| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pituitary |
Department of Biological Sciences,4 Macquarie University, Sydney, New South Wales 2109, Australia
Unité Evolution des Régulations Endocriniennes,5 UMR CNRS-MNHN 5166, 75005, Paris, France
Department of Biology,6 School of Education, Waseda University, Tokyo 169-8050, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
follicle-stimulating hormone, luteinizing hormone, pituitary hormones, thyroid-stimulating hormone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunit and a ß subunit that confers biological specificity to the dimer through interaction with a specific receptor. The common subunit and the three members of the glycoprotein hormone ß subunit family were already present at the emergence of the gnathostomes [1]. After dimerization, they produce either one of the two pituitary gonadotropins, follicle stimulating hormone (FSH) and luteinizing hormone (LH) (previously named GTH1 and GTH2, respectively, in teleosts) or thyroid-stimulating hormone (TSH) [2]. The subunit is quite well conserved throughout the gnathostomes, with identity scores higher than 55% [1]. The subunit of the lungfish Neoceratodus shares 69–84% similarity with tetrapod subunits and 57–74% with that of teleosts [3]. Stability of the subunit is probably due to constraints imposed by its obligation to associate with three different ß subunits in a way that allows binding to and activation of three different target receptors. These structural constraints also apply for each ß subunit but a higher degree of freedom seems to be permitted. Indeed, the evolution rate of the FSHß subunit in teleosts was estimated to be 1.3 times higher than for LHß subunit [4]. If the FSHß subunit evolved more rapidly than the LH in the lineage leading to teleosts, a somewhat reverse situation has been suggested for tetrapods from sequence comparisons [1]. Of significance for these considerations should be the glycoprotein ß subunits produced in the pituitary of lungfish. Current molecular phylogenies favor lungfish as the closest living relative to the ancestors of the tetrapods [5–7]. Data from lungfish should then help give an orientation to sequence modifications observed in the tetrapod lineage. In this paper, we describe the cloning of the three glycoprotein hormone ß subunits from the lungfish Neoceratodus forsteri. These sequences were used with those from other vertebrate groups, both fish and tetrapod, to tentatively estimate the relative evolution rate of glycoprotein hormone ß subunits in these groups.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adult mature (about 4 kg and 90 cm in length) Australian lungfish (N. forsteri) were collected from a tributary of the Mary River in Queensland, Australia. Some were anesthetized and killed by decapitation in the field. The pituitary was removed and kept in RNA later (Ambion Inc., Woolward Austin, TX). Others were transported back to the laboratory in a tank and held in outdoor tanks before being killed. The pituitary glands from these fish were kept in liquid nitrogen until extraction. Animal experimentations were carried out under animal ethics Permit #600, issued by Macquarie University Animal Ethics Committee.
Construction of Pituitary cDNA Library (LHß and TSHß Subunits)
Cloning of FSHß subunit and of LHß and TSHß subunits was performed on two different pituitary libraries. For details on construction of the library for the FSH cloning, see Arai et al. [3].
