differential effects of corticosterone on antimycobacterial activity and expression of the Bcg gene (candidate Nramp). Inject Immun 63, 2983-2988.
Daynes, RA, Araneo, BA, Hennebold, J, Enioutina, J and Mu, HH (1995) Steroids as regulators of the
mammalian immune response. J Invest Dermatol 105, 14S-19S.
de Peretti, E and Forest, MG (1978) Pattern of plasma dehydroepiandrosterone sulfate levels in human
from birth to adulthood: evidence for testicular production. J Clin Endocrinol Metab 47, 572-577.
Donald, PR, Beyers, N and Rook, GAW (1995) Adolescent tuberculosis. S Afr Med J 86, 231-233.
Dougherty, TF, Berliner, ML and Berliner, DL (1960) 1 l fl-hydroxy dehydrogenase system activity in thymi
of mice following prolonged cortisol treatment. Endocrinology 66, 550-558.
Hernandez-Pando, R and Rook, GAW (1994) The role of TNFo~ in T cell-mediated inflammation depends
on the Thl/Th2 cytokine balance. Immunology 82, 591-595.
Hubbard, WC, Bickel, C and Schleimer, RP (1994) Simultaneous quantitation of endogenous levels of cortisone and cortisol in human nasal and bronchoalveolar lavage fluids and plasma via gas chromatographynegative ion chemical ionization mass spectrometry. Anal Biochem 221, 109-117.
Kitada, M, Kamataki, T, ltahashi, K, Rikihisa, T and Kanakubo, Y (1987) P-450 HFLa, a form of cytochrome
P-450 purified from human foetal livers, is the 16o~-hydroxylase of dehydroepiandrosterone 3-sulphate.
J Biol Chem 262, 13534-13537.
Koch, R (1891) Fortsetzung/Jber ein Heilmittel gegen Tuberculose. Deutsch Med Wochenschr 17, 101-102.
Mason, D (1991) Genetic variation in the stress response: susceptibility to experimental allergic encephalomyelitis and implications for human inflammatory disease, lmmunol Today 12. 57-60.
McCune, RM, Feldmann, FM, Lambert, HP and McDermott, W (1966) Microbial persistence. I. The
capacity of tubercle bacilli to survive sterilization in mouse tissues. J Exp Med 123, 445-468.
Porter, JR and Svec, F (1995) DHEA diminishes fat food intake in lean and obese Zucker rats. Ann NY
Acad Sci 774, 329-331.
Rook, GA, Steele, J, Ainsworth, M and Leveton, C (1987) A direct effect of glucocorticoid hormones on
the ability of human and murine macrophages to control the growth of M. tuberculosis. Eur J Respir
Dis 71, 286--291.
Rook GAW, Bloom BR (1994) Mechanisms of pathogenesis in tuberculosis. In: Bloom BR (Ed) Tuberculosis: Pathogenesis, Protection and Control. ASM Press, Washington DC, pp. 485-501.
Rook, GAW and Hernandez-Pando, R (1996) The pathogenesis of tuberculosis. Ann Rev Microbiol 50,
259-284.
Rook, GAW, Hernandez-Pando, R and Lightman, S (1994) Hormones, peripherally activated prohormones,
and regulation of the Thl/Th2 balance. Immunol Today 15, 301-303.
Rook, GAW, Honour, J, Kon, OM, Wilkinson, RJ, Davidson, R and Shaw, RJ (1996) Urinary steroid
metabolites in tuberculosis; a new clue to pathogenesis. Q J Med 89, 333-341.
Sanchez, FO, Rodriguez, Jl, Agudelo, G and Garcia, LF (1994) Immune responsiveness and hymphokine
production in patients with tuberculosis and healthy controls. Infect Immun 62, 5673-5678.
Sarma, GR, Chandra, I, Ramachandran, G, Krishnamurthy, PV, Kumaraswami, V and Prabhakar, R (1990)
Adrenocorticol function in patients with pulmonary tuberculosis. Tubercle 71, 277-282.
Schauf, V, Rom, WN, Smith, KA, Sampaio, EE Meyn, PA, Tramontana, JM, Cohn, ZA and Kaplan, G
(1993) Cytokine gene activation and modified responsiveness to interleukin-2 in the blood of tuberculosis
patients. J Infect Dis 168, 1056-1059.
Schleimer, RP (1991) Potential regulation of inflammation in the lung by local metabolism of hydrocortisone.
Am J Respir Cell Mol Biol 4, 166-173.
Stewart, PM, Burra, P, Shackleton, CHL, Sheppard, MC and Elias, E (1993) 11/3-hydroxysteroid dehydrogenase deficiency and glucocorticoid status in patients with alcoholic and non-alcoholic chronic liver
disease. J Clin Endocrinol Metab 76, 748-751.
Walker, BR (1993) Defective enzyme-mediated receptor protection: novel mechanisms in the pathophysiology of hypertension. Clin Sci 85, 257-263.
Wegmann, TG, Lin, H, Guilbert, L and Mosmann, TR (1993) Bidirectional cytokine interactions in the
maternal-fetal relationship: is succesful pregnancy a Th2 phenomenon? hnmunol Today 14, 353-356.
Wright, BE, Porter, JR, Browne, ES and Svec, F (1992) Antigluocorticoid action of dehydroepiandrosterone
in young obese Zucker rats. Int J Obes 16, 579-583.
Yong, A J, Grange, JM, Tee, RD, Beck, JS, Bothamley, GH, Kemeny, DM and Kardjito, T (1989) Total and
anti-mycobacterial IgE levels in serum from patients with tuberculosis and leprosy. Tubercle 70, 273-279.