Eur. J. Biochem. 200,679-687 (1991)
Isolation, characterization, cDNA cloning and gene expression
of an avian transthyretin
Implications for the evolution of structure and function of transthyretin in vertebrates
Wei DUAN', Marc G. ACHEN', Samantha J. RICHARDSON', Michael C. LAWRENCE', Richard E. H. WETTENHALL',
Anthony JAWOROWSKI and Gerhard SCHREIBER'
' Russell Grimwade School of Biochemistry, University of Melbourne, Australia
CSIRO Division of Biomolecular Engineering, Australia
(Received March 13, 1991) - EJB 91 0353
A chicken liver cDNA library was constructed in bacteriophage AgtlO. A full-length transthyretin cDNA clone
was identified by screening with rat transthyretin cDNA and was sequenced. A three-dimensional model of chicken
transthyretin was obtained by computer-graphics-based prediction from the derived amino acid sequence for
chicken transthyretin and from the structure of human transthyretin determined by X-ray diffraction analysis
[Blake, C. C. F., Geisow, M. J., Oatley, S. J., Rtrat, B. & Rtrat, C. (1978) J. Mol. Biol. 121, 339-3561. The
similarity of the amino acid sequences of chicken and human transthyretins was 75% overall and 100% for the
central channel containing the thyroxine-binding site. Also, the organization of the transthyretin gene into exons
and introns and the tissue specificity of expression of the transthyretin gene were similar in chicken and mammals,
despite an evolutionary distance of about 3 x lo8 years from their common ancestor, the Cotylosaurus. By far the
highest levels of transthyretin mRNA were found in choroid plexus. The data suggest a fundamental role for the
cerebral expression of transthyretin in all vertebrates. It has been proposed that this role is the transport of
thyroxine from the bloodstream to the brain [Schreiber, G., Aldred, A. R., Jaworowski, A., Nilsson, C., Achen,
M. G. & Segal, M. B. (1990) Am. J. Physiol. 258, R338-R345].
Transthyretin is one of the three thyroid hormone-binding
proteins in the blood of larger mammals [l]. It differs from
the other two thyroid hormone-binding plasma proteins, thy-
roxine-binding globulin and albumin, in three characteristic
features: firstly, the gene encoding transthyretin is expressed
at a high rate in the choroid plexus in the brain [2-111;
secondly, transthyretin is enriched in the cerebrospinal fluid
; and thirdly, no genetically caused absence has ever been
reported for transthyretin, in contrast to the observations for
albumin and thyroxine-binding globulin [13 - 181. Therefore,
the question arises as to whether transthyretin has an essential
function in the brain, distinct from that in the bloodstream.
The transport of thyroxine in the brain has been proposed as
an important function of transthyretin synthesis in the choroid
plexus [2, 191.
The structures of transthyretin and its gene are known for
various mammals [3, 4, 20-281, but not for other species.
The purpose of this investigation was the characterization of
transthyretin, its gene and pattern of expression in a species
whose evolution was sufficiently different from that of the
Correspondence to G. Schreiber, Russell Grimwade School of
Biochemistry, University of Melbourne, Parkville, Victoria, Australia
Enzymes. Restriction endonucleases (EC 22.214.171.124); DNA poly-
merase (EC 126.96.36.199); SP6 RNA polymerase (EC 188.8.131.52); DNase I
(EC 3.1.21 .I); pancreatic RNase (EC 184.108.40.206); ribonuclease TI (EC
220.127.116.11); reverse transcriptase (EC 18.104.22.168).
Note. The novel nucleotide sequence data published here have
been deposited with the EMBL sequence data bank.
mammals to allow the evaluation of both structural and func-
tional conservation during evolution. Birds and mammals had
a common ancestor in the stem reptiles (Cotylosaurus), living
about 3 x lo8 years ago [29,30] (for review see ). This time
is sufficient for two identical nucleotide sequences, which are
not subjected to selection pressure, to diverge completely [32,
331. Therefore, chicken transthyretin was isolated from serum,
its N-terminus sequenced, a cDNA library constructed from
chicken liver, a full-length transthyretin cDNA isolated, se-
quenced and used to analyze the tissue specificity of trans-
thyretin gene expression. A three-dimensional model was de-
veloped for chicken transthyretin based on the derived pri-
mary structure of chicken transthyretin and the crystal struc-
ture of human transthyretin. The model was used for the
interpretation of the function of transthyretin in vertebrates.
Animals and materials
White Leghorn chickens, 6 weeks old, were obtained from
a local slaughterhouse. Thermus aquaticus (Taq) DNA poly-
merase was purchased from Cetus Corporation. The sources
of other chemicals and enzymes were as described previously
[4, 5, 7, 341. Tissues were removed from animals within 5 min
after death. Brains were kept on ice and dissection was com-
pleted within 4 h after death. All tissue samples were frozen
in liquid nitrogen immediately after dissection and stored at
- 70 "C until use.
Purification of chicken transthyretin
Chicken plasma was dialyzed overnight at 4°C against
25 mM Tris/HCl pH 6.8 (200 volumes), then centrifuged at
12000xg for 15 min to remove any precipitated material.
Transthyretin was purified from 3 ml dialyzed plasma (0.13 g
total protein) by preparative disc electrophoresis in polyacryl-
amide gel at pH 8.6, 4°C. The method of Ornstein  and
Davis  was adapted to large-scale preparation with con-
tinuous elution, in a Buchler preparative polyacrylamide gel
electrophoresis apparatus (Buchler Instruments, Fort Lee,
NJ); 10 mM Tris/glycine pH 8.3 and 0.375 M Tris/HCl
pH 8.9 were used as upper and elution buffers, respectively.
