Expression, purification, and characterization of human hemoglobins Gower-1 (zeta(2)epsilon(2)), Gower-2 (alpha(2)epsilon(2)), and Portland-2 (zeta(2)beta(2)) assembled in complex transgenic-knockout mice

Embryonic zeta- and epsilon-globin subunits assemble with each other and with adult alpha- and beta-globin subunits into hemoglobin heterotetramers in both primitive and definitive erythrocytes. The properties of these hemoglobins-Hbs Gower-1 (zeta(2)epsilon(2)), Gower-2 (alpha(2)epsilon(2)), and Portland-2 (zeta(2)beta(2))-have been incompletely described as they are difficult to obtain in quantity from either primary human tissue or conventional expression systems. The generation of complex transgenic-knockout mice that express these hemoglobins at levels between 24% and 70% is described, as are efficient methods for their purification from mouse hemolysates. Key physiological characteristics-including P(50), Hill coefficient, Bohr effect, and affinity for 2,3-BPG-were established for each of the 3 human hemoglobins. The stability of each hemoglobin in the face of mechanical, thermal, and chemical stresses was also determined. Analyses indicate that the zeta-for-alpha exchange distinguishing Hb Portland-2 and Hb A alters hemoglobin O(2)-transport capacity by increasing its P(50) and decreasing its Bohr effect. By comparison, the epsilon-for-beta exchange distinguishing Hb Gower-2 and Hb A has little impact on these same functional parameters. Hb Gower-1, assembled entirely from embryonic subunits, displays an elevated P(50) level, a reduced Bohr effect, and increased 2,3-BPG binding compared to Hb A. The data support the hypothesis that Hb Gower-2, assembled from reactivated epsilon globin in individuals with defined hemoglobinopathies and thalassemias, would serve as a physiologically acceptable substitute for deficient or dysfunctional Hb A. In addition, the unexpected properties of Hb Gower-1 call into question a common hypothesis for its primary role in embryonic development.     

Proteasome implication in phorbol ester- and GnRH-induced selective down-regulation of PKC (alpha, epsilon, zeta) in alpha T(3)-1 and L beta T(2) gonadotrope cell lines

We investigated mechanisms underlying selective down-modulation of PKC isoforms (alpha, epsilon, zeta): 1) during 12-O-tetradecanoyl-phorbol-13 acetate (TPA) (10(-7) M) or GnRH (10(-7) M) desensitization conditions (2- to 6-h treatments) in two gonadotrope cell lines (alpha T(3)-1, L beta T(2)) and 2) in primary pituitary cell cultures from male rats during long-term phorbol ester administration. We demonstrated that, as in alpha T(3)-1 cells, in a more differentiated gonadotrope cell line L beta T(2) the GnRH-receptor coupling (PLC, PLA2, PLD) generated second messengers essential for PKCs activation; the characterized isoforms (alpha, beta II, delta, epsilon, zeta) were selectively and differentially down-regulated by TPA (alpha, beta II, delta, epsilon) or GnRH (delta, epsilon). In whole cell lysates, proteasome inhibitors (proteasome inhibitor I and II, Lactacystin, beta-Lactone, Calpain inhibitor I) prevented in both gonadotrope cell lines the TPA-induced depletion of PKC alpha, epsilon, and the GnRH-elicited PKC epsilon down-regulation; they counteracted in mixed pituitary cell cultures as well, the TPA-evoked PKC alpha, epsilon depletion. In contrast, the inhibitors of calpain(s) and lysosomal proteases (Calpeptin, E64d, Calpain inhibitor II, and PD150606), were ineffective. As shown in alpha T(3)-1 subcellular fractions, proteasome abrogation did not affect membrane translocation of TPA- and GnRH- target isoforms (alpha, epsilon) but, preventing their degradation, favored enzyme accumulation to the membrane compartment. Proteolysis processing of PKCs may be dependent upon their phosphorylated state and/or catalytic activity. Inhibition of PKC catalytic activity (GF109203X, G?6976), selectively prevented the TPA-evoked PKC alpha depletion in both mixed pituitary cells and alpha T(3)-1 gonadotropes; in alpha T(3)-1 subcellular fractions, PKC alpha inactivation overcame the TPA-evoked isoenzyme degradation by inducing a pronounced membrane accumulation of the isoform without affecting its membrane relocalization. Thus, the proteasome system by adjusting PKC cellular levels, may represent a regulatory proteolytic pathway implicated in the adaptive mechanisms of the time dependent cell responses.     

