Heterogeneous Nuclear Ribonucleoprotein-A2/B1 Modulate Collagen Prolyl 4-Hydroxylase, α (I) mRNA Stability*

Collagen prolyl 4-hydroxylase (C-P4H) α-subunit is of regulatory importance in the assembling of C-P4H tetramers, which are necessary for the hydroxylation of procollagen chains. Change in collagen expression by hypoxia or iron diminishment is a significant issue in extracellular matrix remodeling. It was proposed that C-P4H-α (I) is regulated at the posttrancriptional level under these conditions. Here we report that the induction of C-P4H-α (I) in human fibrosarcoma cells HT1080 by the iron chelator 2,2-dipyridyl is predominantly caused by an enhancement of mRNA stability. This effect is mediated by an increased synthesis and binding of heterogeneous nuclear ribonucleoprotein (hnRNP)-A2/B1, which interacts with a (U)16 element located in the 3′-untranslated region of C-P4H-α (I) mRNA. Luciferase reporter gene assays depending on C-P4H-α (I) 3′-untranslated region and co-transfection with hnRNP-A2/B1 provide evidence that the (U)16 element is necessary and sufficient for posttranscriptional control of C-P4H-α (I) synthesis under the analyzed conditions. Further indication for the significance of hnRNP-A2/B1 in C-P4H-α (I) induction was obtained by micro array experiments. In a data set representing 686 independent physiological conditions, we found a significant positive correlation between hnRNP-A2/B1 and C-P4H-α (I) mRNAs.

Collagen-modifying enzymes such as collagen prolyl-hydroxylases (C-PHs) 2 or collagen-degrading enzymes such as matrix metalloproteinases are important factors in collagen metabolism. Collagen turnover is a central issue for maintaining of the extracellular matrix, which influences cell adhesion, chemotaxis, and migration, as well as extracellular matrix remodeling during growth, differentiation, morphogenesis, wound healing, and many pathologic states, such as fibrosis (1).
One of the most important factors in collagen synthesis is the collagen prolyl 4-hydroxylase, which is necessary for proline hydroxylation of procollagen chains. The hydroxylation is required for correct folding of the triple-helical structure of procollagen molecules. The activity of C-P4H depends on Fe 2ϩ ions, 2-oxoglutarate, oxygen, and ascorbic acid (2,3). C-P4H is an ␣ 2 ␤ 2 tetramer. The ␤-subunit is synthesized in large excess over the ␣-subunit and is identical to the enzyme and chaperone protein disulfide isomerase (4). The ␣-subunit is regarded as the limited and regulated part and determines the rate of C-P4H-formation (5). Three different isoforms of the ␣-subunit, called type I, II, and III, are known (4,6). Type I isoform is the most abundant in most cell types including fibroblasts (7), which are mainly responsible for the production of extracellular matrix compounds.
In general, gene expression is regulated at different levels, including variation of the transcription rate, mRNA stability, efficiency of translation, or posttranslational modifications. In the recent past, it has become increasingly clear that the regulation of mRNA turnover can be a crucial step in the adaptation of gene expression to alterations of environmental conditions (8 -12). mRNA lifetime is modulated in particular by trans-acting proteins interacting with cis-elements residing frequently in the 3Ј-untranslated region (3Ј-UTR).
Important sequences controlling mRNA turnover are adenylate/uridylate-rich (AU-rich) elements (AREs), which interact with several ARE-binding proteins such as AUF-1, HuR, Hsp70, or hnRNP-A1 (10). Although the nonamer UUAUUUA(U/A)(U/A) has been proposed to form the optimal binding site for ARE-binding factors, a variety of AUUUA clusters and also oligo(U) tracts are targets (13)(14)(15)(16). In general, AREs are known as destabilizing elements controlling immediate early genes, but they do not act solely as mRNA destabilization determinants. The nature of ARE-binding proteins determines whether the mRNA stability is increased or decreased.
In this study, we present data that C-P4H-␣ (I) mRNA is stabilized by interaction of RNA-binding proteins hnRNP-A2/B1 with a U (16) element within the 3Ј-UTR. mRNA stabilization by this phenomenon seems to be the key mechanism responsible for P4H induction under hypometabolic conditions caused by diminishment of Fe 2ϩ ions by the iron chelator 2,2-dipyridyl.

