possible etiology for unexplained chronic fatigue

[Guest guest]

I thought some of you might be interested is this article that was on one of

the CFS lists. A possible mechanism for the thyroid problems that many

develop.

Cheryl

Source: Medical Hypotheses

Volume 60, Issue 2, Pages 175-180

Date: February 2003, Pages 175-180

URL: http://www.sciencedirect.com/science/journal/03069877

Type I interferons induce proteins susceptible to act as thyroid

receptor (TR) corepressors and to signal the TR for destruction by

the proteasome: possible etiology for unexplained chronic fatigue

------------------------------------------------------------------

P. Englebienne(Corresponding Author), M. Verhas, C. V. Herst and K.

De Meirleir

University of Brussels (ULB/VUB), and RED Laboratories, N.V.,

Brussels, Belgium

Correspondence to: P. Englebienne PhD, Department of Nuclear Medicine

and Fatigue Clinic, Free University of Brussels (ULB/VUB), Brugmann

Hospital, Place van Gehuchten 4, B-1020 Brussels, Belgium. Phone:

+32-2-477-2651; Fax: +32-2-477-2432, E-mail: penglebi@...

Received 23 November 2001; accepted 8 May 2002.; Available online 10

December 2002.

Abstract

In some patients complaining of chronic fatigue such as those suffering

from the chronic fatigue syndrome (CFS), no underlying physical cause

can be clearly identified and they typically present a normal thyroid

function. Several studies indicate a dysregulation in the type I

interferons (IFN-alpha/beta) pathway in CFS resulting in a sustained

upregulation of 2',5'-oligoadenylate synthetases (2-5OAS). Likewise,

patients treated with IFN-alpha/beta usually complain of severe

fatigue as a limiting side effect. Beside the 2-5OAS, IFN-alpha/beta

induce also the expression of three closely related proteins of

unknown function termed the 2-5OAS-like (2-5OASL) proteins. The amino

acid sequences of the 2-5OASL proteins display 96% identity with the

partial sequence of the thyroid receptor interacting protein (TRIP)

14, further contain two typical thyroid hormone receptor (TR)

coregulator domains and feature two ubiquitin C-terminal domains.

>From these observations, we raise the hypothesis that the 2-5OASL

proteins are TRIPs capable of, respectively, repressing TR

transactivation and/or signaling the receptor for destruction by the

proteasome. Such molecular mechanisms could explain the development

of a clinical hypothyroid state in presence of a normal thyroid

function.

Hypothesis

Thyroid hormones (T3) play important roles in metabolism and are required

for the normal function of nearly all tissues, with major effects on

oxygen consumption and metabolic rate [1]. Consequently, chronic fatigue

is the hallmark symptom of primary or secondary hypothyroidism. In these

clearly defined clinical situations, abnormalities in the

hypothalamicpituitary-thyroid axis explain the disorders. In some less

well-defined clinical situations such as chronic fatigue syndrome (CFS),

the hallmark symptom, i.e., severe chronic fatigue, is not supported by

any objective biochemical origin and the syndrome is diagnosed by

exclusion [2]. Whilst biochemically euthyroid, CFS patients are

clinically hypothyroid [3] and some clinicians have even advocated T3

therapy [4] which was forcefully controversed [5].

The presence of a dysregulation in the type I interferon (IFN- small

alpha/beta) antiviral pathway in CFS has been repeatedly observed [6

and 7], which involves the upregulation of 2',5'-oligoadenylate

synthetase (2-5OAS) [8]. Interestingly, severe chronic fatigue is the

major adverse side-effect experienced by patients, respectively,

treated with IFN-alpha/beta for cancer [9], chronic hepatitis C [10],

or multiple sclerosis [11]. Although there is some evidence that

systemic IFN-alpha/beta therapy induces hormonal deficiency

syndromes, they do not explain the occurrence of fatigue in all

patients [12].

This parallelism leads us to raise the hypothesis that a cross-talk

takes place between the interferon antiviral pathway and transcription

mechanisms by the thyroid hormone receptor (TR). This cross-talk could

be mediated by 2-5OAS-like proteins (2-5OASL), also induced by

IFN-alpha/beta, and susceptible to repress transactivation of the TR

and/or of targeting the receptor for destruction by the proteasome.