Dissection of the voltage-activated potassium outward currents in adult mouse ventricular myocytes: I(to,f), I(to,s), I(K,slow1), I(K,slow2), and I(ss)

Voltage-activated outward K⁺ currents (I _Kv) are essential for cardiac repolarization and are major factors in the electrophysiological remodeling and arrhythmias seen in heart disease. Mouse models have been useful for understanding cardiac electrophysiology. However, previous methods for separating and quantifying the components of I _Kv in mouse myocardium have yielded inconsistencies. In this study, we developed a statistically rigorous method to uniquely quantify various I _Kv in adult mouse ventricular myocytes, and concluded that tri-exponential functions combined with depolarizing pulses of duration greater than 20 s are essential to adequately separate the different I _Kv components. This method enabled us to reliably dissect the kinetic components of the decay phase of I _Kv into fast (I _to), intermediate (K_V1.5-encoded I _K,slow1) and slow (K_V2-encoded I _K,slow2) components. The most rapid kinetic phase, I _to, can be further dissected into fast (K_V4-encoded I _to,f) and slow (K_V1.4-encoded I _to,s) components by measuring recovery from inactivation, voltage-dependence of activation and sensitivity to HpTx-2 and 4-AP. The applicability of our dissection method was validated using transgenic mice over-expressing dominant-negative K_V1.1 transgene which largely abolished the 4-AP-sensitive portion of I _to (i.e., I _to,s) and the I _K,slow1 component. We also applied our method to Irx5-deficient mice and verified selective elevations of I _to in endocardial myocytes. Our method should prove useful in future electrophysiological studies using mouse.     

The VLA-2 (alpha 2 beta 1) I domain functions as a ligand-specific recognition sequence for endothelial cell attachment and spreading: molecular and functional characterization

The integrin VLA-2 (alpha 2 beta 1), generally considered to represent the specific collagen receptor on human endothelial cells, contains an alpha 2-subunit inserted I domain with structural similarity to the type A domains found within the recently described superfamily of receptor-ligand recognition proteins. This region of the cDNA has now been isolated and used for molecular and functional characterization of this heterodimeric receptor complex. Comparative sequence analysis with the porcine homologue revealed 93% amino acid sequence identity, suggestive of a developmentally conserved function. To complete structure/function studies, this region of the human cDNA was expressed as a chimeric protein in Escherichia coli, and a rabbit polyclonal antibody (anti-I domain) was used to study determinants of endothelial cell attachment and spreading in vitro. Quantifiable and visual disruption of endothelial cell attachment to gelatin, type I collagen, and laminin was evident using the specific anti-I domain antibody, with minimal inhibitory effects demonstrable using fibronectin or fibrinogen matrices. Therefore, these data would suggest that the alpha 2 beta 1 I domain confers ligand-binding specificity for both known alpha 2 beta 1 substrates (laminin and collagen), and that this region subserves a regulatory function in the molecular processes controlling endothelial cell attachment and spreading in vitro.     

Isolation and characterization of cDNAs encoding the heavy chain of human inter-alpha-trypsin inhibitor (I alpha TI): unambiguous evidence for multipolypeptide chain structure of I alpha TI.

Human inter-alpha-trypsin inhibitor (I alpha TI) is a plasma glycoprotein of Mr 180,000, which has been described as a single polypeptide chain. Recently, however, we proposed that I alpha TI might be composed of a heavy (H) chain (Mr = 95,000) and a light (L) chain (Mr = 40,000) synthesized by two separate mRNAs. In the present study we have characterized cDNAs for the H chain of I alpha TI. These cDNAs collectively covered two sequences (977 and 1450 base pairs in length) with single open reading frames. The deduced amino acid sequences were highly homologous to each other and well matched with partial amino acid sequences obtained from purified serum I alpha TI. RNA blot analyses of liver RNAs with H- or L-chain cDNAs as probes clearly identified two distinct mRNAs of 3.3 and 1.3 kilobases, which corresponded to H or L chain, respectively. Poly(A)+ RNAs hybrid-selected with H-chain cDNAs coded for polypeptide chains of Mr 90,000-95,000. These results unambiguously establish that I alpha TI is made of multipolypeptides, possibly including one H and two L chains. The H chain contains potential calcium-binding sites and also regions homologous to the proposed reactive site for thiol-proteinase inhibitors. These data indicate that I alpha TI is a complex, multifunctional protein. mRNAs for both the H and L chains were found only in liver.     

Transcutaneous measurement of oxygen and carbon dioxide tension (TcPO2 and TcPCO2) during treadmill exercise in patients with arterial occlusive disease (AOD)--stages I and II

Transcutaneous oxygen and carbon dioxide tensions were measured during treadmill exercise in 30 patients with stages I and II arterial occlusive disease and 28 normal controls. Measuring sites were the dorsum of the foot and the frontal thorax. The calculated oxygen index at maximum exercise/rest (foot/thorax) differentiated with 100% sensitivity and specifity between the two groups. The oxygen index was the only parameter with a significant positive correlation with the pain-free walking distance, whereas, eg, oscillography and Doppler analysis did not correlate. Carbon dioxide tensions showed significant differences only in the recovery time between patients and normal controls.     

Single base mutation in the pro alpha 2(I) collagen gene that causes efficient splicing of RNA from exon 27 to exon 29 and synthesis of a shortened but in-frame pro alpha 2(I) chain.

Previous observations demonstrated that a lethal variant of osteogenesis imperfecta had two altered alleles for pro alpha 2(I) chains of type I procollagen. One mutation produced a nonfunctioning allele in that there was synthesis of mRNA but no detectable synthesis of pro alpha 2(I) chains from the allele. The mutation in the other allele caused synthesis of shortened pro alpha 2(I) chains that lacked most or all of the 18 amino acids encoded by exon 28. Subclones of the pro alpha 2(I) gene were prepared from the proband's DNA and the DNA sequence was determined for a 582-base-pair (bp) region that extended from the last 30 bp of intervening sequence 26 to the first 26 bp of intervening sequence 29. Data from six independent subclones demonstrated that all had the same sequence as a previously isolated normal clone for the pro alpha 2(I) gene except that four subclones had a single base mutation at the 3' end of intervening sequence 27. The mutation was a substitution of guanine for adenine that changed the universal consensus sequence for the 3' splicing site of RNA from -AG- to -GG-. S1 nuclease experiments demonstrated that about half the pro alpha 2(I) mRNA in the proband's fibroblasts was abnormally spliced and that the major species of abnormal pro alpha 2(I) mRNA was completely spliced from the last codon of exon 27 to the first codon of exon 29. The mutation is apparently unique among RNA splicing mutations of mammalian systems in producing a shortened polypeptide chain that is in-frame in terms of coding sequences, that is used in the subunit assembly of a protein, and that contributes to a lethal phenotype.     

Hereditary pyropoikilocytosis and elliptocytosis in a white French family with the spectrin alpha I/74 variant related to a CGT to CAT codon change (Arg to His) at position 22 of the spectrin alpha I domain

We describe a white French family in which 12 subjects presented with hereditary elliptocytosis (HE) or hereditary pyropoikilocytosis (HPP). Eight of these subjects were shown to be heterozygous for a spectrin (Sp) alpha I/74 variant, as demonstrated by analysis of partial tryptic digestion fragments of spectrin. This abnormal peptide pattern was associated with a decreased ability of Sp dimers to self-associate. In this kindred, in which four generations were available for study, the clinical expression varied from mild HE to HPP with an intermediate status of hemolytic HE. The severity of the disease appeared to be correlated both with the estimated amount of variant Sp (42% to 65%) and the excess of Sp dimers found in the membrane (30% to 51%, with a normal value of 3.7% +/- 1.6%). Reassociation studies using isolated Sp alpha and beta chains from an affected patient and an unaffected control subject showed that the Sp alpha I/74 Kd abnormal tryptic peptide resulted from a defect in the Sp alpha chain. Partial amino acid sequencing showed that the Sp alpha I/74 Kd peptide resulted from cleavage at lysine residue 42 of the Sp alpha I/80 Kd domain. Knowledge of the exon/intron organization of the human alpha Sp gene allowed us to amplify by the polymerase chain reaction the second exon of the alpha Sp gene in total cellular DNA of the HPP proposita. The amplified fragment was subcloned and sequenced. We found a G to A base substitution in the 22nd codon (CAT for CGT), which changes the normal arginine to a histidine. Hybridization of amplified DNAs with allele-specific oligonucleotides corresponding to the normal and mutant sequences confirmed the presence of the mutation in six other HE and HPP members of the family. The identification of this mutation at the DNA level confirmed the transmission of the same molecular defect in Sp through four generations but with different patterns of clinical expression.     