EXPERIMENTAL PROCEDURES
Cell Culture and RNA/Protein Isolation-Human fibrosarcoma line HT1080 (ATCC, passages 16 -21) cells were maintained in Dulbecco's modified Eagle's medium (high glucose; PAA Laboratories GmbH) supplemented with 10% heat-inactivated fetal calf serum, 50 units/ml penicillin, 50 g/ml streptomycin, 15 mM Hepes, and 2 mmol/liter glutamine at 37°C, 5% CO 2 . Before use in experiments, cells were maintained in a medium containing 0.4% fetal calf serum for at least 24 h. Measurements started with the application of fresh medium containing 0.4% fetal calf serum. For 2,2-dipyridyl (2,2-DP, Sigma) treatment, the iron chelator was dissolved in ethanol and supplemented to a final concentration of 100 M 2,2-DP and 0.1% ethanol. Control cells were also supplemented with ethanol (0.1% final concentration). For RNA and protein isolation, cells were washed with ice-cold phosphate-buffered saline. RNA was prepared using RNA-Bee (Biozol Diagnostica Vertrieb GmbH) according to the manufacturer's protocol. Protein extracts (10,000 ϫ g supernatants, S10) were isolated using lysis buffer (10 mM Tris, pH 7.5, 140 mM NaCl, 1 mM EDTA, 25% glycerol, 0.1% SDS, 0.5% * This work was supported by Deutsche Forschungsgemeinschaft (DFG Grants GRK754 and TH459/5) and Bundesministerium fü r Bildung und Forschung (BMBF Grants NGFN2, KG-CV1.2, and Charité , Core Facility Proteomics). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  Nonidet P40, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1ϫ Complete protease inhibitor mix (Roche Diagnostics)).
Western Blotting-Protein extracts (30 g/sample) were separated by SDS-PAGE. After electrophoresis, proteins were transferred to Hybond TM -P membrane (Amersham Biosciences) using a Bio-Rad Mini Trans-Blot transfer cell. The membranes were blocked for 1 h with 5% BLOT-QuickBlocker (Chemicon). Following the blocking step, the membranes were incubated in 1% blocking solution containing a primary antibody (anti-C-P4H-␣: Acris Antibodies GmbH, AF0210; anti-hnRNP-A2/B1: Acris Antibodies GmbH, BM4520) for 1.5 h. The membranes were washed three times with Tris-buffered saline with Tween and incubated with an anti-mouse secondary-antibody (Promega) for 1 h. Bands were detected using the ChemiGlow TM West detection kit (Alpha Innotech Corp.). Membranes were stripped for 5 min with distilled water, 5-15 min with 0.2 M NaOH, and 5 min with distilled water and reprobed with anti-␤-actin antibody (Chemicon) to detect relative ␤-actin levels as loading control.
mRNA Stability-To test mRNA stability, cells were incubated for 18 h with or without the iron chelator 2,2-dipyridyl (100 M). Native cytosolic extracts were isolated using a lysis buffer (20 mM Tris, pH 7.4, 150 mM KCl, 30 mM MgCl 2 , 0.25% Nonidet P40, 1 mM dithiothreitol, 1ϫ Complete protease inhibitor mix) at 0 -4°C. Following the isolation step, the native extracts were incubated at room temperature up to 4 h.
RNA was prepared using RNA-Bee (Biozol Diagnostica Vertrieb GmbH) according to the manufacturer's protocol and reverse-transcribed using random hexamer primers. The estimation of mRNA concentration was performed by RT-PCR. To compare the RNA decay under control conditions and in the lack of iron ions, mRNA levels were adjusted to equal amounts at the starting point (0 h).
C-P4H-␣ (I) UTR-dependent Reporter Gene Constructs-For reporter gene assays, the luciferase vector pGL3-promoter (Promega, constitutive SV40 promoter) was used. The vector-specific 5Ј-and 3Ј-UTRs of luciferase mRNA were replaced by the human C-P4H-␣ (I) UTRs (Gen-Bank TM gi:63252885). The 5Ј-UTR of C-P4H-␣ (I) mRNA (133 nt) was cloned using the vector-specific HindIII (5Ј-end) and NcoI (3Ј-end) restriction sites and the 3Ј-UTR (999 nt, including the poly-A signal) using the vector-specific XbaI (5Ј-end) and BamHI (3Ј-end) restriction sites. These restriction sites were added to the UTR sequences by primer extension. 3Ј-UTR variants correspond to: 3Ј-A, first 500 nt of 3Ј-UTR; 3Ј-B, terminal 522 nt of 3Ј-UTR. Deletion of the U-rich element was performed by PCR technique. Modified vectors were confirmed by sequencing. The resulting vector constructs represent constitutively transcribed luciferase transcripts with or without the specific C-P4H-␣ (I) UTRs/UTR variants.

hnRNP-A2/B1 Modulate C-P4H-␣ (I) mRNA Stability
was used in a ratio of 1:3 (firefly luciferase vector:hnRNP expression vector). Luciferase activity was detected using the Dual-Glo TM luciferase assay system (Promega) and a luminometer (Labsystems Luminoscan RS) programmed with individual software (Luminoscan RII, Ralf Mrowka). The co-transfection with the Renilla-luciferase expression vector served as a control.
UV Cross-linking Experiments-1-2 ng of [␣-32 P]U/C/A-or [␣-32 P]GTP-labeled in vitro transcripts representing 100,000 cpm were incubated with 35 g of cytosolic protein extract for 30 min at room temperature in 10 mM Hepes pH 7.2, 3 mM MgCl 2 , 5% glycerol, 1 mM dithiothreitol, 150 mM KCl, and 2 units/l RNaseOUT (Invitrogen) in the presence of rabbit rRNA (0.5 g/l). For competition assays, a 50-fold excess of unlabeled in vitro transcripts was added. Then the samples were exposed to UV light (255 nm, 1.6 J, UV-Stratalinker) on ice, treated with RNase A (30 g/ml final concentration) and RNase T1 (750 units/ml final concentration) for 15 min at 37°C, and subjected to 12% SDS-PAGE and autoradiography.
mRNA Level Correlation Study-Normalized human mRNA expression levels for hnRNP-A2/B1 mRNA/C-P4H-␣ (I) mRNA were obtained from the Stanford Microarray data base as used in Stuart et al. (17), which contains a collection of different independent investigations. The difference of the sum of all microarrays in the data base and the number of data points used for the correlation analysis in this study results from the fact that some entries for hnRNP-A2/B1 and C-P4H-␣ (I) were missing in the microarray data sets. Regression and correlation analysis was performed using the math module of the open source/GPL program xmgrace, and the null hypothesis was rejected at the 0.05 level.
Statistical Analysis-Autoradiographic signals were scanned and quantified using the Scion Image software (Scion Corp.). Results appear as means, and in Figs. 1-3, 6, and 7, error bars represent the standard deviation (S.D.). Data were analyzed using the Student's t test, and the null hypothesis was rejected at the 0.05 level.

RESULTS
Human HT1080 fibroblasts were treated with or without 100 M 2,2-DP, and synthesis of P4H-␣ (I) was analyzed at the mRNA and protein level. As demonstrated in Fig. 1, the lack of Fe 2ϩ ions resulted in an ϳ4-fold increase at the mRNA level and an ϳ6.5-fold increase at the protein level after 18 h. The analysis of time-dependent C-P4H-␣ (I) mRNA decay confirmed that the elevated mRNA concentration was not only attributed to an increase at the transcriptional level but also to an enhanced mRNA stability (Fig. 2). The mRNA half-life time under control conditions is 1.9 h (Ϯ0.3 h, S.D.) but was elevated after 18 h of Fe 2ϩ diminishment to 4.2 h (Ϯ0.8 h, S.D.). The data indicated that in response to the lack of Fe 2ϩ ions, the C-P4H-␣ (I) mRNA stability increased about 2-fold. Interestingly, we did not observe an alteration in C-P4H-␣ (I) mRNA half-life time after 7 h of 2,2-DP treatment, indicating the requirements of trans-acting factors that have to be induced. However, the mRNA turnover of glyceraldehyde-3-phosphate dehydrogenase (Fig. 2B), C-P4H-␤, or C-P4H-␣ (II) (data not shown) was not altered under the same conditions. To confirm the posttranscriptional regulation of C-P4H-␣ (I) expression, we performed UTR-dependent reporter gene assays (pGL3-promoter vector) in which the 5Ј-and/or 3Ј-UTR of luciferase mRNA were replaced by specific 5Ј-and 3Ј-UTRs of C-P4H-␣ (I) mRNA or by artificial 3Ј-UTR variants (for a schematic illustration, see Fig. 3A). The transcription rate is controlled by the constitutive SV40 promoter. Therefore, differences in luciferase activity depend only on posttranscriptional control, mediated by the UTRs, and involve mainly mRNA stability and translational efficiency.
2,2-DP treatment neither changed luciferase activity resulting from the original pGL3-promoter vector (control) nor changed luciferase activity in the case when luciferase mRNA contains the C-P4H-␣ (I) 5Ј-UTR. By contrast, the C-P4H-␣ (I) 3Ј-UTR significantly enhanced luciferase activity (Fig. 3B), indicating a 3Ј-UTR-mediated posttranscriptional control. Interestingly, the combination of 5Ј-and 3Ј-UTR showed a significantly stronger increase in luciferase activity. This observation correlated with a higher increase at C-P4H-␣ (I) protein level when compared with the mRNA level after 18 h of 2,2-DP treat-  APRIL 7, 2006 • VOLUME 281 • NUMBER 14 ment. These data indicated the involvement of translational control in the adaptation of C-P4H-␣ (I) expression in response to Fe 2ϩ diminishment. Polysomal gradient analyses, however, indicated that translational control is of minor importance (data not shown). Thus, the 3Ј-UTR-mediated posttranscriptional control in the lack of Fe 2ϩ ions correlated mainly with an increased mRNA stability and is attributed to the 3Ј-terminal ϳ500 nt of the 3Ј-UTR (called 3Ј-B, Fig. 3B). We concluded that the enhanced mRNA stability is attributed to a cis-element within the terminal half of the 3Ј-UTR.

hnRNP-A2/B1 Modulate C-P4H-␣ (I) mRNA Stability
Further experiments concentrated on the investigation of C-P4H-␣ (I) 3Ј-UTR-protein interaction, with the aim to detect possible mRNA stabilization factors. As demonstrated in Fig. 4A, UV cross-linking assays revealed that two 3Ј-UTR-binding proteins with a molecular mass of ϳ36 and ϳ38 kDa showed an increased binding behavior in response to 18 h of Fe 2ϩ diminishment. These trans-acting factors were identified by affinity chromatography and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry analyses as RNAbinding proteins hnRNP-A2 and hnRNP-B1. Both proteins represent splicing variants resulting from one gene. Western blotting analyses showed that the concentrations of both proteins were elevated under Fe 2ϩ deficiency after 18 h (Fig. 5B), which correlated well with the increased binding seen in cross-linking experiments (Fig. 4A). The enhanced protein concentrations of the RNA-binding proteins hnRNP-A2 and -B1 can be attributed to an elevated mRNA concentration (Fig. 5A). We did not observe a significant increase in hnRNP-A2/B1 mRNA stability (data not shown), indicating that the induction of hnRNP-A2/B1 expression in response to Fe 2ϩ diminishment is controlled mainly at the transcriptional level.
Trans-factors such as hnRNP-A2/B1 interact with discrete cis-mRNA elements or specific three-dimensional mRNA structures. UV cross-linking signals depend on the direct interaction of trans-acting factors with cis-acting elements. The observed signal intensity depends on the quality and quantity of nucleotides involved in the RNAprotein interaction. The separate labeling of C-P4H-␣ (I) 3Ј-UTR transcripts with the four possible nucleotides [ 32 P]U/C/A or [ 32 P]GTP is an approach to get detailed information concerning which nucleotides are preferentially represented in the binding site. Such labeling experiments revealed that both hnRNP-A2 as well as hnRNP-B1 only interacted with uracil in the recognition element (Fig. 4B). There was no label transfer from C-P4H-␣ (I) 3Ј-UTR to hnRNP-A2/B1 proteins when the other possible nucleotides were used for the labeling reaction. From the UTRdependent reporter gene assays, however, it was evident that only the 3Ј-terminal half of the 3Ј-UTR is responsible for mRNA stabilization. UV cross-linking competition assays in which non-labeled in vitro transcripts representing the first 500 nt of 3Ј-UTR (3Ј-A) or terminal 522 nt (3Ј-B) were added in an excess when compared with the radioactive labeled 3Ј-UTR showed that hnRNP-A2/B1 significantly interacted only with the 3Ј-terminal part of the 3Ј-UTR (Fig. 4C). Although both 3Ј-UTR parts contain several U (3)(4)(5)(6)(7) elements (see Fig. 7A, sequence), only the 3Ј-B part contains a U-rich element, which consists of a continuous stretch of 16 uridines (U (16) element). As shown by UTR-dependent reporter gene assays, the deletion of the 3Ј-UTR-located U (16)

hnRNP-A2/B1 Modulate C-P4H-␣ (I) mRNA Stability
element significantly suppressed the posttranscriptional influence in response to the diminishment of Fe 2ϩ ions by 2,2-dipyridyl (Fig. 6). These data confirm that the increased binding amount of hnRNP-A2/B1 to the 3Ј-UTR-located U (16) element is of crucial importance in the adaptation of C-P4H-␣ (I) gene expression under hypometabolic conditions, caused by the lack of Fe 2ϩ ions.
Furthermore, the influence of the trans-factors hnRNP-A2/B1 by interaction with the C-P4H-␣ (I) 3Ј-UTR U (16) cis-element on gene expression was demonstrated by UTR-dependent reporter gene assays. As shown in Fig. 7B, overexpression of hnRNP-A2/B1 lead to a significant increase in C-P4H-␣ (I) 3Ј-UTR-dependent luciferase activity. This influence was abolished by the deletion of the U (16) element. In contrast, another known RNA-binding protein, hnRNP-E1, did not alter C-P4H-␣ (I) 3Ј-UTR-mediated reporter gene activity, neither if the U (16) element was present nor if it was deleted.
To test the hypothesis whether hnRNP-A2/B1 is also important as a regulatory factor under other physiological or pathophysiological conditions, we performed an hnRNP-A2/B1 mRNA/C-P4H-␣ (I) mRNA correlation study based on microarray data, which depend on 686 different experimental conditions (17). The elevated hnRNP-A2/B1 expression level was associated with an increased mRNA concentration (Fig. 5A) and has a positive influence on C-P4H-␣ (I) mRNA stability. Thus, an elevated hnRNP-A2/B1 mRNA level should cause on average an elevated C-P4H-␣ (I) mRNA level. Therefore, we would expect a positive correlation between both mRNA signals. Indeed, statistical analysis revealed a significant positive correlation between hnRNP-A2/B1 and C-P4H-␣ (I) mRNA levels (Fig. 8). This supports the view that hnRNP-A2/B1 is not only a factor of C-P4H-␣ (I) mRNA stabilization under conditions of Fe 2ϩ deficiency but is also relevant for a large variety of other changes in cell metabolism.

DISCUSSION
Posttranscriptional control affecting mRNA properties such as mRNA localization, translational efficiency, or mRNA stability modulates the expression rate, probably of each gene. All eukaryotic mRNAs contain 5Ј-and 3Ј-UTRs of different length. The human average of the   APRIL 7, 2006 • VOLUME 281 • NUMBER 14 5Ј-UTR is 300 nt (median: 240 nt), and that of the 3Ј-UTR is ϳ770 nt (median: 400 nt) (18). mRNA UTRs constitute a "hotspot" of regulatory elements (cis-elements), which interact with RNA-binding factors (trans-factors) to form specific ribonucleoprotein (RNP) complexes (19 -22). mRNP complexes are dynamic, and trans-factors provide a link between extracellular signals and mRNA metabolism (23). It has been shown that posttranscriptional control is important in development (24) or in adaptation to stress conditions such as heat shock (25) or hypoxia (26 -28).

