SEARCH

SEARCH BY CITATION

Keywords:

  • tissue inhibitor of metalloproteinase 1;
  • NvTIMP1;
  • limb regeneration;
  • matrix metalloproteinases;
  • MMPs;
  • newt;
  • Notophthalmus viridescens

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Matrix metalloproteinase (MMP) activity is important for newt limb regeneration. In most biological processes that require MMP function, MMP activity is tightly controlled by a variety of mechanisms, including the coexpression of natural inhibitors. Here, we show that gene expression of one such inhibitor, tissue inhibitor of metalloproteinase 1 (NvTIMP1), is upregulated during the wound healing and dedifferentiation stages of regeneration when several MMPs are at their maximal expression levels. Newt MMPs and NvTIMP1 also exhibit similar spatial expression patterns during the early stages of limb regeneration. NvTIMP1 inhibits the proteolytic activity of regeneration-related newt MMPs and, like human TIMP1, can induce a weak mitogenic response in certain cell types. These results suggest that NvTIMP1 may be functioning primarily to maintain optimal levels of MMP activity during the early stages of limb regeneration, while possibly serving a secondary role as a mitogen. Developmental Dynamics 235:606–616, 2006. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Newts and other salamanders can regenerate several organs and structures, including their limbs, tails, jaws, spinal cords, retinas, lenses, optic nerves, heart ventricles, and intestines (Piatt,1955; O'Steen,1958; Turner and Singer,1974; Brockes and Kumar,2002). Limb regeneration in salamanders is accomplished through a complex process involving the migration of epithelial cells to cover the wound and the dedifferentiation and proliferation of internal cells to form a regeneration blastema (Thornton,1938a,b; Chalkley,1954; Bodemer,1958; Hay and Fischman,1961). The thickened epithelium that covers the wound is known as the apical epithelial cap (AEC) and is essential for the outgrowth of the regenerating salamander limb (Thornton,1957). The blastema contains the progenitor cells that are required to generate all the internal tissues of the regenerated newt limb, except the nerve axons and blood vessels, which are thought to be formed primarily through the regrowth of existing nerve axons and angiogenesis, respectively (Rageh et al.,2002).

To identify genes that might be involved in establishing or maintaining the blastema, we have performed differential display analysis between RNA extracted from regenerating and intact newt limbs. This analysis has identified tissue inhibitor of metalloproteinase 1 (NvTIMP1) as a gene that is upregulated within hours of limb amputation and is maintained at elevated levels through most of the blastemal stage of limb regeneration. NvTIMP1 is an interesting candidate for regulating some of the cellular and molecular processes that occur during the initial stages of limb regeneration. First, TIMP1 in other species inhibits the proteolytic activity of many of the matrix metalloproteinases (Sternlicht and Werb,2001). The proteolytic activity of MMPs is required for a variety of regenerative processes, including foot regeneration in hydra (Leontovich et al.,2000), intestinal regeneration in sea cucumbers (Quinones et al.,2002), limb regeneration in newts (Vinarsky et al.,2005), and fin regeneration in zebrafish (Bai et al.,2005). NvTIMP1 may function to regulate this activity at an optimal level during newt limb regeneration. Second, human TIMP1 and TIMP2 have been shown to act as mitogens on a variety of human, bovine, and mouse cell types, including erythroid precursor cells, keratinocytes, fibroblasts, smooth muscle cells, endothelial cells, and chondrocytes (Gasson et al.,1985; Bertaux et al.,1991; Hayakawa et al.,1992,1994; Nemeth and Goolsby,1993; Yamashita et al.,1996; Wang et al.,2002). These dual functions of TIMP1 are independent of each other, given that point mutations that abolish the proteolytic inhibitory function of human TIMP1 do not ablate its mitogenic activity (Chesler et al.,1995). Human TIMP1 has also been shown to be a cell survival factor for several cell types (Alexander et al.,1996; Li et al.,1999; Murphy et al.,2002; Yoshiji et al.,2002), as well as an enhancer or inhibitor of bone resorption depending on whether TIMP1 concentrations are low or high, respectively (Sobue et al.,2001).

To characterize the function(s) of NvTIMP1 during newt limb regeneration, we have performed temporal RNA expression analyses of NvTIMP1 over the first 40 days of limb regeneration and have determined the spatial expression patterns of NvTIMP1 during its peak expression. We have shown that purified recombinant NvTIMP1 protein can inhibit the newt MMPs that are known to be upregulated during the earliest stages of limb regeneration and have found evidence that NvTIMP1 can act as a mitogen on at least one cell type. These results suggest that NvTIMP1 may be functioning primarily to maintain MMP activity at the optimal level for normal limb regeneration, while possibly playing a secondary role as a mitogenic factor during regeneration.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Differential Display Analysis and Cloning of NvTIMP1