Activation of the platelet collagen receptor integrin alpha(2)beta(1): its mechanism and participation in the physiological functions of platelets

When platelets are stimulated by agonists, integrin alpha(2)beta(1) (GP Ia/IIa), one of the platelet collagen receptors, is activated to forms with high affinities for its ligand collagen. Here we describe our studies to characterize the binding kinetics of the activated integrin forms and the activation mechanism. Under low agonist concentrations, integrin alpha(2)beta(1) is activated through a mechanism involving ADP/ADP receptors; and under high agonist concentrations, multiple signaling pathways are involved in its activation. Such differences in mechanism at low and high agonist concentrations are also suggested in the activation of integrin alpha(IIb)beta(3), the platelet fibrinogen receptor. We describe our flow adhesion studies, from which evidence was obtained about the involvement of integrin alpha(2)beta(1) activation in the physiological function of platelets, adhesion and thrombus formation.     

Skeletal muscle PGF(2)(alpha) and PGE(2) in response to eccentric resistance exercise: influence of ibuprofen acetaminophen

PGs have been shown to modulate skeletal muscle protein metabolism as well as inflammation and pain. In nonskeletal muscle tissues, the over the counter analgesic drugs ibuprofen and acetaminophen function through suppression of PG synthesis. We previously reported that ibuprofen and acetaminophen inhibit the normal increase in skeletal muscle protein synthesis after high intensity eccentric resistance exercise. The current study examined skeletal muscle PG levels in the same subjects to further investigate the mechanisms of action of these drugs in exercised skeletal muscle. Twenty-four males (25 +/- 3 yr) were assigned to 3 groups that received the maximal over the counter dose of ibuprofen (1200 mg/d), acetaminophen (4000 mg/d), or a placebo after 10-14 sets of 10 eccentric repetitions at 120% of concentric 1 repetition maximum using the knee extensors. Preexercise and 24 h postexercise biopsies of the vastus lateralis revealed that the exercise-induced change in PGF(2alpha) in the placebo group (77%) was significantly different (P < 0.05) from those in the ibuprofen (-1%) and acetaminophen (-14%) groups. However, the exercise-induced change in PGE(2) in the placebo group (64%) was only significantly different (P < 0.05) from that in the acetaminophen group (-16%). The exercise-induced changes in PGF(2alpha) and PGE(2) were not different between the ibuprofen and acetaminophen groups. These results suggest that ibuprofen and acetaminophen have a comparable effect on suppressing the normal increase in PGF(2alpha) in human skeletal muscle after eccentric resistance exercise, which may profoundly influence the anabolic response of muscle to this form of exercise.     

Involvement of protein kinase C alpha (PKC alpha) in the early action of angiotensin II type 2 (AT2) effects on neurite outgrowth in NG108-15 cells: AT2-receptor inhibits PKC alpha and p21ras activity

The aim of the present study was to investigate whether protein kinase C (PKC) isoforms may be among the putative candidates implicated in the primary effects of the Ang II type 2 (AT2) receptor. Western blot analyses revealed the presence of PKC alpha,epsilon, iota, and zeta in NG108-15 cells. After a 3-d treatment with 3 nm G?6976, a specific inhibitor of classical PKC isoforms, cells were characterized by the presence of one elongated process similar to that observed after treatment with Ang II or with CGP42112, a selective AT2 receptor agonist. Similar findings were observed in cells expressing a dominant-negative mutant of PKC alpha (K368A). Inhibition of PKC alpha in NG108-15 cells also decreased cell number and proliferation. In conditions of acute stimulation, Ang II induced a time-dependent and transient inhibition of PKC alpha activity, as well as a decrease in PKC alpha levels associated with the membrane. Treatment of cells with G?6976 was also found to inhibit p21(ras) (between 1-10 min) but stimulated Rap1 activity (1-5 min) in a time-course similar to that of Ang II. Incubation of NG108-15 cells with G?6976 (3 nm) inhibited basal p42/p44(mapk) phosphorylation, but failed to interfere with its activation by the AT(2) receptor, indicating that inhibition of PKC alpha is not directly involved in the Rap1-MEK-p42/p44(mapk) cascade. Taken together, these results indicate that PKC alpha is a primary target of the AT2 receptor. Inhibition of PKC alpha leads to a decrease in both p21(ras) activity and cell proliferation, which may facilitate AT2 receptor signaling through p42/p44(mapk), thereby leading to neurite outgrowth.     

Electrolyte Tuning in Iron(II)-Based Dye-Sensitized Solar Cells: Different Ionic Liquids and I2 Concentrations

The effects of different I₂ concentrations and different ionic liquids (ILs) in the electrolyte on the performances of dye-sensitized solar cells (DSCs) containing an iron(II) N-heterocyclic carbene dye and containing the I^–/I₃^– redox shuttle have been investigated. Either no I₂ was added to the electrolyte, or the initial I₂ concentrations were 0.02, 0.05, 0.10, and 0.20 M. The short-circuit current density (J_SC), open-circuit voltage (V_OC), and the fill factor (ff) were influenced by changes in the I₂ concentration for all the ILs. For 1-hexyl-3-methylimidazole iodide (HMII), low V_OC and low ff values led to poor DSC performances. Electrochemical impedance spectroscopy (EIS) showed the causes to be increased electrolyte diffusion resistance and charge transfer resistance at the counter electrode. DSCs containing 1,3-dimethylimidazole iodide (DMII) and 1-ethyl-3-methylimidazole iodide (EMII) showed the highest J_SC values when 0.10 M I₂ was present initially. Short alkyl substituents (Me and Et) were more beneficial than longer chains. The lowest values of the transport resistance in the photoanode semiconductor were found for DMII, EMII, and 1-propyl-2,3-dimethylimidazole iodide (PDMII) when no I₂ was added to the initial electrolyte, or when [I₂] was less than 0.05 M. Higher [I₂] led to decreases in the diffusion resistance in the electrolyte and the counter electrode resistance. The electron lifetime and diffusion length depended upon the [I₂]. Overall, DMII was the most beneficial IL. A combination of DMII and 0.1 M I₂ in the electrolyte produced the best performing DSCs with an average maximum photoconversion efficiency of 0.65% for a series of fully-masked cells.     

New alpha 2 globin chain variant with low oxygen affinity affecting the N-terminal residue and leading to N-acetylation [Hb Lyon-Bron alpha 1(NA1)Val --> Ac-Ala

Hemoglobin Lyon-Bron was found in two members of a family of German ascent presenting with a moderate normocytic anemia. In this alpha 2 globin variant, the N-terminal valine of the chain was replaced by an alanine. Electrospray mass spectrometry of the alpha chain showed that, as normally, the initiator methionine was cleaved during globin processing but that the N alpha-terminal group was totally acetylated. This resulted in structural modifications of a region crucial for oxygen binding. As a consequence, hemoglobin Lyon-Bron displayed both a reduced chloride effect and a decreased oxygen affinity, this last point explaining the apparent anemia.     

A cytochrome P450 immunochemically related to P450c,d (P450I) localized to the smooth microsomes and inner zone of the guinea pig adrenal

In addition to their capacity for steroid synthesis, guinea pig adrenal microsomes have a well documented ability to metabolize foreign compounds. The capacity for metabolism of foreign compounds is localized to the smooth endoplasmic reticulum-filled cells of the inner zone. However, it has not been clear whether they possess cytochrome P450(s) specific for this function, distinct from the two known steroid hydroxylases, P450(21) and P450(17)alpha. Multiple prominent protein bands in the mol wt range of known cytochrome P450s are seen on sodium dodecyl sulfate gels of guinea pig adrenal microsomes. Most are more intense in smooth microsomes, where the concentration of cytochrome P450 is highest. However, one band (52K) appears unique to the smooth microsomes. This band is also characteristic of microsomes obtained from the inner zone. This protein and two others (54K and 50K) are concentrated in the membrane pellet after carbonate treatment of the microsomes, indicating that they are integral membrane proteins. All three decrease in intensity after treatment of the animals with spironolactone, a compound known to cause depletion of adrenal cytochrome P450s. On Western blots of microsomal proteins the 54K and 50K proteins react with antibodies specific for P450(17) alpha and P450(21), respectively. The 52K protein, characteristic of the smooth microsomes and inner zone, does not react with anti-P450(21) or anti-P450(17) alpha, but does react with polyclonal antibody raised against microsomal cytochrome P450s induced by methylcholanthrene in rat liver (P450c,d). These results suggest that there is at least one additional cytochrome P450 in adrenal microsomes which is immunochemically distinct from P450(21) and P450(17) alpha. Its localization to the smooth microsomes and inner zone microsomes correlates with the high activity for ethylmorphine metabolism detected in these fractions. This suggests that this cytochrome P450, which is immunochemically related to members of the P450I subfamily, may be associated with the ability of guinea pig adrenal microsomes to metabolize foreign compounds.     

Rat hepatic stellate cells contribute to the acute-phase response with increased expression of alpha1(I) and alpha1(IV) collagens, tissue inhibitor of metalloproteinase-1, and matrix-metalloproteinase-2 messenger RNAs

The acute-phase response (APR) represents a systemic reaction of the organism to multiple nonspecific inflammatory stimuli. In general, it is protective for the host, and hepatocytes are the main cells responding with alterations in the expression of a set of liver-specific proteins named the acute-phase proteins. We have previously shown that although a turpentine-induced APR is not fibrogenic per se, it enhances collagen deposition in rats treated with CCl(4) and up-regulates expression of hepatic alpha1(I) collagen and tissue inhibitor of metalloproteinases 1 (TIMP-1) messenger RNAs (mRNAs). In this report we extended our studies and showed that turpentine induced, in a time-dependent manner, expression of alpha1(I) and alpha1(IV) collagens, TIMP-1, and matrix-metalloproteinase 2 (MMP-2) mRNAs. We further showed that expression of these mRNAs occurs in hepatic stellate cells, but not in hepatocytes obtained 6 hours after the induction of an APR episode. These changes were accompanied by increased blood levels of tumor necrosis factor alpha (TNF-alpha) and interleukin 6 (IL-6) without noticeable immediate changes in the expression of their respective mRNAs in the liver. In contrast to CCl(4)-induced liver damage, turpentine alone, whether administered as a single dose or as a weekly dose for 3 weeks did not up-regulate expression of transforming growth factor beta1 (TGF-beta1) mRNA and did not result in excess collagen deposition. Overall, these findings suggest that collagen deposition in the livers of rats with repeated APR episodes may be enhanced only when given together with a fibrogenic stimulus that activates hepatic stellate cells (HSCs) and/or up-regulates TGF-beta1 mRNA expression.     

Insulin-like growth factor I gene promoter polymorphism, collagen type II alpha1 (COL2A1) gene, and the prevalence of radiographic osteoarthritis: the Rotterdam Study

OBJECTIVE: To examine the role of an IGF-I gene promoter polymorphism in the prevalence of radiographic osteoarthritis (ROA), and study its interaction with the COL2A1 gene. METHODS: Individuals genotyped for IGF-I (n = 1546) and COL2A1 gene polymorphisms (n = 808) were selected from a random sample (n = 1583) derived from the Rotterdam study. The presence of ROA was defined as a Kellgren score of 2 or more in at least one of four joints (knee, hip, hand, and spine). Genotype specific odds ratios (OR) were adjusted for age, sex, body mass index, and bone mineral density using logistic regression. Interaction with the COL2A1 genotype was tested. RESULTS: Overall, no association was found between the IGF-I polymorphism and ROA. In subjects aged 65 years or younger (n = 971), the prevalence of ROA increased with the absence of the 192 base pair (bp) allele (p for trend = 0.03). Compared with homozygotes for the 192 bp allele, the prevalence of ROA was 1.4 times higher in heterozygotes (95% confidence interval, 1.0 to 1.8) and 1.9 times higher in non-carriers (1.1 to 3.3). There was evidence of interaction between the IGF-I and COL2A1 genes. Individuals with the risk genotype of both genes had an increased prevalence of ROA (OR 3.4 (1.1 to 10.7)). No effect was observed in subjects older than 65 years. CONCLUSIONS: SUBJECTS: with genetically determined low IGF-I expression (non-carriers of the 192 bp allele) may be at increased risk of ROA before the age of 65 years. Furthermore, an interaction between the IGF-I and COL2A1 genes is suggested.