Pituitary samples were separately sonicated in 20 volumes of 6 M urea, 3 M LiCl, 50 mM NaAc, 200 µg/ml (30 units/ml) heparin (cat #101932; ICN Biomedicals Inc., Costa Mesa, CA). Sodium sarkosinate (Sigma-Aldrich, St. Louis, MO) was then added to 0.1% final volume and the samples were kept at 4°C for 48 h. After centrifugation (30 min at 15 000 x g at 4°C), the pellet was resuspended in 8 M urea, 4 M LiCl, 50 mM NaAc and centrifuged for another 30 min. The pellet was dissolved in water and extracted twice with one volume of phenol/chloroform/isoamyl alcohol (24:24:1 by volume). The aqueous phase was decanted and precipitated at -20°C with three volumes of ethanol after addition of 0.05 volumes of LiCl 8 M. The RNA was recovered by centrifugation (30 min, 15 000 x g, 4°C) and dissolved in water. PolyA+ RNA was isolated using the PolyATract mRNA Isolation System (Promega Corporation, Madison, WI). Synthesis of cDNA was performed using ZAP-cDNA Library Construction Kit (Stratagene, La Jolla, CA) on 1µg polyA+ RNA from the two pooled samples. A one-fifth scaled-down procedure was used for optimization. Size fractionation of cDNAs was performed on Size Sep 400 Spin columns (Amersham Biosciences, Piscataway, NJ). One hundred nanograms of cDNA were directionally ligated to 1 µg lambda Zap II EcoRI-XhoI digested arms for 48 h at 8°C in 5 µl. Two microliters of the ligation were used for packaging. The primary library was titered at 8 x 106 positive plaque forming units (pfu). An aliquot of 1.3 x 106 pfu was plated for amplification that was titered at 1.5 x 107 pfu/µl.
Isolation of cDNA and Sequence Analysis
LHß cDNA A 23-mer nondegenerated oligonucleotide, LH51 (Fig. 1), was designed based on Rana ridibunda LHß cDNA sequence (EMBL AJ311355) in the region around the cleavage site of the signal peptide, and on a 23-amino-acid–long N-terminal peptide sequence of Neoceratodus LHß subunit (Kawauchi and Joss, unpublished). This oligonucleotide was used together with one corresponding to the T7 promoter sequence (melting temperature = 53°C) to amplify the 3' end of LHß cDNA from 1 µl of the amplified library using Hot Start Taq DNA polymerase (Sigma-Aldrich). The following program was used: 15 min at 95°C followed by four cycles with 1 min at 94°C, 1 min at 68°C, 1 min at 72°C, then directly 15 sec at 55°C and 30 sec at 72°C, followed by 34 cycles with 15 sec at 94°C, 15 sec at 55°C, 30 sec at 72°C, and ended with 15 min at 72°C. The 400-base pair (bp) fragment was separated on a 1.5% agarose gel in 1x TAE (40 mM Tris-acetate, 1 mM EDTA) and extracted using MinElute QiaGel extraction kit (Qiagen AS, Oslo, Norway) and subcloned into pGEM-T Easy vector (Promega) and sequenced. An antisense oligonucleotide (LH31, Fig. 1) was designed from this sequence and used in conjunction with M13 reverse primer to amplify the 5' end using the same program as above except that the annealing temperature for the first four cycles was decreased to 65°C. The amplified fragment was gel purified, extracted, and subcloned into pGEM-T Easy vector and sequenced. Sequencing was performed on ABI Prism 377 DNA sequencer (Perkin Elmer Applied Biosystem, Foster City, CA).
|
TSHß cDNA Two sense (TSH51) and antisense (TSH31), nondegenerated oligonucleotide primers were designed based on conserved regions of TSHß subunits to amplify a 243-bp fragment. This fragment was gel purified, extracted, and sequenced. From this sequence, two specific sets of primers were designed. TSH52 and TSH32 were used with primers corresponding with M13 forward and reverse primers, respectively, to amplify the 5' and 3' ends using the same program as for LHß, with annealing temperatures of 65 and 62°C, respectively, for the first four cycles. Nested amplifications were performed on purified fragments using TSH53 and TSH33 primers and primers corresponding with T3 and T7 promoters, respectively. Resulting amplified fragments were subcloned into pGEM-T Easy and sequenced. Sequencing was performed on a CEQ 2000 sequencer (Beckman Coulter Instruments, Fullerton, CA).