A TIMP gene was initially identified as being upregulated during the first few days of limb regeneration using differential display analysis between RNAs extracted from regenerating and intact newt limbs. A modified RNA Ligase Mediated-Rapid Amplification of cDNA Ends (RLM-RACE) procedure (Vinarsky et al.,2005) was used to clone the full-length TIMP gene. The predicted sequence for the newt TIMP protein was 219 amino acids in length and contained a 25–amino acid signal peptide. The protein exhibited 40–42% identity and 61–62% similarity to mammalian TIMP1 proteins (Fig. 1) and 35% identity and 56–58% similarity to mammalian TIMP2. Furthermore, the sequence revealed two potential N-glycosylation sites that are found on mammalian TIMP1 proteins but not TIMP2 proteins (Kossakowska et al.,1998). For these reasons, we have classified this gene as the Notophthalmus viridescens ortholog of TIMP1 (NvTIMP1).

thumbnail image

Figure 1. A comparison of the NvTIMP1 and human TIMP1 sequences. The double arrowed bar above the sequence denotes the signal peptide required for secretion of the TIMP1 proteins from the cell. The amino acids represented in bold letters are the conserved cysteine residues that form the disulfide bonds. Disulfide bonds between cysteine residues are shown by connecting lines and are based upon the disulfide bonds present in human TIMP1. Predicted N-glycosylation sites are noted by asterisks. The GenBank accession number for the sequence reported here is DQ286430.

Download figure to PowerPoint

Temporal Expression Patterns of NvTIMP1 During Newt Limb Regeneration

We assessed the temporal expression patterns of NvTIMP1 during limb regeneration to determine the stage of regeneration in which NvTIMP1 might play an important role. Total RNA was extracted from regenerating tissues at various time points over the first 40 days of limb regeneration, and Northern blot, microarray, and real-time RT-PCR were used to assess NvTIMP1 expression levels (Fig. 2). These analyses revealed that NvTIMP1 was upregulated more than 2-fold within 6 hr of limb amputation (microarray analysis; data not shown) and that the peak expression level (8.5–11-fold upregulation over levels found in intact newt limbs as determined by real-time RT-PCR) was reached by 24 hr postamputation. Elevated expression levels were maintained through about regeneration day 20. These expression data suggest that NvTIMP1 primarily functions in the early phases of limb regeneration during the formation and maintenance of the regeneration blastema.

thumbnail image

Figure 2. Temporal expression of NvTIMP1 during newt limb regeneration. A: Northern blot analysis demonstrates that NvTIMP1 is upregulated within hours of limb amputation and remains upregulated through about day 20 postamputation. Bottom: The 28S and 18S rRNA bands following methylene blue staining of the northern blot following RNA transfer. B: Real-time RT-PCR analysis determines relative levels of NvTIMP1 upregulation through the blastemal stage of limb regeneration. Two different primer sets that recognized NvTIMP1 were used to assess relative increases in NvTIMP1 expression levels.

Download figure to PowerPoint

Spatial Expression Patterns of NvTIMP1 During the Early Stages of Newt Limb Regeneration

Spatial expression patterns for NvTIMP1 were obtained during its period of maximum expression (regeneration days 1–5) by performing RNA in situ hybridization on 1-, 3-, and 5-day limb regenerates (Fig. 3). At regeneration day 1, NvTIMP1 was highly expressed in the proximal epidermis, dermis, wound epithelium, myofibers, and periosteal cells. NvTIMP1 was also expressed in endosteal cells at 1 day postamputation (data not shown, but see data for day 5 in Fig. 3). By regeneration day 3, the general spatial expression pattern remained the same but the signal was reduced in most tissues, except muscle. In myofibers, expression appears to increase on day 3 with most of the NvTIMP1 transcript localized to the nuclear and perinuclear regions of the myofiber. By the 5th day of regeneration, NvTIMP1 remained high in periosteal and endosteal cells but had dropped to lower levels in the AEC (data not shown) and was undetected in the myofibers and proximal skin tissues.

thumbnail image

Figure 3. Spatial expression patterns of NvTIMP1 during newt limb regeneration. A,C,E,G,I: Antisense NvTIMP1 probe (AS) used in RNA in situ hybridization of tissue sections from regenerating newt limbs. B,D,F,H,J: Sense NvTIMP1 probe (S) used in RNA in situ hybridization of tissue sections from regenerating newt limbs. A: A day 1 limb regenerate showing NvTIMP1 expression in the wound epithelium (top right arrow), proximal epidermis (top left arrow), and dermal tissues (bottom left arrow). C: A day 1 limb regenerate demonstrating NvTIMP1 expression in muscle tissues (left arrow and periosteum (right arrow). E: A day 3 limb regenerate exhibiting NvTIMP1 expression in muscle tissues (arrow). G: A high-power photomicrograph of a day 3 limb regenerate illustrating that the NvTIMP1 transcripts (arrows) are located primarily in the perinuclear region of the myofibers. The tissues were counterstained with hematoxylin to identify nuclei. I: A day 5 limb regenerate exhibiting expression of NvTIMP1 in the periosteum (top arrow) and endosteum (bottom arrows). Note that muscle, dermis, and epidermis do not express NvTIMP1 on day 5. B, D,F,H,J: Sense strand NvTIMP1 probes do not exhibit staining. e, epithelium; b, bone; m, muscle. The black sections in D, I, and J are newt pigments. Scale bar in A is for A–F. Scale bar in G is for G and H. Scale bar in I is for I and J.