Electrophoresis was performed at 150 V with an elution rate
of 1 ml/min. Individual fractions (8 ml each) were analyzed
spectrophotometrically at 280 nm and by SDS/polyacryl-
amide gel electrophoresis. Fractions containing transthyretin,
but free of detectable serum albumin, were pooled, then con-
centrated by ultrafiltration using a Diaflow concentrator
(Amicon) with a PMlO membrane. The yield of transthyretin
from 3 ml plasma was 0.5 mg.
For the analysis of the N-terminal amino acid sequence,
200 pg purified transthyretin was exhaustively dialyzed at 4°C
against 50 mM NH4HC03, then freeze-dried and resuspended
in a total volume of 1 .O ml ultrapure water (Milli Q, Millipore,
Australia). All glassware used for the preparation of protein
for sequence analysis was washed in 'Chromerge' (Fisher
Scientific Company, USA) and rinsed extensively in ultrapure
Analysis of N-terminal sequence of chicken transthyretin
Automated Edman degradation of the transthyretin (5 pg)
immobilised in a Biobrene (Applied Biosystems) matrix was
carried out on an Applied Biosystems model 477A gaslliquid-
phase sequenator  and the amino acid sequence determined
by on-line quantitative HPLC analysis of phenylthiohydan-
toin derivatives , using an applied Biosystems model 120
HPLC analyzer. The analysis gave a single unambiguous se-
quence with an initial yield of 120 pmol, which was equivalent
to 40% of the starting material, a typical yield for sequence
analysis under these conditions (see ).
Molecular mass of chicken transthyretin
The molecular mass of chicken transthyretin was estimated
by FPLC/gel-permeation chromatography using a Superose-
12 column (HR 10/30, Pharmacia), equilibrated in 0.15 M
NH4HC03 pH 8.0. A sample of 0.25 ml purified transthyretin
(containing 60 pg protein) was chromatographed at a flow
rate of 0.25 ml/min. Fractions of 0.5 ml were collected, and
the absorbance was measured at 280nm. The column was
calibrated with porcine insulin (5 kDa), horse heart
cytochrome c (13 kDa), horse heart myoglobin (17 kDa), hen
ovalbumin (43 kDa) and bovine serum albumin (67 kDa).
Assays of protein concentrations
The apparent protein content of serum samples was esti-
mated by the microbiuret method of Itzhaki and Gill  and
that of solutions of purified transthyretin by the method of
Lowry et al. , using defatted bovine serum albumin as the
Polyacrylamide gel electrophoresis
Protein was analysed by SDS/polyacrylamide gel electro-
phoresis using a 15% polyacrylamide resolving gel and a 4.5%
stacking gel in the discontinuous buffer system of Laemmli
and Favre . Proteins were visualised, after electrophoresis,
with Coomassie brilliant blue R250 followed, when higher
sensitivity was required, by the silver staining procedure of
Merril et al. .
Analysis of the binding of thyroxine to serum proteins
and to purijied transthyretin
Aliquots of chicken plasma and purified transthyretin were
incubated with 1 nM ['251]thyroxine for 3 h at room tempera-
ture, then loaded onto a 10% polyacrylamide gel, which had
been subjected to prior electrophoresis for 90 min at 125 V,
using 0.05 M Tris/glycine pH 8.3. An aliquot of chicken
plasma, without radioactive thyroxine, was applied to an ad-
jacent well. Electrophoresis was carried out at 125 V for 5 h
at room temperature with buffer recirculation between
chambers. The track containing chicken plasma was cut out
and the proteins were stained with Coomassie brilliant blue
R250 to determine the positions of albumin and transthyretin.
The remainder of the gel was dried and autoradiographed.
Construction of cDNA library, isolation
of a transthyretin cDNA and nucleotide sequencing
A chicken liver cDNA library was constructed in bacterio-
phage igtlO as described previously . Approximately
20000 plaques were screened by hybridization  with a rat
transthyretin cDNA probe  labelled with [R-~'P]~ATP
random hexanucleotide-primed synthesis [46, 471. A 0.67-kb
cDNA was subcloned into plasmid pGEM-3Zf(+). The re-
combinant pGEM-3Zf(+) plasmid was purified by gel fil-
tration using a Superose-6 column . The entire sequence
of the cDNA was determined on both strands by the method
of Sanger et al. .
Northern and Southern analyses
For Northern analysis, RNA was prepared [50, 511 and
fractionated according to size by electrophoresis in denaturing
agarose/formaldehyde gel, as described by Sambrook et al.
, except that Hepes was substituted in the electrophoresis
buffer for Mops and that the buffer was circulated. RNA was
transferred to nitrocellulose filters and hybridized at 42 O C for
18 h with 32P-labelled cDNA probes. Conditions for hy-
bridization and washing were as described by others ,
except that washing was carried out at 45 "C. Southern analysis
was carried out as described by others . DNA probes were
labelled with [R-~~PI~ATP by random hexanucleotide-primed
synthesis [46, 471.
Polymerase chain reaction
For reverse transcription coupled to polymerase chain re-
action, the following procedure was used: 10 pg total cellular
RNA was treated with 20 units ribonuclease-free DNase I in
40 p120 mM Tris/HCl pH 7.6, 5 mM MgCI2, at 37 "C for 4 h.