FSHß cDNA
Total RNA was extracted from a single pituitary gland with a commercial kit, ISOGEN (Nippon Gene, Tokyo, Japan). A reverse transcription-PCR was performed using FSH51 and FSH31 as sense and antisense primers, respectively. A 200-bp fragment was obtained and used as a probe to screen the library. The amplified Australian lungfish pituitary cDNA library constructed by Arai et al. [3] was screened by a plaque hybridization method. The probe was labeled with [-32P]-dCTP (AA0005; Amersham, South Clearbrook, IL) by the random priming method using the Rediprime II DNA Labelling System (Amersham). Hybridization was performed overnight at 55°C in 6x SSC (900 mM NaCl, 90 mM sodium citrate), 0.1% SDS, 1x Denhardt reagent (0.02% ficoll, 0.02% bovine serum albumin, 0.02% polyvinylpyrrolidone). Filters were washed twice with 1x SSC containing 0.1% SDS for 20 min at 60°C. Hybridization signals on the membranes were analyzed on the BAS-2000II Bio-Imaging Analyzer (Fuji Film, Tokyo, Japan). Four positive clones were obtained and subcloned into the EcoRI site of the pBluescript plasmid vector for the sequence analysis. Sequencing was performed on the DNA sequencer model 4000L (LI-COR, Lincoln, NE) with Thermo Sequenase Cycle Sequencing Kit (Amersham). Nucleotide sequences of the four positive clones were identical.
Phylogenetic Analysis
Phylogenetic analysis was performed using PAUP 3 (Phylogenetic Analysis Using Parsimony, version 3.1.1; Sinauer Associates, Sunderland, MA). In addition to those of Neoceratodus, amino acid sequences of the three glycoprotein hormone ß subunits from representatives of the diversity of vertebrates (Table 1) were aligned (Fig. 2). Only the portion between conserved cysteine residues numbers 1 and 12 were considered (42 sequences, 107 characters, out of which 17 were parsimony constant and 3 were parsimony uninformative). A heuristic search was performed under the constraint of a backbone tree in which the relationships between the different groups (birds, mammals, amphibians, etc.) were imposed for each subunit cluster. All TSHß subunits were used as a monophyletic outgroup. All characters were given the same weight and were unordered. Two slightly different but equally parsimonious, unrooted trees were obtained. The one giving the same topology as the species true tree for the three subunit clusters was retained. The cumulated assigned branch lengths (from the subunit cluster basal node to the terminal taxon) were computed with accelerated transformation settings for character-state optimization. Accelerated transformation settings were chosen over those with deleted transformations because the trifurcation of subunit clusters occurred long before the radiation of the considered vertebrate groups did and also because one can assume that most of the major transformations should have occurred before the species within each vertebrate group radiated. In any case, results obtained using deleted transformation settings gave rise to very similar conclusions in terms of cumulated branch lengths.
|
|
Identity scores were obtained from the pairwise amino acid differences computed from the distance settings using PAUP 4 (Phylogenetic Analysis Using Parsimony [*and Other Methods], version 4.0b10; Sinauer Associates).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The sequence of the 476-bp full-length cDNA for LHß subunit (EMBL AJ578038) was reconstructed from the two overlapping 5' and 3' cDNA ends. It encoded a 131 amino acid sequence (Fig. 2), including a signal peptide of 19 amino acid residues. The position of the cleavage site was obtained from the N-terminal peptide sequence (Kawauchi and Joss, unpublished), which started with Arg His Met Cys. The same position was used in Rana catesbeiana, giving a peptide with Arg His Val Cys at its N-terminal end [8]. The 3' end was rather short, including a polyadenylation signal site located 17 nucleotides before the poly(A) tail. Comparison with LHß subunits from different vertebrate representatives showed that the lungfish peptide exhibited most of their features (Fig. 2). Neoceratodus LHß sequence notably shared the proline residue in position 16 (according to numbering in Fig. 2) with all other LHß subunits except that of Fundulus heteroclitus ([9], EMBL M87015) and choriogonadotropin (CG) ß subunits. No proline residue was present at that position in any FSH or TSHß subunit sequence. However, three amino acids in the Neoceratodus LHß subunit had physicochemical properties quite different from those conserved in the corresponding position of other LHß subunits: alanine 55 (numbered according to Fig. 2), a nonpolar amino acid residue, in place of the negatively charged glutamate or aspartate; lysine 56 (positively charged), in place of isoleucine, leucine, or valine (aliphatics); and alanine 113, in place of a proline that tends to bend the polypeptide chain. Highest identity scores (Fig. 3) were obtained with the shark Scyliorhinus sequence (73% identity between the two external conserved cystein residues of the mature peptides) and amphibian sequences (62–72%). The lowest identity scores were observed with mammalian (41–44%) and avian (48–49%) sequences, whereas those of teleosts (60%) and the chondrostean Acipenser (59%) were in between.