Download figure to PowerPoint

Inhibition of Newt MMP Proteolytic Activity by NvTIMP1

NvTIMP1 was tagged on the 3′-end of the open reading frame with a DNA sequence that encodes the eight–amino acid FLAG sequence (DYKDDDDK). This tagged cDNA was then cloned into the pCMV-SPORT6 mammalian expression vector and expressed in human embryonic kidney 293H (HEK 293H) cells. The NvTIMP1.flag protein was purified from conditioned medium by affinity chromatography and its concentration was estimated by comparison to a FLAG-tagged bovine alkaline phosphatase standard following Western blotting (Fig. 4). The purified recombinant protein had an apparent molecular weight of 30 kDa, while the predicted size was 22.5 kDa. This suggests that post-translational modification of the protein had occurred and is consistent with both the presence of potential N-glycosylation sites in NvTIMP1 as well as the known glycosylation of TIMP1 in other species (Kossakowska et al.,1998).

thumbnail image

Figure 4. Purification of recombinant FLAG-tagged NvTIMP1 (NvTIMP1.flag) from HEK293H conditioned medium. A: Western blot of 30-kDa NvTIMP1.flag protein using anti-FLAG M2 antibody. B: SDS-PAGE gel showing degree of purification of the NvTIMP1.flag protein following affinity chromatography. A lane showing the conditioned medium taken from HEK293H cells before purification illustrates the degree of purification that can be achieved using this technique.

Download figure to PowerPoint

Reverse zymography, rat-tail type I collagen activity assays, and fluorescence dequenching assays conducted in the presence of NvTIMP1.flag and the 4-aminophenylmercuric acetate (APMA)-activated newt MMPs nCol, MMP9, MMP3/10a, and MMP3/10b (Vinarsky et al.,2005) demonstrated that NvTIMP1.flag can inhibit the proteolytic activity of all four of these newt MMPs (Fig. 5). Reverse zymography was performed in the presence of either MMP9-embedded gelatin gels or MMP3/10a-embedded casein gels. The dark 30-kDa bands indicate that NvTIMP1.flag is able to inhibit the MMP-directed cleavage of the embedded gelatin or casein substrate (Fig. 5A). Conditioned medium containing the untagged recombinant NvTIMP1 protein functioned similarly to the corresponding purified TIMP1 FLAG-tagged protein in the reverse zymography assays, suggesting that the FLAG-tag had no observable effect on protein function (data not shown).

thumbnail image

Figure 5. Functional analyses of purified recombinant NvTIMP1.flag protein. A: Reverse zymography illustrates that the 30-kDa NvTIMP1.flag protein can inhibit newt MMP9 digestion of gelatin and newt MMP3/10a digestion of casein. B: Newt MMP3/10b is inhibited by NvTIMP1.flag using fluorescence quenching assays. As the various substrates (DQ-Collagen I, DQ-Collagen IV, DQ-Gelatin, and BODIPY-Casein) are digested by NvMMP3/10b over time, the fluorescence signals increase (blue diamonds and lines). However, in the presence of equimolar (pink squares and lines) or 3-fold excess (yellow triangles) NvTIMP1.flag, digestion of the substrates are inhibited and the fluorescence remains quenched. In these assays, 1.6 pmol of newt MMP3/10b and either 1.6 pmol or 4.8 pmol of NvTIMP1.flag were used. C: A native rat-tail type I collagen digestion assay demonstrates that NvTIMP1.flag can inhibit the newt collagenase nCol. α1 and α2 are the two type I collagen strands that comprise the triple helical form of native rat-tail type I collagen. TCA and TCB represent the ¾ and ¼ digestion products, respectively, of α1 and α2. p-nCol is unactivated pro-nCol, while a-nCol is the activated form of nCol following the removal of the propeptide by APMA.

Download figure to PowerPoint

Fluorescence dequenching assays were performed to determine whether NvTIMP1.flag could inhibit the function of newt MMP3/10b. In these assays, either equimolar or 3-fold molar excess of NvTIMP1.flag was premixed with MMP3/10b and 10 min later the quenched MMP substrate was added to the mixture. In the absence of NvTIMP1.flag, MMP3/10b readily digested the substrate as measured by the increase in fluorescent signal as the fluorescent dyes were dequenched. However, this activity was almost completely eliminated in the presence of equimolar NvTIMP1.flag, and a 3-fold molar excess of NvTIMP1.flag abolished all protease activity (Fig. 5B). Under these conditions, it would be expected that equimolar NvTIMP1.flag would almost completely inhibit MMP3/10b, given that TIMP1 binds and inhibits MMPs in a 1:1 stoichiometric manner and the two proteins were premixed before adding the substrate (Murphy and Willenbrock,1995).