The mixture was heated at 80°C for 3 min and the RNA was
precipitated with ethanol. RNA was then dissolved in 42 ~l
water. One half of the RNA sample was hybridized to 2 pg
(dT)12-18 in 24 pl water at 75°C for 10 min, then chilled on
ice. As a control, the other half of the RNA sample was
digested with 0.8 pg RNase A and 1.6 units RNase TI in 40 pl
water at 37°C for 2 h. The control sample was then dried in
vucuo, resuspended in 22 p1 water and hybridized to (dT)12 - 18
as above. For both samples, first-strand cDNA synthesis was
then carried out in 40 p1 cDNA buffer (50 mM Tris/HCl
pH 8.3, 75 mM KCl, 3 mM MgC12, 5 mM dithiothreitol,
0.25 mM each dATP, dGTP, dCTP, dTTP) in the presence of
40 units ribonuclease inhibitor and 400 units Moloney murine
leukemia virus reverse transcriptase, at 42°C for 1 h. Both
samples were then incubated at 95°C for 5 min to inactivate
reverse transcriptase, rapidly chilled on ice and precipitated
with ethanol. Polymerase chain reactions were carried out in
10 mM Tris/HCl pH 8.3,50 mM KCl, 1.5 mM MgCl,, 0.01 %
(mas/vol.) gelatin, 200 pM of each dNTP, 100 ng of both 5’
and 3’ primers (intron I 5’ and intron 111 3’ primers, see below)
and 2.5 units Taq DNA polymerase, in a volume of 100 pl.
The reaction mixtures were overlaid with 100 p1 mineral oil.
The amplification involved 40 cycles of denaturation at 94°C
for 30 s, annealing at 45°C for 30 s and extension at 70°C for
Amplification of introns of the transthyretin gene
from chicken liver genomic DNA
Chicken liver genomic DNA (1 pg) was prepared  and
subjected to 35 cycles of amplification by the polymerase chain
reaction as described above, except that the extension time
was 2.5 min. The regions of the chicken liver transthyretin
cDNA (Fig. 2) which corresponded to the sequences of the
oligonucleotides used for polymerase chain reactions were as
follows : intron I 5’ primer (5’-GTTTTCTTAGCTGGACT-
GGT-3’), bases 50 - 69; intron I 3’ primer (5’-ACCATGAG-
AGGGCATTTGGA-3’), complementary to bases 135 - 116;
intron I1 5’ primer (5’-GGAACCTGGCAGGACTTTGC-3’),
bases 209 - 228; intron I1 3’ primer (5’-ACTGTTCTTCTGT-
TGTGAGC-3’), complementary to bases 284 - 265; intron
111 5’ primer (5’-CCTTTCCCCATTCCATGAAT-3’), bases
343 - 362; intron 111 3’ primer (5’-CTGAGGAGAGCAGC-
GATGGT-3’), complementary to bases 429 - 41 0. The prod-
ucts of the amplifications were separated by electrophoresis
in 1% agarose gel and analyzed by Southern analysis, using
32P-labelled chicken transthyretin cDNA as a probe.
R NA quantification
Quantification of transthyretin mRNA in chicken choroid
plexus and liver was achieved using the ribonuclease protec-
tion assay of Lee and Costlow , with monitoring of trans-
thyretin mRNA recovery during RNA purification. The ex-
perimental details were as described previously [2,34] with the
following exceptions : (a) the tracer RNA (approximately 720
bases) was transcribed in vitro  with bacteriophage SP6
RNA polymerase from a DNA template constructed by inser-
tion of a conglutin 6 cDNA from Lupinus angustifolius  at
the unique EcoRI restriction endonuclease site of plasmid
pGEM-4Z; (b) the transthyretin antisense RNA probe used
in ribonuclease protection assays was transcribed from a DNA
template constructed by insertion of a chicken liver trans-
thyretin cDNA (described in Results) at the unique EcoRI
restriction endonuclease site of plasmid pGEM-3Zf(+); (c)
the molecular mass of chicken transthyretin mRNA was as-
sumed to be 240 kDa (based on the size of chicken trans-
thyretin mRNA and the average molecular mass of
Derivation of accepted point mutations for proteins
Complete amino acid sequences of the proteins cited were
obtained from the Protein Database, version 25, or from the
Provisional Protein Database, version 43, of the National
Biomedical Research Foundation (USA) using the Protein
Sequence Query retrieval program . The sequences were
aligned using ALIGN and COMPARE programs [56 - 581
to allow maximum possible matches. A sequence break or
deletion was counted as one mutation event, irrespective of
the number of amino acids involved. For each protein, the
number of accepted point mutations was derived from the
number of observed amino acid mutations using a conversion
table devised by Dayhoff . For this analysis, the divergence
time between mammals and birds was assumed to be 300
million years [29, 301. Mammalian radiation was assumed to
have occurred 85 million years ago .
Isolation of chicken transthyretin
Transthyretin was isolated from chicken plasma with the
aim of obtaining its N-terminal amino acid sequence. Chicken
transthyretin has a higher electrophoretic mobility at pH 8.3
than any other abundant plasma protein in chicken. Thus,
considerable purification was achieved using preparative
polyacrylamide gel electrophoresis under non-denaturing con-
ditions. Gel permeation chromatography, using a Superose-
12 column, indicated an apparent molecular mass of 56.4 kDa,
agreeing with previous estimates of the molecular mass of the
native protein . Upon electrophoresis in SDSIPAGE, the
protein dissociated into subunits of mass 15 kDa, agreeing
with previous estimates by others  and confirming that
chicken transthyretin, like mammalian transthyretins, is a
tetramer. Further evidence for the identity of the purified
protein was obtained by demonstration that the protein bound
[ ‘251]thyroxine and migrated with the same electrophoretic
mobility under non-denaturing conditions at pH 8.6 as the
complex of thyroxine and transthyretin from serum.
Amino terminus of chicken transthyretin
The unambiguous N-terminal amino acid sequence of
purified chicken transthyretin obtained by automated Edman
degradation was Ala-Pro-Leu-Val-Ser-His-Gly-Ser-Val-Xaa-
identification of an amino acid residue was not possible in
cycles 10 and 13. The amino acid sequence showed that the
alanine residue, representing residue 21 in the primary trans-
lation product, was the N-terminus of the mature protein.
Therefore, processing of the signal peptide occurs at the Ala-
Ala bond predicted from the cDNA sequence. The site of
processing agrees with the predictive rules of von Heijne .
However, our data remove any ambiguity as to which alanine
residue is the start of the mature polypeptide. The cDNA
sequence of chicken transthyretin indicates the intercalation
of three extra amino acids, -Val-Ser-His-, near the amino
terminus, compared to various mammalian transthyretins (see
Fig. 3). The analysis of the amino acid sequence of the protein
confirmed the intercalation of these three extra amino acids
in chicken transthyretin.
Isolation and characterization of chicken transthyretin cDNA
A chicken transthyretin cDNA was isolated from a chicken
liver cDNA library and the nucleotide sequence was analyzed
as described in Experimental Procedures. The sequencing
strategy is shown in Fig. 1. The nucleotide and deduced amino
acid sequences are depicted in Fig. 2. The cDNA contained
656 nucleotides followed by 14 adenylate residues. After re-
moval of poly(A) segments, the. sizes of the transthyretin
mRNAs from chicken liver and choroid plexus were 0.65 kb
(Northern analysis not shown). Thus, the chicken trans-
500 6w 700 base pairs
Fig. 1. Strategy for nucleotide sequencing of the chicken liver trans-
thyretin cDNA. The chicken transthyretin cDNA is depicted as a
hatched bar. The EcoRI cloning sites flanking the cDNA and the Sac1
site used for subcloning are shown. Arrows indicate the direction and
the extent of sequencing
thyretin cDNA described above is near full-length. The first
ATG codon is located 26 bases downstream from the 5' end
of the cDNA. It is followed by an open reading frame of 450
nucleotides ending with the stop codon TGA at nucleotides
476 - 478. Compared with mammalian transthyretin cDNAs,
the chicken transthyretin cDNA has an extra nine nucleotides
which encode the amino acids Val, Ser and His after the third
amino acid of the mature protein (see above). The
untranslated sequence at the 3'-end (181 bases) preceding the
poly(A) segment possesses two additional in-frame stop
codons (nucleotides 527 - 529 and 635 - 637). The sequence
CAGAATGG (nucleotides 22 - 29) which contains the first
ATG codon agrees with the favored sequence that flanks the
translational start sites found in eukaryotic mRNAs . A
polyadenylation motif, AATAAA , is present 18 nucleo-
tides upstream from the 5' end of the poly(A) segment. The
entire chicken transthyretin cDNA sequence displays 63%,
63%, 63% and 62% similarity to the sequences of human ,
sheep 1261, mouse , and rat  transthyretin cDNAs,
respectively. The sequence similarities within the coding re-
ATG GCC TTT CAC TCT ACT CTT CTT GTT W C TTA GCT GGA CTG GTA TTT CTC TCC
Ket Ala Phe H i s Ser Thr Leu Leu Val Phe Leu Ala Gly Leu Val Phe Leu Ser
Glu Ala Ala Pro Leu Val Ser H i s Gly Ser Val Asp Ser Lys Cys Pro Leu Ket
CCA Crc; GTC TCC CAT GGC TCT GTT GAT TCC AAA TCC CCT CTC ATG
GTG AAA GTG CTG GAT C % A GTC AGA GGA AGT CCT GCA GCT AAC GTA GCT GTT AAA
Val Lys Val Leu Asp Ala Val Arg Gly Ser Pro Ala Ala Asn Val Ala Val Lys
GTC TTT AAA AAG GCT GCA GAT GGA ACC TGG CAG GAC TTT GCT ACT GGG AAA ACC
Val Phe Lys Lys Ala Ala Asp Gly Thr Trp Gln Asp Phe Ala Thr Gly Lys Thr
ACA GAG TTC GGT GAA ATC CAT GAG CTC ACA ACA GAA GAA CAG TTT GTA GAA GGA
T h r Glu Phe Gly Glu Ile H i s Glu Leu T h r T h r Glu Glu Gln Phe Val Glu Gly
GTA TAC AGA GTT GAG TTT GAC ACC AGT TCT TAC TGG AAG GGA CTT GGC CTT TCC
Val T y r Arg Val Glu Phe Asp T h r Ser Ser T y r Trp Lys Gly Leu Gly Leu Ser
CCA TTC CAT GAA TAT GCT GAT GTG GTG TTC ACT GCT AAT GAT TCT GGC CAC CGC
Pro Phe H i s Glu Tyr Ala Asp Val Val Phe Thr Ala Asn Asp Ser Gly H i s Arg
CAT TAT ACC ATC GCT GCT CTC CTC AGT CCT TTC TCT TAC TCA ACA ACT GCT GTT
H i s Tyr Thr Ile Ala Ala Leu Leu Ser Pro Phe Ser Tyr Ser Thr Thr Ala Val
GTC AGT GAT CCC CAG GAA TGAAAACATACTGTSWTAGATATATATACCCTCCTCETAGGACT
Val Ser Asp Pro Gln Glu ***
G T T T T T C A ~ T G T M C C ~ ~ G T S W T W T G T A A C T T C T T T W ~
Fig. 2. Nucleotide sequence of chicken liver transthyretin cDNA and deduced amino acid sequence o f chicken transthyretin. Nucleotides are
numbered at the right of the figure in the 5' to 3' direction, beginning with the first nucleotide of the cDNA insert. The deduced amino acid
residues are indicated below the nucleotide triplets. The amino acid residues of the signal peptide are numbered -20 to - 1. The N-terminal
residue of mature chicken transthyretin, as determined by amino acid sequencing, is designated as + 1. The first in-frame stop codon is marked
by three asterisks. A potential polyadenylation signal in the 3'-untranslated region of the cDNA is underlined
Fig. 3. Comparison of the primary structures of six transthyretins. The amino acid sequence of rabbit transthyretin  and those deduced from
cDNA sequences for sheep , rat [3,25], mouse  and chicken (this paper) transthyretins are aligned with the deduced amino acid sequence
of human transthyretin . The N-terminal residue of mature transthyretin (designated + 1) has been identified for human , rabbit ,
rat  and chicken (this paper), but has been inferred for mouse and sheep from this alignment. Numbers below the sequences refer to the
amino acids of chicken transthyretin. Those residues in sheep, rabbit, rat, mouse and chicken transthyretins which are identical to those in
human transthyretin are represented by asterisks. Presegments contain the residues numbered from -20 to - 1. Features of secondary
structure in human transthyretin [22, 721 are indicated above the sequence for human transthyretin. Residues located in the core of the
transthyretin subunit are singly underlined, those located in the central channel of the tetramer are doubly underlined, and those which
participate in the binding of thyroxine  are enclosed in rectangles. The positions of the three arrows below the amino acid sequences
correspond to the positions of the intron-exon boundaries in the human  and mouse  transthyretin genes
gion are 73% (human), 72% (sheep), 74% (mouse) and 73%
(rat). The mass of the subunit of chicken transthyretin calcu-
lated from the amino acid sequence deduced from the cDNA
sequence is 13 836 Da.