If verified, these signal transduction mechanisms could be accounted

for a peripheral T3 resistance syndrome leading to a clinically

hypothyroid but biochemically euthyroid state, as observed in

diseases characterized by dysregulations in the antiviral pathway or

during the therapeutic use of INF-alpha/beta.

Elements supporting the hypothesis

THE 2-5OASL BUT NOT THE 2-5OAS PROTEINS SHARE SEQUENCE IDENTITY WITH

THE THYROID HORMONE RECEPTOR INTERACTING PROTEIN (TRIP) 14

Type I interferons induce the expression of ubiquitous proteins of the

2-5OAS family. This family encompasses the three 2-5OAS enzymes (p41/46,

p69/71, and p100) [13] along with three recently identified 2-5OASL

proteins, respectively, p30, p56, and p59 [14 and 15]. The amino acid

(aa) sequences of all the proteins within the family share a high degree

of similarity amounting up to 43%. However, the C-terminal part of p56

and p59 2-5OASL extends by another 150 residues beyond the sequences of

the three 2-5OAS and of the p30 2-5OASL. The genes coding for 2-5OAS

proteins are found on chromosome 12q24.1, although the genes coding for

the 2-5OASL proteins have been mapped to chromosome 12q24.2 [16]. It has

been suggested that the expansion of the 2-5OAS gene family occurred by

duplication of an ancestral gene, followed by domain coupling and

chromosomal segment duplication [16]. The role of 2-5OAS proteins is well

established. They polymerize ATP into 2',5'-oligoadenylates which

subsequently activate ribonuclease L, resulting in a general RNA

degradation [17]. Using the ClustalW program [18], we aligned the

catalytic site of 2-5OAS with the corresponding region of 2-5OASL. This

alignment shows that the critical Lys residue of 2-5OAS P-loop is

substituted by an Asn in 2-5OASL and that the three Asp residues which

are required for the catalytic activity of 2-5OAS [19] are, respectively,

substituted by Glu and Thr residues in the 2-5OASL proteins. Despite the

high degree of identity between the sequences within this domain, these

subtle differences explain the lack of enzymatic activity of the 2-5OASL

proteins [14], the physiological role of which has so far remained

elusive [15]. Interestingly, a BLAST search [20] performed at the

National Center for Biotechnology Information (NCBI) using the p56 OASL

sequence as the template reveals 96% identity with the partial sequence

(153 amino acids) of TRIP 14 (accession gi703104). This protein was

identified as interacting with the thyroid hormone receptor (TR) using

the yeast two-hybrid system [21]. The interaction was observed in cells

grown in the absence of thyroid hormone, but not in cells grown in

presence of the hormone. As shown in Fig. 1, the sequence alignment of

TRIP14 with the three 2-5OAS and p56/59 2-5OASL proteins (the TRIP14

sequence fragment is upstream to the p30 OASL C-terminal sequence)

indicates a much lower degree of identity with the 2-5OAS enzymes,

suggesting that the members of the family could have, respectively,

differentiated as enzymes (2-5OAS) at the one hand and TRIPs (2-5OASL)

at the other hand.