Native rat-tail type I collagen is a substrate for collagenases such as nCol (Vinarsky et al.,2005). When a collagenase digests rat-tail type I collagen, the two α1 and single α2 chains of the triple helical collagen are each cleaved at a single site producing two fragments: one ¾ the length of the original chain and the other ¼ the length of the chain. The ¾ length peptides are known as TCA fragments and the ¼ length peptides are TCB fragments. When NvTIMP1.flag was mixed with activated newt nCol and rat-tail type I collagen, no TCA or TCB fragments were produced, indicating that NvTIMP1.flag can inhibit the proteolytic function of newt nCol (Fig. 5C). In this assay, concentrations of 20 nM NvTIMP1.flag were required to completely inhibit 5 nM activated nCol. The 4-fold molar excess of NvTIMP1.flag required to completely inhibit the reaction in these assays probably reflects the fact that the substrate, enzyme, and inhibitor were mixed together at the same time. In reactions containing up to 2-fold excess inhibitor, some of the substrate was digested before the inhibitor could bind to and inactivate the enzyme. At 4-fold molar excess of inhibitor, most of the enzyme molecules would have been bound by the inhibitor before they could cleave the substrate.

NvTIMP1 Exhibits Differential Mitogenic Activities on Different Cell Types

Given that human TIMP1 and TIMP2 can act as mitogens on a variety of human, bovine, and mouse cell types (Bertaux et al.,1991; Hayakawa et al.,1992,1994; Nemeth and Goolsby,1993; Hayakawa,2002), we examined the mitogenic potential of NvTIMP1.flag on four mammalian cell lines and one newt cell line using the WST-8 assay (Fukuda et al.,2005; Matsuda et al.,2005; Orita et al.,2005). Viable cells metabolize the WST-8 reagent producing a soluble chromogenic substance that can be assayed by absorbance at 450 nm. The degree of absorbance is directly proportional to the cell number so that cell proliferation can accurately be examined. The cell lines chosen for this study included mouse C2C12 myoblasts (Yaffe and Saxel,1977; Blau et al.,1983), mouse NIH/3T3 embryonic fibroblasts (Jainchill et al.,1969), human IMR-90 fetal lung fibroblasts (Nichols et al.,1977), newt A1 limb cells (Ferretti and Brockes,1988), and human MG-63 osteosarcoma cells (Billiau et al.,1977). The first four cell lines were chosen because they represent the types of cells that are present in limbs, while the latter cell type was chosen because it had been shown to respond to human TIMP1.

NvTIMP1.flag exhibited a weak but reproducible proliferative effect on the osteosarcoma cell line MG-63 when cultured in serum-free medium (Fig. 6A). This cell line has previously been shown to proliferate in response to human TIMP1 (Yamashita et al.,1996) and we have confirmed this previous finding, while demonstrating that mouse Timp1 can also produce a similar mitogenic response in MG-63 cells. However, the three TIMP1 proteins produce a relatively weak mitogenic response on MG-63 cells when compared to the proliferative effects of IGF1. NvTIMP1.flag did not consistently act as a mitogen on any of the other cell lines tested (Fig. 6B,C and data not shown), whether in the absence or presence of low levels of serum (1% horse serum or 0.25% FBS; data not shown). This was also true for the human and mouse TIMP1 proteins (data not shown). These results suggest that NvTIMP1 can have differential proliferative effects on different cell types, presumably due to the presence or absence of the appropriate cell surface receptor (Avalos et al.,1988; Bertaux et al.,1991; Luparello et al.,1999). Whether there are cells present in regenerating newt limbs that can respond to the mitogenic effects of NvTIMP1 has not yet been determined.

thumbnail image

Figure 6. Proliferation assays to assess mitogenic capabilities of NvTIMP1. All proliferation assays shown were performed in the absence of serum following cell cycle synchronization in serum-free medium. A: Newt, human, and mouse TIMP1 (NvTIMP1, HsTIMP1, and MmTimp1) acted as weak mitogens on MG-63 cells when compared to the IGF1 positive control. B: Neither NvTIMP1 nor IGF1 acted as mitogenic factors on newt A1 cells in the absence of serum. C: Neither NvTIMP1 nor IGF1 acted as mitogenic factors on IMR-90 cells in the absence of serum. *P = 0.001, **P < 0.0005, and ***P < 0.0001 using either ANOVA or Student's t-test.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We have previously shown that MMPs are highly upregulated during the early stages of limb regeneration and that inhibition of MMP proteolytic activity during the regenerative process produces either limb stumps that completely fail to regenerate or dwarfed and deformed limb regenerates (Vinarsky et al.,2005). Other studies have shown that MMPs are also required for foot regeneration in hydra, regeneration of intestines in the sea cucumber, and regeneration of caudal fins in zebrafish (Leontovich et al.,2000; Quinones et al.,2002; Bai et al.,2005).