|
FSHß Subunit
There were three potential translation start sites at the 5' end of the 608-bp-long FSHß cDNA (EMBL AJ578040) in positions 51 (txxATGc), 75 (cxxATGg), and 102 (axxATGt). The three of them were equally probable according to ATGPr (www.hri.co.jp/atgpr/). When compared with amphibian sequences Bufo japonicus (EMBL AB085668, [10]) and Cynops pyrrogaster (EMBL AB067752 [11]), the third one is the most likely in that it encodes a signal peptide of approximately the same size (19–20 amino acids) and the sequence 5' to this position is not conserved among them. By analogy with LHs, the position of the signal peptide cleavage site has been tentatively located as for R. catesbeiana FSH [12], just before the cystein residue number 1 (Fig. 2). Most of the amino acid residues usually conserved among FSHs were also present with similar physicochemical properties in the Neoceratodus ß subunit except for isoleucine 52 and particularly histidine 88, a positively charged residue in place of a negatively charged one. The lungfish FSH was quite similar to all tetrapod FSHs, with identity scores ranging from 59% to 66%. When compared with actinopterygian sequences, the identity score was lower (from 52% to 40%) with the evolutionary distance (Fig. 3).
TSHß Subunit
Six potential translation start sites, all in the same reading frame, were located at the 5' end of the 794-bp-long lungfish TSHß subunit cDNA (EMBL AJ578039) at positions 30 (txxATGg), 69 (cxxATGt), 177 (cxxATGt), 183 (txxATGa), 189 (axxATGa), and 201 (axxATGa). The first of them potentially initiates the translation of a short peptide (MGGYFCCVRALFLMFNIstop) comprising the second ATG. A BLAST search (www.ebi.ac.uk/blast2/) with this sequence against EMBL data bank did not yield any hits. The signal peptide might start from any of the four remaining ATG codons. The one in position 201 would give a signal peptide of 20–21 amino acids in length (Fig. 1), provided that the mature peptide is cleaved just before the first conserved cysteine residue as it is in most tetrapod species. When compared with the other two lungfish ß subunits, the TSHß subunit appeared as the most divergent (Fig. 2). In eight positions that are usually quite conserved among TSHß subunits or even among all ß subunits, the lungfish TSHß subunit presented amino acid residues with unique physicochemical properties. The highest overall identity score was only 55% as compared with 72% and 66% for LH and FSHß subunits (Fig. 3).
Sequence alignment (Fig. 2) also showed that Neoceratodus TSH had 41% amino acid residues conserved in position with Neoceratodus LH and 43% with Neoceratodus FSH, when FSH and LH shared up to 51% of them.