THE 2-5OASL PROTEINS COULD INTERACT WITH UNLIGANDED TR AS COREPRESSORS

THROUGH THE SPECIFIC PHI XX PHI PHI MOTIF

The TR regulates target gene expression directly through DNA T3 response

elements (TREs). The TR can interact with TREs as an homodimer but the

major form of TR bound to TREs is the heterodimer with the retinoid X

receptor (RXR) [22]. Transcriptional activation of liganded TR is

mediated by coactivator proteins that associate with the receptors in a

ligand-dependent manner and repression is mediated by the association of

the unliganded receptors with corepressor proteins. The coactivator

proteins interact with TR and RXR receptors through conserved hydrophobic

Phi xx Phi Phi (x stands for any aa) motifs made of a LeuxxLeu/IleLeu

pattern [23 and 24]. Two such receptor-binding domains (RBD) are

required, each interacting with one receptor of the heterodimeric

complex [25]. From the N-terminus of coactivator proteins, RBD1

displays preference for RXR, and RBD2 exhibits a strong preference

for TR [25] resulting from the presence of basic/polar residues

downstream to the Phi xx Phi Phi motif. The corepressor proteins

interact with unliganded receptors through an homologous hydrophobic

core motif containing the Ile/LeuxxIleIle pattern [26] and the

specificity of their interaction with TR is likewise governed by

distinct aa sequences within the interacting domains [27]. The

distinction between coactivators and corepressors is likely to be

governed by the nature of the nine aa upstream to the Phi xx Phi Phi

motif since chimeric peptide made of the coactivator motif in which

these nine aa are replaced by those of a corepressor interact with

TR in a ligand-independent manner [28]. The 2-5OASL proteins contain

the hydrophobic core motif and its alignment ( Fig. 2) with the

corresponding sequences of the RBD2 of the human TR corepressor Alien

[29] and of the coactivator TRAP220 [30] indicates that the 2-5OASL

motif is very similar to that of the corepressor. The similarity is

particularly remarkable in the sequence upstream to the hydrophobic

core motif. Consequently, that observation leads us to argue that the

2-5OASL proteins could interact as possible corepressors with TR in a

ligand-independent manner [21] through the conserved hydrophobic core

motif present in their aa sequence.

THE 2-5OASL PROTEINS COULD ALSO INTERACT WITH TR/RXR THROUGH AN

ADDITIONAL SCAN DOMAIN

In order to interact with either TR/TR homodimers or TR/RXR heterodimers,

corepressors and coactivators aa sequences must contain at least twice

the Phi xx Phi Phi motif, which is not the case in the 2-5OASL

proteins. Consequently, we examined further the sequence of the

2-5OASL proteins with an aim at identifying other possible interacting

domains. As shown in Fig. 3, we identified a strong homology with the

consensus sequence of the SCAN domain [31] in the N-terminal sequence

of the 2-5OASL proteins. In particular, the 2-5OASL proteins retain

most of the conserved residues of the SCAN proteins which are

relevant characteristics of the domain [31], including the central

GlnTrpLeuxPro and Leux6Leu motifs flanked by Asp, Glu, Leu, and His

residues. This homology is much less matched by the corresponding

sequences of the 2-5OAS proteins which do not contain the characteristic

motifs. The SCAN domain is a highly conserved Leu-rich motif of

approximately 60 aa originally found at the N-terminal end of Cys2His2

zinc finger transcription factors and more recently identified also in

proteins devoid of zinc fingers [32], such as the 2-5OASL. The SCAN

domain of these latter proteins, even partial, mediates their

oligomerization with zinc finger transcription factors of the Cys2Cys2

type such as the peroxisome proliferator-activated receptor-gamma

TR-sbeta and the estrogen receptor-alpha [33]. These observations

consequently suggest that the three 2-5OASL proteins could bind the

second nuclear receptor in the homo- or hetero-dimer through their

SCAN domain.

P56/59 2-5OASL PROTEINS CONTAIN TWO UBIQUITIN DOMAINS THROUGH WHICH

THEY COULD SIGNAL THE TR FOR DESTRUCTION BY THE PROTEASOME

Finally, an e-motif search [34] allowed us to identify two ubiquitin

(Ub) domains in the C-terminal part of p56 and p59 2-5OASL sequences

( Fig. 4). Domain 1 displays a strong conserved identity with the Ub

consensus and domain 2 contains two clear overlapping Ub signatures.

The Ub system targets proteins for proteolytic degradation by the 26S

proteasome [35]. Ub-mediated degradation of cellular proteins occurs

in both the cytosol and the nucleus and plays important roles in the

regulation of cell-cycle progression, signal transduction,

transcriptional regulation, and endocytosis [35 and 36]. Ub is a

small 76-aa protein which is covalently attached as a polymeric chain

to the target substrate. The normal process involves three steps,

namely Ub activation by E1 enzymes, transport and finally ligation by

Ub-ligases to the target substrate [36]. Alternatively, the second

step is skipped provided that a polyUb chain is already anchored to

the protein substrate [36]. This latter scenario could be retained in

the present case since we have shown above that the p56/59 2-5OASL

can bind the TR and RXR through their Phi xx Phi Phi and SCAN

domains. Activation of Ub for covalent attachment involves the

cleavage of a C-terminal GlyGly dipeptide. Albeit rich in Gly

residues, the C-terminus of the p56/59 OASL is devoid of such

dipeptide (Fig. 4). Intriguingly, IFN-alpha/beta induce also the

expression of an Ub homolog (UCRP) which is activated by E1 enzymes

in a way involving different, yet unidentified C-terminal residues

[37]. The absence of the GlyGly dipeptide in the C-terminal part of

the p56/59 OASL consequently does not necessarily precludes its

activation for covalent attachment to the TR/RXR by an Ub-ligase. In

order to attach Ub, the Ub-ligases bind the substrate on its

N-terminus. They contain two distinct sites that recognize either

basic or bulky-hydrophobic residues [36]. Although the N-terminus of

the TR-alpha isoforms is rich in basic residues (approximately 30%),

TR-beta and the RXR-alpha/beta and gamma forms contain bulk

hydrophobic motifs (i.a. ValxProAlaAla, AlaAlaxProProPhe, and

PheMetxPheProAla). Consequently, all the members of this receptor

family contain recruitment sites for Ub-ligases and the presence of a

polyUb-like protein bound to their complexes may signal them for

destruction by the proteasome.

Proposed mechanisms for induction of T3 resistance by the 2-5OASL

proteins

As depicted schematically in Fig. 5A, the unliganded TR/RXR heterodimer

could interact with the p30 OASL, which contains both the SCAN and

Phi xx Phi Phi motifs but is devoid of Ub homology. Due to the

presence upstream to the Phi xx Phi Phi motif of an aa stretch

analogous to the corresponding corepressor sequence, the complex could

activate the histone deacetylases which mediate TR repression [22].

Similarly, the TR/RXR unliganded complex could also interact with the

p56/p59 OASL proteins through the same SCAN and Phi xx Phi Phi motifs

(Fig. 5B). Moreover, these latter proteins contain two Ub domains.

Upon proper covalent attachment of these domains to the N-terminus of

one of the receptors by an Ub-ligase, the complex would be signaled

for degradation by the 26S proteasome. These two mechanisms provide a

possible logical explanation for the biochemically euthyroid,

clinical hypothyroid states which occur in ill-defined diseases

involving IFN-alpha/beta dysregulations such as CFS [3], or

during intensive IFN therapy [9, 10 and 11]. The further

documentation of these mechanisms would undoubtedly allow the

development of means intended at counteracting these unwanted effects

of signal transduction by IFN-alpha/beta.

Figure captions

Fig. 1. Alignment of the respective aa sequences (single letter code)

of gi703104 (partial TRIP14 sequence), p56, and p59 2-5OASL

proteins and the three 2-5OAS enzymes. Identities are displayed

in black and conservative replacements in gray.

Fig. 2. Alignment of the 20 aa stretch containing the Phi xx Phi Phi

motif of the 2-5OASL proteins with the receptor binding domain

2 (RBD2) of the TR nuclear corepressor Alien and the

coactivator TRAP220, respectively. Identities are in black,

conservative replacements in gray and similar amino acids

are boxed. The aa stretch of the 2-5OASL proteins displays a

significant level of similarity with the corresponding sequence

of the Alien corepressor, particularly in the nine aa upstream

to the hydrophobic core.

Fig. 3. Alignment of the 60 amino acid stretch of the 2-5OAS family

members with the SCAN domain consensus (above). The amino acids

conserved in all the SCAN proteins identified so far are marked

by blacktriangle up tri, filled at the bottom of the alignment.

It is worth noting that beside the high degree of identity with

the consensus, the 2-5OASL proteins contain also the Leux6Leu

motif specific of this domain.

Fig. 4. Alignment of the C-terminal sequence of p56 and p59 OAS with

the ubiquitin (Ub) consensus and the Ub signatures,

respectively. The Xs stand for any aa.

Fig. 5. Proposed mechanisms for the induction of thyroid resistance by

the p30 (part A) and the p56/59 (part 2-5OASL proteins.

References

1. P.M. Yen , Physiological and molecular basis of thyroid hormone

action. Physiol. Rev. 28 (2001), pp. 1097-1142.

2. A.L. Komaroff and D. Buchwald , Chronic fatigue syndrome: an update.

Annu. Rev. Med. 49 (1998), pp. 1-13.

3. C. Shepherd , Giving thyroid hormones to clinically hypothyroid but

biochemically euthyroid patients: long term treatment is being

used. BMJ 315 (1997), p. 8-14.

4. G.R.B. Skinner, R. , M. et al., Thyroxine should be

tried in clinically hypothyroid, but biochemically euthyroid

patients. BMJ 314 (1997), p. 17-64.

5. G.J. Beckett , Giving thyroid hormones to clinically hypothyroid but

biochemically euthyroid patients: supporting author's view would be

unwise. BMJ 315 (1997), p. 8-13.

6. R.J. Suhadolnik, D.L. , K. O'Brien et al., Biochemical

evidence for a novel low molecular weight 2-5A-dependent RNase L in

chronic fatigue syndrome. J. Interferon Cytokine Res. 17 (1997), pp.

377-385.

7. K. De Meirleir, C. Bisbal, I. Campinne et al., A 37 kDa 2-5A-binding

protein as a potential biochemical marker for chronic fatigue

syndrome. Am. J. Med. 108 (2000), pp. 99-105.

8. R.J. Suhadolnik, N. Reichenbach, P. Hitzges et al., Upregulation of

the 2-5A synthetase/RNase L antiviral pathway associated with chronic

fatigue syndrome. Clin. Infect. Dis. 18 suppl. 1 (1994), pp.

S96-S104.

9. M. Kankuri, T.T. Pelliniemi, S. Pyrhonen et al., Feasibility of

prolonged use of interferon-apha in metastatic kidney carcinoma:

a phase II study. Cancer 92 (2001), pp. 761-767.

10. E.B. Melian and G.L. Plosker , Interferon alfacon-1: a review of its

pharmacology and therapeutic efficacy in the treatment of chronic

hepatitis C. Drugs 61 (2001), pp. 1661-1691.

11. K. Gottberg, A. Gardulf and S. Fredrikson , Interferon-beta,

treatment for patients with multiple sclerosis: the patients'

perceptions of the side-effects. Mult. Scler. 6 (2000), pp.

349-354.

12. T.H. , S. Walder and K.H. Hupart , Endocrine-mediated mechanisms

of fatigue during treatment with interferon-alpha. Semin. Oncol.

25 suppl. 1 (1998), pp. 54-63.

13. J. Chebath, P. Benech, A. Hovanessian, J. Galabru and M. Revel , Four

different forms of interferon-induced 2',5'-oligo(A) synthetase

identified by immunoblotting in human cells. J. Biol. Chem. 262

(1987), pp. 3852-3857.

14. D. Rebouillat, I. Marie and A.G. Hovanessian , Molecular cloning and

characterization of two related and interferon-induced 56- and 30-kDa

proteins highly similar to 2'5' oligoadenylate synthetase. Eur. J.

Biochem. 257 (1998), pp. 319-330.

15. R. Hartmann, H.S. Olsen, S. Widder, R. nsen and J. Justesen ,

p59OASL, a 2'5' oligoadenylate synthetase- like protein: a novel

human gene related to the 2'5' oligoadenylate synthetase family.

Nucleic Acids Res. 26 (1998), pp. 4121-4127.

16. S. Kumar, C. Mitnick, G. Valente and G. Floyd- , Expansion and

molecular evolution of the interferon-induced 2'5' oligoadenylate

synthetase gene family. Mol. Biol. Evol. 17 (2000), pp. 738-750.

17. Wreschner D. H., McCauley J. W., Skehel J. J., Kerr I. M. Interferon

action-sequence specificity of the ppp(A2'p)nA-dependent ribonuclease.

Nature 1981; 289: 414-417.

18. J.D. , D.J. Higgins and T.J. Gibson , ClustalW: improving the

sensitivity of progressive multiple sequence alignment through

sequence weighting, position-specific gap penalties, and weight

matrix choice. Nucleic Acids Res. 22 (1994), pp. 4673-4680.

19. S.N. Sarkar, A. Ghosh, H.W. Wang, S.S. Sung and G.C. Sen , The nature

of the catalytic domain of 2'5'-oligoadenylate synthetases. J. Biol.

Chem. 274 (1999), pp. 25535-25542.

20. S.F. Altschul, W. Gish, W. , E.W. Myers and D.J. Lipman , Basic

local alignment search tool. J. Mol. Biol. 215 (1990), pp. 403-410.

21. J.W. Lee, H.S. Choi, J. Gyuris, R. Brent and D.D. , Two classes

of proteins dependent on either the presence or absence of thyroid

hormone for interaction with the thyroid hormone receptor. Mol.

Endocrinol. 9 (1995), pp. 243-254.

22. J. Zhang and M.A. Lazar , The mechanism of action of thyroid

hormones. Annu. Rev. Physiol. 62 (2000), pp. 439-466.

23. L. Ko, G.R. Cardona and W.W. Chin , Thyroid hormone receptor-binding

protein, an LXXLL motif-containing protein, functions as a general

coactivator. Proc. Natl. Acad. Sci. USA 97 (2000), pp. 6212-6217.

24. M.A. Mhajan and H.H. s , A new family of nuclear receptor

coregulators that integrate nuclear receptor signaling through

CREB-binding protein. Mol. Cell Biol. 20 (2000), pp. 5048-5063.

25. Y. Ren, E. Behre, Z. Ren, J. Zhang, Q. Wang and J.D. Fondell ,

Specific structural motifs determine TRAP220 interactions with

nuclear hormone receptors. Mol. Cell Biol. 20 (2000), pp. 5433-5446.

26. P. Webb, C.M. , C. Valentine et al., The nuclear receptor

corepressor (NcoR) contains three isoleucine motifs (I/LXXII)

that serve as receptor interaction domains (IDs). Mol.

Endocrinol. 14 (2000), pp. 1976-1985.

27. R.N. Cohen, S. Brzostek, B. Kim, M. Chorev, F.E. Wondisford and A.N.

Hollenberg , The specificity of interactions between nuclear hormone

receptors and corepressors is mediated by distinct amino acid

sequences within the interacting domains. Mol. Endocrinol. 15

(2001), pp. 1049-1061.

28. X. Hu and M.A. Lazar , The CoRNR motif controls the recruitment of

corepressors by nuclear hormone receptors. Nature 402 (1999),

pp. 93-96.

29. U. Dressel, D. Thormeyer, B. Altincicek et al., Alien, a highly

conserved protein with characteristics of a corepressor for

members of the nuclear hormone receptor superfamily. Mol. Cell

Biol. 19 (1999), pp. 3383-3394.

30. C.X. Yuan, M. Ito, J.D. Fondell, Z.Y. Fu and R.G. Roeder , The

TRAP220 component of a thyroid hormone receptor-associated

protein (TRAP) coactivator complex interacts directly with

nuclear receptors in a ligand-dependent fashion. Proc. Natl.

Acad. Sci. USA 95 (1998), pp. 7939-7944.

31. A.J. , L.M. Khachigian, T. Shows and T. , Isolation

and characterization of a novel zinc-finger protein with

transcriptional repressor activity. J. Biol. Chem. 270 (1995),

pp. 22143-22152.

32. C. Schumacher, H. Wang, C. Honer et al., The SCAN domain mediates

selective oligomerization. J. Biol. Chem. 275 (2000), pp. 17173-

17179.

33. G. Castillo, R.P. Brun, J.K. Rosenfield et al., An adipogenic

cofactor bound by the differentiation domain of PPAR gamma.

EMBO J. 18 (1999), pp. 3676-3687.

34. C.G.N. Nevill-Manning, T.D. Wu and D.L. Brutlag , Highly specific

protein sequence motifs for genome analysis. Proc. Natl. Acad.

Sci. USA 95 (1998), pp. 5865-5871.

35. A. Hershko and A. Ciechanover , The ubiquitin system. Annu. Rev.

Biochem. 67 (1998), pp. 425-479.

36. A. Ciechanover , The ubiquitinproteasome pathway: on protein death

and cell life. EMBO J. 17 (1998), pp. 7151-7160.

37. J. Narasimhan, J.L. Potter and A.L. Haas , Conjugation of the 15-kDa

interferon-induced ubiquitin homolog is distinct from that of

ubiquitin. J. Biol. Chem. 271 (1996), pp. 324-330.

--------

© 2003 Elsevier Science Ltd.

_________________________________________________________________

Add photos to your messages with MSN 8. Get 2 months FREE*.

http://join.msn.com/?page=features/featuredemail