We show here that NvTIMP1, which encodes a natural inhibitor of newt MMPs, is also upregulated during the early stages of limb regeneration. NvTIMP1 reaches a maximal expression level (9- to 11-fold upregulation over intact limb tissues) at 24 hr postamputation, which coincides with the time of maximal expression for nCol, MMP3/10b, MMP9, and MMP3/10a (Vinarsky et al.,2005). RNA in situ hybridization studies indicate that by 1 day postamputation, NvTIMP1 is expressed in the wound epithelium, proximal epidermis, dermis, myofibers, periosteal cells, and endosteal cells. By day 3, the expression levels of NvTIMP1 decrease slightly to 5- to 7-fold above the levels found in intact limbs. This overall decrease in NvTIMP1 expression is reflected in a reduction of NvTIMP1 transcripts in the AEC, proximal epidermis, and dermis. In contrast, expression in muscle tissues is markedly increased on day 3 with most of the transcripts localizing to the nuclear and perinuclear regions of myofibers. However, not all myofiber nuclei transcribe NvTIMP1 simultaneously. The localization of transcripts to the perinuclear/nuclear region of myofibers, as well as the differential expression of transcripts between nuclei within a single myofiber, represents a common pattern of gene expression within myofibers (Newlands et al.,1998; Sacks et al.,2003) and may reflect the fact that a substantial portion of the secretory Golgi complex is located in the perinuclear regions of myofibers (Ralston,1993). By day 5 postamputation, the expression levels of NvTIMP1 decrease further to about 3- to 5-fold above the levels found in intact limbs. At this time, NvTIMP1 expression is confined to periosteal and endosteal cells and the AEC.

During the first 5 days of limb regeneration, MMPs are expressed in the same tissues as NvTIMP1 (Vinarsky et al.,2005). nCol, MMP3/10b, MMP9, and MMP3/10a are expressed in the wound epithelium/AEC, while nCol, MMP3/10b, and MMP9 are expressed in both the periosteum and endosteum and MMP3/10b and MMP9 are expressed in muscle tissue. This coexpression in tissues, coupled with the demonstration that NvTIMP1 can inhibit all four newt MMPs that are known to be upregulated during the early stages of limb regeneration, suggests that NvTIMP1 may be important for the regulation of MMP activity during limb regeneration. Although we have previously demonstrated that MMPs are required for normal newt limb regeneration and may act to prevent scar formation (Vinarsky et al.,2005), unregulated proteolytic activity might be detrimental to the regenerative process. In many physiologic processes, both MMPs and TIMP1 are coexpressed, thereby producing a tightly regulated response required for optimal MMP activity (Gardner and Ghorpade,2003). During limb regeneration, an optimal level of proteolytic activity could be maintained, in part, by the action of MMP inhibitors such as NvTIMP1. Other mechanisms for maintaining proper MMP activity during regeneration include the well-established regulation of mRNA levels (Yang and Bryant,1994; Miyazaki et al.,1996; Yang et al.,1999; Kato et al.,2003; Vinarsky et al.,2005) as well as the post-translational modification, secretion, activation, pericellular location, and catabolism and clearance of MMP proteins (Sternlicht and Werb,2001).

Previous studies have shown that human TIMP1 can act as a mitogen on several mammalian cell lines (Gasson et al.,1985; Bertaux et al.,1991; Hayakawa et al.,1992; Yamashita et al.,1996; Wang et al.,2002). Therefore, we tested whether NvTIMP1 could act as a mitogen on one newt and four mammalian cell lines. We show that one of these cell lines, the human osteosarcoma cell line MG-63, can respond to NvTIMP1 stimulation. The proliferative response is weak when compared to the response achieved following stimulation with a strong mitogen, such as IGF1, but is similar to the proliferative activity produced by human and mouse TIMP1. The other cell lines tested did not proliferate in response to human TIMP1, mouse TIMP1, or NvTIMP1, suggesting that these cells do not carry the required signaling machinery to respond to TIMP1 proteins. Human TIMP1 has been shown to exhibit high affinity binding (Kd value in the picomolar to low nanomolar range) to an 80-kDa transmembrane protein expressed on responding cells (Avalos et al.,1988; Bertaux et al.,1991; Luparello et al.,1999), suggesting that TIMP1 acts directly through cell surface receptors to induce the proliferative response. It has also been shown that more tumorigenic cell lines respond to the mitogenic effects of human TIMP1 (Luparello et al.,1999). This latter finding may have implications for the functions of NvTIMP1 during limb regeneration, given that cells produced during blastemal formation share many of the characteristics of tumor cells and may, therefore, respond to TIMP1 stimulation.

A recent study has shown that a related TIMP gene, ztimp2, is upregulated during zebrafish fin regeneration and is coexpressed with the Mmp genes zmt1-mmp (zmmp14) and zmmp2 (Bai et al.,2005). The coexpression of TIMPs and MMPs in at least two regenerating systems suggests that tight regulation of MMP activity is necessary for many, if not all, regenerative processes. However, TIMP2 is also known to function in the activation latent MMP2 proteins. Most MMPs are secreted as zymogens that are activated by the proteolytic cleavage of their propeptides. This cleavage is usually accomplished by the action of other activated MMPs or by serine proteases (Sternlicht and Werb,2001). Latent MMP2, however, is not activated by serine proteases and is instead cleaved at the cell surface by the interaction of TIMP2, activated MMP2, and membrane-type MMPs, such as MMP14 (Strongin et al.,1995; Sternlicht and Werb,2001). The coexpression of these three interacting proteins during fin regeneration implies that Timp2 may be functioning to assist in the activation of latent Mmp2.