Comparison of the primary structures of transthyretins
from six different species
The amino acid sequence of chicken transthyretin derived
from the nucleotide sequence of a chicken liver transthyretin
cDNA (Fig. 2) was compared with the derived amino acid
sequences of the transthyretins reported earlier for rat [3, 251
and sheep , and the sequences published by others for
human , mouse  and rabbit , as shown in Fig. 3.
Three extra amino acids result in an overall length of 130
amino acids for the subunit of chicken transthyretin, com-
pared with 127 amino acids for the subunit of mammalian
transthyretins. The overall similarity of the amino acid se-
quence of chicken and those of mammalian transthyretins is
between 75-80%. The amino acids which are thought to
participate in the binding of thyroxine  are conserved
Comparison of the three-dimensional structures of human
and chicken transthyretins
The X-ray structure of human transthyretin is known for
residues 10- 123 . In this region, 90 residues of the chicken
sequence are identical to those of human transthyretin; 12 out
of the 24 substitutions are conservative. This suggests that the
secondary, tertiary and quaternary structures of chicken and
human transthyretin are highly similar. A model for the three-
dimensional structure of chicken transthyretin, based on the
X-ray structure of human transthyretin , was built using
the interactive molecular-graphics program QUANTA (Poly-
gen Corporation). The a-carbon backbone trace of the chicken
transthyretin model is presented in Fig. 4. The transthyretin
tetramer contains a central channel with the binding site for
thyroxine . None of the amino acid substitutions were
found to occur in this channel, indicating that the structure
of the thyroxine-binding site in the two species is highly simi-
lar. The majority of the substitutions occur on the outer sur-
face of the protein. All amino acid replacements could be
accommodated easily without steric hindrance. This result
was confirmed by energy minimization calculation . The
restrained minimization gave an energy of - 18.46 MJ/mol
(- 441 7 kcal/mol) and - 20.56 MJ/mol ( - 491 8 kcal/mol) for
the human and chicken transthyretin dimers, respectively,
indicating that the model-building of the chicken structure
leads to a structure energetically comparable with that of the
human protein. The root-mean-square shifts of the backbone
atom positions from their corresponding human transthyretin
X-ray coordinates, as a result of the minimization, are 9.3 pm
and 8.9 pm for the human and chicken species, respectively.
The human and chicken transthyretin structures obtained by
energy minimization have a root-mean-square difference of
4.6 pm in their backbone atoms. These differences are small,
again indicating that the substitutions can readily be accom-
modated in the X-ray structure of human transthyretin. In
summary, therefore, it is highly likely that the secondary and
tertiary structures of human and chicken transthyretins are
Conservation o f the organization o f the transthyretin gene
into exons and introns in mammals and chickens
Since the boundaries between introns and exons are fre-
quently correlated with the structural organization of a pro-
tein into domains, and positions of introns are often
maintained during phylogenesis [67, 681, it was of interest to
study the organization of the chicken transthyretin gene. A
comparison of the structure of the transthyretin gene in hu-
mans , mice  and rats  reveals conservation of the
relative positions of the exon - intron junctions of the trans-
thyretin gene in the three species. Therefore, the positions of
introns of the chicken transthyretin gene corresponding to
the three introns found in mammals were analyzed. For this
analysis, oligonucleotide primers for polymerase chain reac-
tions were synthesized which hybridized to chicken cDNA
near regions corresponding to the exon - intron boundaries
in the mammalian transthyretin genes (see Fig. 3 and Exper-
imental Procedures). Polymerase chain reactions were carried
out using genomic DNA from chicken liver and chicken trans-
thyretin cDNA as templates. The sizes of the fragments ampli-
fied from the cDNA were as predicted from the nucleotide
sequence of the cDNA, whereas amplification of genomic
DNA gave rise to much larger fragments. The identities of the
products of the genomic DNA amplifications were checked
by Southern analysis using the chicken transthyretin cDNA
as a hybridization probe. The sizes of introns of the chicken
transthyretin gene estimated from the Southern blots were
0.8 kb for intron I (compared with 0.92 kb for humans ,
0.95 kb for mice , and 1 kb for rats ), 2 kb for intron I1
(compared with 2.09 kb for humans , 3.4 kb for mice 
and 2.6 kb for rats ) and 2.4 kb for intron 111 (compared
with 3.3 kb for humans , 3.6 kb for mice  and 2.8 kb
Fig. 4. Trace of the a-carbon backbone o f the human transthyretin
tetramer. A model for the three-dimensional structure of chicken
transthyretin, based on the X-ray structure of human transthyretin
, was built using the interactive molecular-graphics program
QUANTA (Polygen Corporation). The coordinates for human trans-
thyretin were obtained from the Brookhaven Protein Data Base 
and, using the ‘mutate’ facility in QUANTA, were replaced on a
residue-by-residue basis to obtain the chicken transthyretin sequence.
In the X-ray structure study of human transthyretin, residues 1-9
and 124 - 127 are not defined, hence sequence differences between the
chicken and human species in these rcgions werc not considered.
The plausibility of the model-building was investigated further by
subjecting both the model and the human structure to 40 cycles of
conjugate gradient energy minimization using the program XPLOR
. To avoid spurious conformational changes during the minimiza-
tion, the coordinates of the backbone atoms of both structures were
restrained in a harmonic fashion to the X-ray coordinates of the
human structure. The structure is viewed down the crystallographic
z-axis (A) and x-axis (B). Residues that differ in the chicken and
human transthyretin sequences are shown as bold segments. The
absence of changes in the central channel containing the thyroxine-
binding site is apparent
Fig. 5. Northern analysis of chicken transthyretin mRNA. Total RNA
(20 pg, unless specified otherwise) purified from various tissues was
subjected to Northern analysis as described in Experimental Pro-
cedures, using 32P-labelled chicken transthyretin cDNA as a probe.
Autoradiographic exposure was for 36 h at - 70°C. The sizes of RNA
molecular mass markers (in I000 bases) are shown on the left, the
positions of 28-S and 18-S ribosomal RNA bands on the right
for rats ). In summary, the organization of the transthyretin
gene into exons and introns in chickens reflected that in mam-
mals, suggesting strong conservation of the overall structure
of the gene for the last 3 x 10’ years.
Pattern o f tissue-specific expression o f the transthyretin gene
Total RNA was extracted from various tissues of the
chicken and studied by Northern analysis. The results are
shown in Fig. 5. By far the greatest abundance of transthyretin
mRNA was found in choroid plexus RNA. The signal re-
sulting from only 1.0 pg choroid plexus total RNA was
stronger than that for 20 pg total RNA from liver or eye.
For detection of low levels of transthyretin mRNA, the
most sensitive approach available, reverse transcription cou-
pled to polymerase chain reaction (see Experimental Pro-
cedures), was employed. A signal for transthyretin mRNA was
detected in intestine, lung, kidney and spleen. No transthyretin
mRNA could be detected in ‘rest of brain’ (the brain after
choroid plexus had been removed), stomach, skeletal muscle,
testis and heart. Since 40 cycles of polymerase chain reaction
can usually detect 100 - 1000 target molecules in the starting
materials , the absence of a signal for transthyretin mRNA
suggest that the transthyretin gene is not expressed, or is
expressed at extremely low rates, in these tissues. This con-
firmed previous observations made with other species [3, 5,
81 showing that the choroid plexus is the exclusive site of
transthyretin synthesis in the brain.
Quantification o f transthyretin mRNA
in choroidplexus and liver
The absolute concentrations of transthyretin mRNA in
the choroid plexus and liver of the rat have been measured
previously by ribonuclease protection assay . That analysis
showed much higher concentrations of transthyretin mRNA
in the choroid plexus than in liver, suggesting a special role of
transthyretin gene expression in this organ. Therefore, it was
of interest to measure transthyretin mRNA concentrations in
the choroid plexus and liver from chicken, using the same
The amounts of transthyretin mRNA in chicken lateral
ventricle choroid plexus and liver per wet mass were found
to be 7.2 pg (30 pmol) and 0.33 pg (1.4 pmol), respectively.
Determinations were done in duplicate (for details see Exper-
imental Procedures) and differed from the given values by less
than 2.5%. A gram of RNA from choroid plexus contained
47 times more transthyretin mRNA than a gram of RNA from
liver. These results are similar to those obtained previously for
the rat .
The analysis of the consequences of gene alterations, ex-
perimentally produced or arising during evolution, is a power-
ful approach for the evaluation of the functional importance
of a protein. A drastic alteration is the complete disappearance
of a protein from an organism due to a genetic mutation.
For the three proteins which bind thyroid hormones in the
bloodstream of higher animals, genetically caused absence
is known only for albumin [ 13 - 161 and thyroxine-binding
globulin [17,18]. Surprisingly, in both deficiencies, the individ-
uals affected survive in good health and, in particular, in
an euthyroid state. To our knowledge, a genetically caused
absence of transthyretin has never been observed. This
suggests that transthyretin, in contrast to the other two thy-
roid hormone-binding proteins, has a critical function, the
loss of which leads to the death of the embryo.
It has previously been shown that the transthyretin gene
is very strongly expressed in the choroid plexus in various
mammals [2 - 111 and that the direction of the secretion of
transthyretin synthesized by the choroid plexus was towards
the brain . Based on those results, it was proposed that
transthyretin synthesized by the choroid plexus is involved in
the transport of thyroid hormones from the bloodstream to
the brain . The data presented in this paper demonstrate
high transthyretin mRNA levels also in the chicken choroid
plexus. This suggests that this role of transthyretin appeared
in evolution at least as early as during the stage of the stem
The rates of evolutionary change of the amino acid se-
quence at the surface of transthyretin and around the binding
site for thyroxine in the central channel were calculated from
the data presented in Figs 3 and 4. They are shown in Table 1.
The amino acids lining the central channel of the transthyretin
molecule, which include the binding site for thyroxine, are
almost completely conserved. Compared with the rates of
evolution of the structures of various other plasma proteins,
Table 1. Rates of evolutionary change for different sections ofthe trans-
The derivation of accepted point mutations is described in Experimen-
100 residues in
~ 3 , 2 5 1
a Derived from only one substitution per subunit.
transthyretin is one of the most strongly conserved plasma
The presence of thyroid hormones in the embryo is a
mandatory requirement for normal brain development .