hnRNP-A2/B1 Modulate C-P4H-␣ (I) mRNA Stability
The lack of Fe 2ϩ ions mimics hypoxia/anoxia by inhibiting the respiratory chain, which subsequently leads to hypometabolism. Consequently, iron chelators such as desferrioxamine or 2,2-dipyridyl are used when studying hypoxia response pathways (29 -31). However, the adap-tation of C-P4H-␣ (I) gene expression in the lack of Fe 2ϩ ions is different when compared with hypoxic conditions (1% oxygen) and seems to be mediated by different mechanisms. Hypoxia induces C-P4H-␣ synthesis mainly by activation of transcription and increases the efficiency of translation (32,33). As we show here, Fe 2ϩ diminishment induces C-P4H-␣ (I) synthesis by a different molecular effect, namely by increasing mRNA stability via RNA-binding proteins hnRNP-A2/B1, which interact with a U (16) (16) element is shown in bold print. The artificial division of the 3Ј-UTR as described in the legend for Fig. 3 is marked as follows. The dotted line indicates a sequence that is present in both 3Ј-A and 3Ј-B, and the underlined part represents the 3Ј-B part. B, HT1080 cells were transfected with the original pGL3-promoter (pGL3p) vector or transgenic variants, in which the original luciferase 3Ј-UTR was replaced by the specific C-P4H-␣ (I) 3Ј-UTR, with or without the U (16) element. Furthermore, cells were co-transfected with expression vectors encoding the hnRNP-A2/B1 or hnRNP-E1 proteins. Controls were co-transfected with empty expression vectors. The influence of hnRNP-A2/B1 or hnRNP-E1 overexpression on luciferase activity, depending on C-P4H-␣ (I) 3Ј-UTR (Luc-3Ј-UTR) or 3Ј-UTR without the U (16) element (Luc-3Ј-U(16) del), was measured. Relative values were normalized to the influence of hnRNPs on the original pGL3-promoter vector. hnRNP-A2/B1 co-transfection leads to a significant increase in C-P4H-␣ (I) 3Ј-UTR-dependent luciferase activity (p Ͻ 0,01), and the influence is abolished by deletion of the U (16) element. hnRNP-E1 shows no significant influence on C-P4H-␣ (I) 3Ј-UTR-dependent luciferase activity, independently of whether the U (16) element is present or not. n ϭ 8, **, p Ͻ 0.01.

hnRNP-A2/B1 Modulate C-P4H-␣ (I) mRNA Stability
A2/B1 is not induced by hypoxia. 3 Differences in the alteration of gene expression mediated by UTR-dependent posttranscriptional control in response to hypoxic conditions (1% oxygen) and hypometabolic conditions caused by Fe 2ϩ diminishment are also described in the adaptation of matrix metalloproteinase-9 (MMP-9) gene expression (34). The interaction of RNA-binding factors with mRNA UTRs therefore seems to be crucial and determines the expression rate under stress conditions substantially.
Several efforts were made to find mRNA ligands bound by hnRNP-A2. A large variety of binding mRNAs were characterized that all contained the pentamer AUUUA, the nonamer UUAUUUA(U/A)(U/A), or poly(U) tracts (16). C-P4H-␣ (I) mRNA, however, was not found among the candidates. In vitro studies using homopolymers revealed a strong binding of hnRNP-A2 and -A1 to poly(U) at physiological salt concentration (43). Thus, Au richness seems not to be the only feature of an hnRNP-A2/B1-binding sequence. A quite different type of hnRNP-A2/B1-binding motif of 11 bases, which is not AU-rich, was described in myelin basic protein mRNA. There it plays a role in mRNA trafficking (39). The up-regulation of hnRNP-A2/B1 by Fe 2ϩ deficiency has an interesting feature. Both proteins are overexpressed at an early stage in a variety of tumors and have been proposed as early markers for cancer (44).
Our observation that the elevated hnRNP-A2/B1 expression starts with an enhanced mRNA concentration enables us to test the hypothesis whether hnRNP-A2/B1 modulate the C-P4H-␣ (I) mRNA stability not only after Fe 2ϩ diminishment. We have identified a positive correlation between hnRNP-A2/B1 and C-P4H-␣ (I) gene expression in a large data set of 686 independent microarray experiments. The data include a broad variety of physiological and pathophysiological conditions such as the human cell cycle, response to serum, response to infectious challenges, and profiles of multiple cancers. This supports independently the idea of hnRNP-A2/B1 as a positive effector of C-P4H-␣ (I) gene expression.
Taken together, we have shown that hnRNP-A2/B1 regulate C-P4H-␣ (I) mRNA stability. The RNA-protein interaction is mediated by a U-rich element located within the 3Ј-UTR of the mRNA. Thus, hnRNP-A2/B1 are demonstrated as factors participating in collagen synthesis and extracellular matrix remodeling.