Our results suggest that the primary role of NvTIMP1 during newt limb regeneration is to maintain the MMP proteolytic activity at a level that promotes the regenerative process. Our data also suggest that NvTIMP1 could act as a mitogenic factor if cells in the regenerating newt limb express the required receptor(s). Further studies are necessary to determine whether such cells exist during the preblastemal or blastemal stage of newt limb regeneration.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Newt Limb Amputations

All procedures involving newts were performed with approval from the University of Utah Institutional Animal Care and Use Committee. Newts were anesthetized by submersion in Tris-buffered 0.1% tricaine (pH 7.3) for 10 min at room temperature and then placed on ice for at least 15 min. Limbs were amputated through the stylopodium midway between the shoulder and elbow using a bone microtome blade. The newts were again placed on ice for 1 hr to reduce bleeding, allowed to recover from anesthesia in shallow water, and then housed in dechlorinated tap water until the limbs were reamputated. The procedure for reamputation was the same as described above, except that the reamputation plane was 1.5–2.5 mm proximal to the distal tip of the regenerating limb.

Differential Display, RLM-RACE, and Sequencing

Differential display, RLM-RACE, and sequencing of the full-length cDNAs were performed as described previously (Vinarsky et al.,2005). Differential display initially yielded a 162-bp cDNA product that represented an RNA that appeared to be upregulated on days 1, 3, and 5 postamputation but was not present in intact limbs. Primers for RLM-RACE included the Ambion 5′-RACE outer and inner primers, 5′-GCTGATGGCGATGAATGAACACTG-3′ and 5′-CGCGGATC- CGAACACTGCGTTTGCTGGCTTTGATG-3′, respectively, and the specific outer and inner primers constructed from the sequence of the 162-bp differentially-displayed cDNA product. The specific outer primer sequence was 5′-GAAAGCGGCTGGGTAGTGCTG-3′ and the inner primer sequence was 5′-CTGCACATGGGCTTCATTGCTAACAC-3′. A 2.1-kb RLM-RACE product produced from the final round of amplification using the inner primers was TA-cloned into pBluescript II and sequenced. Protein sequence comparisons were performed using the BLAST 2 sequences program (Tatusova and Madden,1999), Clustal W (Thompson et al.,1994), and T-Coffee (Notredame et al.,2000). N-glycosylation sites were predicted using the PROSITE program on ExPASy (Gasteiger et al.,2003) and the NetNGlyc 1.0 Server (R. Gupta, E. Jung, and S. Brunak, 2004), while the signal peptide sequence was predicted using the SignalP 3.0 Server (Nielsen and Krogh,1998; Bendtsen et al.,2004).

Northern Blot Analysis

RNA was extracted from regenerating newt limbs at 0, 18, 24, and 72 hr postamputation and at 5, 10, 15, 20, 25, 30, and 40 days postamputation. Two μg of total RNA was used for the Northern blot analysis. RNA samples were separated at 80V for 2–4 hr in a 1% denaturing agarose gel and electrophoresis buffer containing 1× MOPS (20 mM MOPS, 5 mM sodium acetate, 1 mM EDTA, pH 7) and 0.5 M formaldehyde. The separated RNAs were equilibrated and transferred overnight in 20× SSC (3 M NaCl, 0.34 M sodium citrate, pH 7) to Hybond N+ nylon membrane (Amersham Pharmacia), cross-linked using a UV Stratalinker 1800 at 120,000 μJoules, and stained with methylene blue to assess transfer efficiency and RNA loading equivalency between lanes. The blot was hybridized in ExpressHyb (Clonetech, Palo Alto, CA) at 65°C for 4 hr using [32P]-α-dCTP-labeled probes, washed, and placed on Kodak BioMax MS film.

Real-Time RT-PCR

Reverse transcription of total RNA was performed according to the manufacturer's instructions using the Bio-Rad (Richmond, CA) iScript cDNA synthesis Kit and a mixture of poly-dT and random primers. Specific gene primers for real-time PCR were designed using the Beacon Designer 3.01 program (Premier Biosoft International). The primer sequences were as follows: (1) Primer set 1—sense primer, 5′-GCCATCAATGTTATTACGCTACC-3′ and antisense primer— 5′-GCTTCCTCTATCTTACGCTCTC-3′ and (2) Primer set 2—sense primer, 5′-CGGCGAATGTATTTGGGAGTC-3′ and antisense primer, 5′-ACAGGAGGTTGGAACAGATGG-3′. Real-time PCR was performed on an ABI Prism 7700 Sequence Detection System using the iTaq SYBR Green Supermix with ROX (Bio-Rad) for detection of amplified DNA. Amplification conditions included an initial 3-min denaturation step at 95°C followed by 45 cycles of 95°C for 20 seconds, 57°C for 20 sec, and 72°C for 30 sec. A denaturation profile for the amplification products was obtained by heating the samples to 95°C for 15 sec, dropping the temperature to 60°C for 15 sec, and then measuring fluorescence as the temperature slowly ramped at 2% its maximum rate until it reached 95°C. Data were analyzed using the standard curve method (Applied Biosystems, Inc., Foster City, CA) and normalized to the control gene histone acetyltransferase 1. Histone acetyltransferase 1 was selected from a group of 10 genes that had exhibited the least variation between time points based on microarray and/or northern blot analyses. Primer sequences for histone acetyltransferase 1 were as follows: (1) sense primer, 5′-CGTGGAGGCTGATGATATTG-3′ and (2) antisense primer, 5′-GCTCGCTGACTCAATGAAC-3′. Normalized values were then converted to relative values by using the intact newt limb controls as the calibrator.