This requirement, the lack of reports of a genetic absence of
transthyretin in the blood, the complete conservation of the
thyroxine-binding site in transthyretin (this paper) and the
high levels of transthyretin mRNA in choroid plexus [2-
111, make it likely that the evolutionarily oldest and most
fundamental function of transthyretin is the transport of thy-
roid hormones from the bloodstream to the brain, whereas
transthyretin synthesis in liver evolved more recently, prob-
ably about or after the appearance of the reptilians .
We thank Drs T. Cole and T. Thomas for technical advice. We
are grateful to Dr C. Blake for supplying a list of surface residues of
human transthyretin. The help of Mrs J. Guest and Mr J. Morley in
preparing the manuscript is also much appreciated. This work was
supported by the National Health and Medical' Research Council of
Australia and the Australian Research Council.
1. Larsson, M., Petterson, T. & Carlstrom, A. (1985) Cen. Comp.
Endocrinol. 58, 360 - 375.
2. Schreiber, G., Aldred, A. R., Jaworowski, A,, Nilsson, C., Achen,
M. G. & Segal, M. B. (1990) Am. J. Physiol. 258, R338-R345.
3. Dickson, P. W., Howlett, G. J. & Schreiber, G. (1985) J. Biol.
4. Fung, W. P., Thomas, T., Dickson, P. W., Aldred, A. R., Milland,
J., Dziadek, M., Power, B., Hudson, P. & Schreiber, G. (1988)
J. Biol. Chem. 263, 480-488.
5. Dickson, P. W., Aldred, A. R., Marley, P. D., Guo-Fen, T.,
Howlett, G. J. & Schreiber, G. (1985) Biochem. Biophys. Res.
6. Dickson, P. W. & Schreiber, G. (1986) Neurosci. Lett. 66, 31 1
7. Dickson, P. W., Aldred, A. R., Marley, P. D., Bannister, D. &
Schreiber, G. (1986) J. Biol. Chem. 261, 3475-3478.
8. Stauder, A. J., Dickson, P. W., Aldred, A. R., Schreiber, G.,
Mendelsohn, F. A. 0. & Hudson, P. (1986) J. Histochem.
687 Download full-text
9. Herbert, J., Wilcox, J. N., Pham, K.-T. C., Fremeau, R. T. Jr,
Zerviani, M., Dwork, A., Soprano, D. R., Makover, A., Good-
man, D. s., Zimmerman, E. A., Roberts, J. L. & Schon, E. A.
(1986) Neurology 36,900-911.
10. Mita, S., Maeda, S., Shimada, K. & Araki, S. (1986) J. Biochem.
(Tokyo) 100, 1215-1222.
11. Tu, G.-F., Cole, T., Southwell, B. R. & Schreiber, G. (1990) Dev.
Brain Res. 55, 203 - 208.
12. Weisner, B. & Rothig, H. J. (1983) Eur. Neurol. 22, 96-105.
13. Bennhold, H., Peters, H. & Roth, E. (1954) Verh. Dtsch. Ges. Inn.
Med. 60,630 - 634.
14. Boman, H., Hermodson, M., Hammond, C. A. & Motulsky, A.
C. (1976) Clin. Genet. 9, 513-526.
15. Dammacco, F., Miglietta, A., D’Addabbo, A., Fratello, A.
Moschetta, R. & Bonomo, L. (1980) Vox Sang. 39, 153-161.
36. Nagase, S., Shimamune, K. & Shumiya, S. (1979) Science 205,
590 - 591.
17. Refetoff, S. (1989) Endocrine Rev. 10, 275-293.
38. Refetoff, S. & Larsen, P. R. (1989) in Endocrinology, 2nd edn
(DeGroot, L. J., ed.) pp. 541-561, W. B. Saunders Co.,
19. Dickson, P. W., Aldred, A. R., Menting, J. G. T., Marley, P. D.,
Sawyer, W. H. & Schreiber, G. (1987) J. Biol. Chem. 262,
13907- I391 5.
20. Kanda, Y., Goodman, D. S., Canfield, R. E. & Morgan, F. J.
(1974) J. Biol. Chem. 249, 6796-6805.
21. Sundelin, J., Melhus, H., Das, S., Eriksson, U., Lind, P.,
Traggrdh, L., Peterson, P. A. & Rask, L. (1985) J. Biol. Chem.
22. Blake, C. C. F., Geisow, M. J., Oatley, S. J., Rerat, B. & Rerat,
C. (1978) J. Mol. Biol. 121, 339-356.
23. Mita, S., Maeda, S., Shimada, K. & Araki, S. (1984) Biochem.
Biophys. Res. Commun. 124, 558 - 564.
24. Wakasugi, S., Maeda, S., Shimada, K., Nakashima, H. & Migita,
S. (1985)J. Biochem. (Tokyo) 98, 1707-1714.
25. Duan, W., Cole, T. & Schreiber, G. (1989) Nucleic Acids Res. 17,
26. Tu, G.-F., Cole, T., Duan, W. & Schreiber, G. (1989) Nucleic
Acids Res. 17, 6384.
27. Tsuzuki, T., Mita, S., Maeda, S., Araki, S. & Shimada, K. (1985)
J. Biol. Chem. 260,12224-12227.
28. Wakasugi, S., Maeda, S. & Shimada, K. (1986) J. Biochem. 100,
29. Margoliash, E. & Smith, E. L. (1965) in Evolving genes andpro-
teins (Bryson, V. & Vogel, H. J., eds) pp. 221 -242, Academic
Press, New York.
30. Dickerson, R. E. (1971) J. Mol. Evol. I, 26-45.
31. Young, J. Z. (1981) The life of the vertebrates, 3rd edn, Clarendon
32. Doolittle, R. F., Feng, D. F., Johnson, M. S. & McClure, M. A.
(1986) Coldspring Harb. Symp. Quant. Biol. 51,447-455.