Phylogenetic Analysis
The relative evolution rate of the three subunits was deduced from a maximum parsimony tree constructed from the amino acid sequences between the cysteine residues numbers 1 and 12 (Fig. 2) to avoid saturations and under the constraint of a backbone tree in order to force the topology of the tree for each subunit cluster to map the real phylogenetic relationships between vertebrate groups (Fig. 4, bold lines). Each assigned branch length represents the number of steps (amino acid substitutions) that differentiates the sequences between two linked nodes. Given that the topology of the tree corresponds to the most parsimonious predictable one, i.e., the one that needs the minimum number of steps of all possible trees, the cumulated branch length between the basal node of a subunit cluster and a terminal species thus represents the number of accumulated substitutions in the sequence from the ancestral subunit up to the considered species. Given that each position may have changed several times during evolution, this cumulated number of substitutions may be higher than the actual number of variable amino acids in the sequence. This model can be reliable only when the topology of the subunit tree reflects the real phylogenetic relationships between the species; hence, the backbone constraint. Figure 5 reports the cumulated branch lengths of the three subunits inferred for each species within each vertebrate group. The average branch lengths for each vertebrate group presented increasing values according to the time of divergence of the groups relative to the emergence of gnathostomes. It showed that saturation events (numerous substitutions at a given character position) had no dramatic influence on the branch length inferences within the sequence window used in the study.
|
|
The cumulated branch length for Neoceratodus TSH was slightly larger than that for FSH and a little less than twice that of LH. It showed that TSH and FSHß subunits evolved more rapidly (higher number of amino acid substitutions) in the lineage leading to Neoceratodus than the LHß subunit did.
Scyliorhinus FSH branch length was more than twice the length of LH, indicating a rapid evolution of FSH in the Selacian lineage. There were no marked differences in the branch lengths between the three subunits in the Chondrostean Acipenser.
LHß was the most diverged subunit in mammals with branch length values about twice those for TSH. Moreover, LHß subunit branch lengths varied according to their phylogenetic distances: 70 for the marsupial Monodelphis, 83 for the rodent Mus, and 90 for human (Fig. 4). Such a relationship was not true for mammalian FSHs and TSHs. LH was also the most diverged subunit in birds and TSH the least. In the lineage leading to teleosts, both FSH and TSH had diverged more rapidly than LH. Branch lengths for FSH and to a lesser extent LH subunits showed a marked progression with phylogenetic distance. In amphibians, LH was the most conserved ß subunit. A clear relationship between branch lengths and phylogenetic distances was nevertheless observed for this subunit in this group.
These results drawn from a parsimony-based tree were consistent with data from distance comparisons (Fig. 3). As an example, identity scores within mammalian LH were 67%, 71%, 80% (mean 72.5%) against 83% (mean value) for FSH and 85.5% for TSH for the same three species, indicating more divergence for LH than for FSH and TSH subunits. For teleosts, FSHß subunits had identity scores of 54.5% against 66.5% for TSH and 82.5% for LH, indicating that FSH is the most divergent subunit in this group.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Comparison of identity scores between subunits might be quite confusing. Neoceratodus LHß subunit is closer to amphibian or shark sequences than to those of any other vertebrate group. This is in agreement with its phylogenetic position, close to the root of tetrapods. More confusing are the relationships between FSHß subunits, where bird sequences show the closest relationship, although those from amphibians are also close. For TSHß subunits, the highest identity scores were obtained with mammalian and bird sequences whereas those with amphibian sequences were as low as with teleost sequences. A somewhat similar result was obtained for the subunit [3]. These results show that identity scores for these subunits are not strictly correlated to phylogenetic distances. Put in other words, it shows that the rate of evolution of these subunits is not constant and may differ from each other.
The relative evolution rate between subunits was inferred from a maximum parsimony tree. Both distance and parsimony methods might have been used. But because the number of informative sites was quite high in our data set (87 informative characters out of 107), the parsimony (or character-state) method was preferred. Results obtained from this method were consistent with data obtained from pairwise distance comparisons (Fig. 5 versus Fig. 3). Even with the sequence window used in this study (excluding signal peptides and carboxy terminal ends that are far too variable between and even within subunit lineages), a number of positions presented saturations and thus lead to distortions in the topology of the unconstrained tree. A backbone constraint was then used to force the topology of the tree to restore the monophyly of the three subunit clusters as determined previously [2] and to map the real phylogenetic relationships between vertebrate groups within each subunit cluster. The cluster of TSHß subunits was used as the outgroup because TSH lineage is believed to have individualized before the split between the two gonadotropin lineages occurred [1]. The maximum parsimony tree obtained should then reflect as closely as possible the real evolutionary history of the sequences.