RNA In Situ Hybridization

Limbs were allowed to regenerate 1, 3, or 5 days. Following this short regenerative period, the newts were again anesthetized and the early limb regenerates were amputated at the shoulder. Limbs were placed in Carnoy's fixative (60% ethanol, 30% chloroform, 10% glacial acetic acid) overnight. The limbs were decalcified, embedded in paraffin, and subjected to analysis by RNA in situ hybridization as previously described (Vinarsky et al.,2005). Digoxigenin probes were used at a concentration of 333 or 375 ng/ml. Following the staining procedure with the NBT/BCIP substrate, the sections were mounted and observed for staining using a Zeiss (Thornwood, NY) Axiophot 2 light microscope. Photographs were taken with a Zeiss Axiocam digital camera.

Preparation of Expression Constructs

A FLAG-tag was placed on the 3′-end of the NvTIMP1 open reading frame by performing PCR with Advantage Taq using the Ambion 5′-RACE inner primer and a FLAG-tagged 3′-primer with the following sequence: 5′-tgtgtgtgtGGCGGCCGCTTACTTATCGT- CGTCATCCTTGTAATCGATCCTGCTGGCTTTAAAGGGGGTTCCA-3′. The lowercase letters represent nonspecific 5′-clamping sequence, underlined sequence is a NotI restriction site, italicized sequence is the complement of the translational stop site, bold sequence is the complement of the FLAG-tag, and regular type is the specific NvTIMP1 primer sequence. This PCR reaction produced a 914-bp product that was then digested with BamHI and NotI and cloned into the pBK-CMV vector (Stratagene, La Jolla, CA) to produce pBK-CMV-NvTIMP1.flag. To achieve higher expression levels, we subcloned a 1-kb HindIII/NotI insert from pBK-CMV-NvTIMP1.flag, which contained the NvTIMP1 open reading frame, into the pCMV-SPORT6 vector (Invitrogen) to produce pCMV-SPORT6-NvTIMP1.flag. Inserts were sequenced to verify that no mutations had occurred during PCR or subcloning.

Purification of NvTIMP1.flag Protein by Affinity Chromatography

pCMV-SPORT6-NvTIMP1.flag was transfected into human embryonic kidney 293H cells using Lipofectamine2000 (Invitrogen, La Jolla, CA), and 100–200 ml of serum-free conditioned DMEM medium was collected from the transfected cells at 24–72 hr post-transfection. Residual cells were removed by centrifugation at 100g. The NvTIMP1.flag protein was purified by affinity chromatography according to the manufacturer's instructions using 300 μl of packed resin that contained conjugated anti-FLAG M2 antibody (Sigma, St. Louis, MO). NvTIMP1.flag was allowed to bind to the resin at 4°C under constant rotational mixing for 14–19 hr. Following elution of the protein with excess FLAG peptide, the concentration of affinity-purified NvTIMP1.flag was estimated by a comparison to known amounts of a FLAG-tagged bovine alkaline phosphatase standard (Sigma) using Western blotting and chemiluminescence detection with the SuperSignal West Femto reagent (Pierce, Rockford, IL).

Reverse Zymography

Reverse zymography was performed on 15% polyacrylamide gels that contained either 0.1% gelatin or 0.2% casein and 20% v/v MMP9 or MMP3/10a conditioned medium, respectively. Purified NvTIMP1.flag was then separated on the polymerized gels by electrophoresis and renatured in two washes of 2.5% Triton-X 100 for 30 min each, followed by a 16–23-hr incubation in development buffer (50 mM Tris-HCl, pH 7.6, 10 mM CaCl2, 200 mM NaCl, 0.05% Brij-35, 1 μM ZnCl2, 0.001% NaN3). Gels were stained with Coomassie blue R250 for 4 hr and then destained in multiple changes of a 40% methanol/10% acetic acid solution until the bands were visible (2–3 hr).