33. Loomis, W. F. (1990) Comp. Biochem. Physiol. 9SB, 21-27.
34. Tu, G.-F., Achen, M. G., Aldred, A. R., Southwell, B. R. &
Schreiber, G. (1991) J. Biol. Chem. 266,6201 -6208.
35. Omstein, L. (1964) Ann. N. Y. Acad. Sci. 121, 321 -349.
36. Davis, B. J. (1964) Ann. N. Y. Acad. Sci. 121, 404-427.
37. Hewick, R. M., Hunkapillar, M. W., Hood, L. E. & Dreyer, W.
J. (1981) J. Biol. Chem. 256,7990-7997.
38. Zimmerman, C. L., Appella, E. & Pisano, J. J. (1977) Anal.
Biochem. 77, 569-573.
39. Wettenhall, R. E. H., Nick, H. P. & Lithgow, T. (1988) Biochemis-
try 27, 170-177.
40. Itzhaki, R. F. & Gill, D. M. (1964) Anal. Biochem. 9,401 -410.
41. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.
(1951) J. Biol. Chem. 193,265-275.
42. Laemmli, U. K. & Favre, M. (1973) J. Mol. Biol. 80, 575-599.
43. Merril, C. R., Goldman, D., Sedman, S. A. & Ebert, M. H. (1981)
Science 211, 1437-1438.
44. Cole, T., Dickson, P. W., Esnard, F., Averill, S., Risbridger, G.
P., Gauthier, F. & Schreiber, G. (1989) Eur. J. Biochem. 186,
45. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular
cloning: a laboratory manual, 2nd edn, Cold Spring Harbor
Laboratory Press, Cold Spring Harbor NY.
46. Feinberg, A. P. & Vogelstein, B. (1983) Anal. Biochem. 132, 6 -
47. Feinberg, A. P. &Vogelstein, B. (1984) Anal. Biochem. 137,266-
48. McClung, J. K. & Gonzales, R. A. (1989) Anal. Biochem. 177,
378 - 382.
49. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Nut1 Acad.
Sci. USA 74, 5463 - 5461.
50. Thomas, T., Schreiber, G. & Jaworowski, A. (1989) Dev. Biol.
51. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J. & Rutter, W.
J. (1979) Biochemistry 18, 5294- 5299.
52. Lee, J. J. & Costlow, N. A. (1987) Methods Enzymol 152, 633-
53. Krieg, P. A. & Melton, D. A. (1987) Methods Enzymol155,397-
54. Gayler, K. R., Kolivas, S., Macfarlane, A. J., Lilley, G. G., Baldi,
M., Blagrove, R. J. &Johnson, E. D. (1990) Plant Mol. Biol.
55. George, D. G., Barker, W. C. & Hunt, L. T. (1986) Nucleic Acids
Res. 14, 11 -15.
56. Orcutt, B. C., George, D. G., Fredrickson, J. A. & Dayhoff, M.
0. (1982) Nucleic Acids Res. 10, 157-174.
57. Dayhoff, M. O., Barker, W. C. & Hunt, L. T. (1983) Methods
Enzymol. 91, 524-5545,
58. Orcutt, B. C., George, D. G. & Dayhoff, M. 0. (1983) Annu. Rev.
Biophys. Bioeng. 12,419-441.
59. Dayhoff, M. O., ed. (1978) Atlas ofprotein sequence andstructure,
vol. 5, suppl. 3, pp. 375, National Biomedical Research Foun-
dation, Washington DC.
60. Remero-Herrera, A. E., Lehmann, H., Joysey, K. A. & Friday,
A. E. (1973) Nature 246, 389-395.
61. Abe, T., Muto, Y. & Hosoya, N. (3975) J. Lipid Res. 16, 200-
62. von Heijne, G. (1988) Biochim. Biophys. Acta 947, 307-333.
63. Kozak, M. (1984) Nucleic Acids Res. 12, 857-872.
64. Breathnach, R. & Chambon, P. (1981) Annu. Rev. Biochem. SO,
349 - 383.
65. Blake, C. C. F. & Oatley, S. J. (1977) Nature 268, 11 5 - 120.
66. Brunger, A. T. (1988) J. Mol. Biol. 203,803-816.
67. Blake, C. (1983) Trends Biochem. Sci. 8, 11 -13.
68. Hewett-Emmett, D., Venta, P. J. & Tashian, R. E. (1982) in
Macromolecular sequences in Systematic and evolutionary bi-
ology (Goodman, M., ed.) pp. 357-405, Plenum Press, New
69. Innis, M. A. & Gelfand, D. H. (1990) in PCR protocols, a guide
to methods and applications (Innis, M. A,, Gelfand, D. H.,
Sninsky, J. J. & White, T. J., eds) pp. 3- 12, Academic Press,
70. Morreale de Escobar, G. (1986) in Frontiers in thyroidology
(Medeiros-Neto, G. & Gaitan, E., eds) vol. 1, pp. 5-17, Ple-
num, New York and London.
71. Navab, M., Mallia, A. K., Kanda, Y. & Goodman, D. S. (1977)
J. Biol. Chem. 252, 5100-5106.
72. Blake, C. C. F., Geisow, M. J., Swan, I. D. A., Rerat, C. & Rerat,
B. (1974) J. Mol. Biol. 88, 1-12.
73. Bernstein, F. C., Koetzle, T. F., Williams, G. J. B., Meyer, E. F.
Jr, Brice, M. D., Rodgers, J. R., Kennard, O., Shimanouchi,
T. & Tasumi, M. (1977) J. Mol. Biol. 112, 535-542.
74. Harms, P., Tu, G.-F., Richardson, S. J., Aldred, A. R.,
Jaworowski, A. & Schreiber, G. (1 991) Comp. Biochem. Physiol.
99 B, 239 - 249.