Strangely, the branch lengths for Neoceratodus FSH and TSHß subunits were not as different as expected from sequence alignment analysis. This might be due to the absence of Scyliorhinus TSHß subunit sequence data, leading to a possible underestimation of the lengths of the branches close to the basal node in the TSH cluster.
The cumulated number of amino acid substitutions for LHß subunits in the lineage leading to mammals and birds was higher than for FSH or TSH. In contrast, the LHß subunit was the least diverged one for amphibians, Neoceratodus, Scyliorhinus, and teleosts. It shows that a burst in evolution rate occurred in the LHß subunit sometime after the emergence of amphibians. The species representation within the bird group is too phylogenetically close to draw any conclusion about the dynamics of evolution within this group. However, for mammals, it seems clear that the rate of evolution of the LHß subunit was high throughout the radiation of the group. It is remarkable also that a duplication of the LHß subunit gene occurred in mammals precisely, leading to the appearance of a chorionic gonadotropin in primates [17]. In contrast with LH, no correlation was observed between cumulated branch lengths and phylogenetic distances for mammalian FSH and TSH, which are very stable within the sequence window used in this study. One can wonder what would have caused such a burst in evolution of LH. Is it a common feature to amniotes? The answer will be known when data from different reptile groups become available. Could it be linked to homeothermy? In this case, FSH and TSHß subunits would certainly be also affected. Does the mammalian burst in LH evolution result from lower or higher constraints imposed on LH function? Lower constraints would allow more flexibility in the hormone structure provided some function is still conserved. On the other hand, a better specificity of action of LH with regard to the other gonadotropin, FSH, would likely be met by increasing differences in LH to prevent it from binding to the FSH receptor. In support of this hypothesis, it has been shown that, while LH is able to displace FSH binding to its receptor in the amphibian R. catesbeiana [18], there is a strict specificity of binding in birds [19] and mammals [20].
Such dynamics in the evolutionary rate has also been shown for LH and FSHß subunits in teleosts (this study and [4]). This is also the case for the LHß subunit in amphibians, even though it is the most stable subunit in this group. More sequence data are needed to evaluate the dynamics of the evolutionary rate for FSH and TSHß subunits in amphibians.
In conclusion, sequencing of the three ß subunits of the pituitary glycoprotein hormones from the lungfish, Neoceratodus forsteri, has provided a further refinement to our understanding of the evolution of this family of pituitary hormones. We were able to demonstrate that the rates of evolution for the three glycoprotein hormone ß subunits differ from each other and are not constant throughout the radiation of vertebrates. LH exhibited a marked acceleration in its rate of evolution sometime after the emergence of amphibians in the tetrapod lineage, whereas FSH did the same sometime after the emergence of chondrosteans in the lineage leading to teleosts.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Bruno Querat, Unité Evolution des Régulations Endocriniennes, UMR CNRS-MNHN 5166, 7 rue Cuvier, 75231 Paris Cedex 05, France. FAX: 33 1 4079 3620; querat{at}mnhn.fr
3 Current address: Department of Anatomy, Kitasato University, School of Medicine, Sagamihara, Kanagawa 228-8555, Japan
Received: 6 August 2003.
First decision: 29 August 2003.
Accepted: 4 September 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, FSHß and LHß) and their expression during the reproductive cycle. Gen Comp Endocrinol 2003 13:216-232This article has been cited by other articles:
|
|
 |
|
  M. Freamat, H. Kawauchi, M. Nozaki, and S. A Sower Identification and cloning of a glycoprotein hormone receptor from sea lamprey, Petromyzon marinus. J. Mol. Endocrinol., August 1, 2006; 37(1): 135 - 146. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