Rat-tail Type I Collagen Assay

Notophthalmus viridescens collagenase (nCol) was preactivated with 1 mM APMA for 20 hr at 30°C (Vinarsky et al.,2005). Native rat-tail type I collagen (0.5 mg/ml, Upstate) was mixed with APMA-preactivated nCol (5 nM) in the presence of varying amounts of NvTIMP1.flag (0–50 nM) and incubated for 24 hr at 30°C in a 40-μl volume. Following the incubation period, an equal volume of 2× reducing SDS loading buffer was added to the reactions and the samples were heated to 95°C for 5 min to denature the proteins. Fifty μl of the reaction sample was added to each lane (each lane contained 12.5 μg of rat-tail type I collagen) of a 4–15% SDS-polyacrylamide gradient gel (Bio-Rad) and electrophoresis was performed at 200 V until the bromophenol blue dye ran off the end of the gel. Gels were stained with Gelcode blue (Pierce) according to the manufacturer's instructions. As a control, native rat-tail type I collagen was mixed with pro-nCol (unactivated nCol) to test whether APMA activation was required for efficient digestion of the type I collagen.

DQ and BODIPY Assays

For testing the inhibitory effects of NvTIMP1, Purified N. viridescens MMP3/10b (1.6 pmol) was mixed with NvTIMP1.flag (1.6 pmol of 4.8 pmol) in the presence of 1 mM APMA and enzyme activation buffer (50 mM Tris-HCl, pH 7.6, 150 mM NaCl, 5 mM CaCl2, 1 μM ZnSO4, and 0.05% Brij-35). Approximately 10 min later, one of the following substrates was added and fluorescence signals were measured on a Perkin-Elmer Fusion α microplate reader at various time points up to 65 hours. Substrates used in this study were DQ-gelatin, DQ-collagen I, or DQ-collagen IV at 100 μg/ml, or BODIPY-Casein at 10 μg/ml (Molecular Probes, Eugene, OR). The final volume for each reaction was 100 μl. Uninhibited reactions were performed in the same manner in the absence of NvTIMP1.flag.

Cell Proliferation Assays

C2C12 cells (mouse myoblast), IMR-90 cells (human fetal lung fibroblasts), NIH/3T3 cells (mouse embryonic fibroblasts), and MG-63 cells (human osteosarcoma cell line) were purchased from the American Type Culture Collection (ATCC, Rockville, MD) and were grown at 37°C/5% CO2 in DMEM (high glucose, 4.5 g/l) supplemented with 4 mM glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% FBS. Cells were plated in quadruplicate in 100 μl volumes onto 96-well plates at the following densities: (1) C2C12 and NIH/3T3 cells, 5,000 cells/well and (2) IMR-90 cells, 4,000 cells/well. Cells were allowed to attach overnight and were then rinsed once with PBS and synchronized for 20–24 hr with medium containing no serum, 0.25% FBS, or 1% horse serum. Following synchronization, some cells were assayed to establish a baseline for proliferation by adding 10 μl of WST-8 reagent (Dojindo) to each well, incubating cells for 2 hr at 37°C/5% CO2, and then collecting spectrophotometric data at 450 nm using an ELISA plate reader. The remainder of the cells were treated with medium containing the same amount of serum used to synchronize the cells, supplemented with no protein, recombinant NvTIMP1.flag (1 to 1,000 ng/ml), human TIMP1 (10 to 100 ng/ml), mouse TIMP1 (10 to 100 ng/ml), or 100 ng/ml IGF1A. After two days, 10 μl of WST-8 reagent was added to each well and cell proliferation was assessed as described above for the baseline readings.

Newt A1 limb cells were a gift from Jeremy Brockes (University College London). They were grown on gelatin-coated plates at 24°C/2% CO2 in 60% EMEM with 10% FBS, 4 mM L-glutamine, 1 mM sodium pyruvate, and 100 U/ml penicillin-100 μg/ml streptomycin. Cells were plated in quadruplicate onto 96-well plates in 100-μl volumes at a density of 2,000 cells/well and allowed to attach for at least 6 hr. Cells were then washed once with PBS and induced into a quiescent phase by 20–24-hr treatment in medium containing 1% horse serum, 0.5% fetal bovine serum, or no serum. Baseline cell readings were obtained as described above, except that the WST-8 reagent was added 7 hr before taking the spectrophotometric reading and the reactions were allowed to proceed at 24°C/2% CO2. Test cells were treated in medium containing the same amount of serum used to induce quiescence, supplemented with no protein, NvTIMP1.flag, human TIMP1, mouse Timp1, or IGF1A at the concentrations used for the mammalian cell assays. After 72 hr, the cells were fed with fresh supplemented medium, and 72 hr later, 10 μl of WST-8 reagent was added to each well. Spectrophotometric readings were taken 7 hr later.

ANOVA was used to compare the differences between the means when three or more concentrations of a particular protein were used in the proliferation assays (all NvTIMP1.flag, human TIMP1, and mouse Timp1 comparisons). Student's t-test was used when only two concentrations were being compared (IGF1A comparisons).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Dr. Jeremy Brockes for the newt A1 cells. We also thank Dr. Dean Li for use of the Axiophot 2 microscope, Dr. Tom McIntyre for use of the Perkin-Elmer Fusion α microplate reader, and Dr. Guy Zimmerman for use of the ELISA plate reader.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES