Molecular Vision 2005; 11:1071-1082 <http://www.molvis.org/molvis/v11/a126/> Received 12 May 2005 | Accepted 5 December 2005 | Published 9 December 2005

Download Reprint

Use of transduction proteins to target trabecular meshwork cells: outflow modulation by profilin I

Azucena Gómez-Cabrero,¹ Nuria Comes,¹ Javier González-Linares,² Joaquín de Lapuente,² Miquel Borras,² Jordi Pales,¹ Arcadi Gual,¹ Xavier Gasull,¹ Miguel Morales¹
 
(The first two authors contributed equally to this publication)
 
 

¹Laboratory of Neurophysiology, Department of Physiological Sciences I-August Pi i Sunyer Biomedical Research Institute (IDIBAPS), Faculty of Medicine, and ²Experimental Toxicology and Ecotoxicology Unit, Parc Científic de Barcelona, University of Barcelona, Barcelona, Spain

Correspondence to: Miguel Morales, PhD, Laboratori de Neurofisiologia, Facultat de Medicina, UB-IDIBAPS, Casanova 143, E-08036 Barcelona, Spain; Phone: (34) 93 4024519; FAX: (34) 93 4035295; email: miguelmorales@ub.edu

Abstract

Purpose: Fusion proteins containing a protein transduction domain (PTD4) are able to cross biological membranes. We tested the applicability of the protein transduction method for study of the aqueous humor trabecular outflow pathway by targeting the actin cytoskeleton, which is known to be involved in outflow facility regulation.

Methods: Expression vectors useful for generating fusion proteins with the PTD4 domain and the actin-binding protein Profilin I were constructed. The transductional and functional properties of these proteins were tested in bovine trabecular meshwork cells in culture. The effects of PTD4-Profilin I on outflow facility were evaluated in perfused bovine anterior segments. PTD4-β-galactosidase was used to visually check correct delivery of fusion proteins to trabecular meshwork cells.

Results: The fusion proteins generated were characterized by western blot. Immunocytochemistry experiments showed intracellular staining for PTD4-Profilin I in trabecular meshwork cells in culture. The fusion protein was found in the cytoplasm associated with actin filaments and in the leading edge of the cellular membrane. In contrast, control Profilin I, without the PTD4 domain, was unable to cross the cell membrane. In perfused anterior segments, 2 μM PTD4-Profilin I increased trabecular outflow facility in a reversible manner, while Profilin I had no significant effect. Anterior segments perfused with PTD4-β-galactosidase showed positive staining in the trabecular meshwork tissue.

Conclusions: Protein transduction technology is a valuable tool for targeting trabecular meshwork tissue, not only for performing physiological studies, but also as a potential drug-delivery method. Profilin I action on the actin cytoskeleton further reinforces the importance of this structure in outflow facility regulation.

Introduction

In the eye, most of the aqueous humor exits the anterior chamber through the trabecular meshwork (TM) and Schlemm's canal. For many years, the TM was believed to be merely a passive filter. Today this tissue, located in the iridocorneal angle, is considered an active network of cells able to modulate its permeability and, therefore, the aqueous humor outflow rate [1-3]. Nevertheless, TM physiology is still largely unknown, probably due to the location, structure, and heterogeneity of the tissue [4].

It has been reported that various mechanisms contribute to aqueous humor (AH) outflow regulation, including contraction/relaxation of both TM and ciliary muscle [1,5], cell volume regulation [6-9], membrane stretch [10], changes in cell shape [11-13], extracellular matrix composition and remodeling [13-15], pore formation in Schlemm's canal endothelium [16,17], and gene expression changes [18-20]. Several authors have demonstrated the involvement of the actin cytoskeleton in AH outflow modulation [13] upon observing that latrunculins and other actin-depolymerizing drugs increased outflow facility [21-23]. In fact, the Rho/Rho-kinase pathway [24,25] appears to regulate actin polymerization, and reorganizes the actomyosin cytoskeleton which finally modulates TM contractility and aqueous outflow facility [20,24,26].

Actin filaments are in a dynamic equilibrium between the filamentous and monomeric forms. The elongating filaments consist of ATP-actin. Along the filament, due to actin's intrinsic ATPase activity, the ATP is slowly hydrolyzed to ADP-actin in the older part of the filament, from where it can be released. Profilin I (Pfn I) is a ubiquitous actin-binding protein with ATP nucleotide exchange properties [27]. In vitro studies have identified three major functions for profilins. First, Pfn I binds and sequesters actin monomers in a 1:1 ratio, thereby decreasing the concentration of free actin. Second, due to its ATP nucleotide exchange function, Pfn I can restore monomeric ADP-actin with ATP, thereby replenishing the pool of actin-ATP in the cell. And third, the profilin-ATP-actin complex can interact with the growing end of the actin filament and release the ATP-actin monomer which is then added to the filament, promoting filament elongation (for a review see [27,28]).

In the present report we evaluate an in vivo protein transduction technique in combination with perfusion of ocular anterior segments to deliver Pfn I to trabecular meshwork cells. We also study the ability of Pfn I to modulate the actin cytoskeleton and its involvement in AH outflow regulation.

Most of the information used to assess TM cell function has been obtained from cell cultures. It is still difficult to understand how the TM works in vivo, and the mechanisms by which different cellular processes modify the tissue's permeability. To evaluate TM function in situ in response to different experimental treatments, several techniques have been used, including in vivo measurements of trabecular outflow facility [29,30], perfusion of anterior segments [31-33], or gene delivery to TM cells using adenovirus [34-37]. Transduction of active proteins into cells in vivo is another powerful experimental approach for the study of cell physiology. Protein transduction was first demonstrated with the transactivating transcriptional activator (TAT) protein from HIV [38]. The sequence responsible for the transduction properties consists of a highly basic region of only 11 amino acid residues from the HIV-TAT protein (YGRKKRRQRRR). The methodology to generate transducible fusion proteins was greatly improved by Dowdy and coworkers and has been used in a wide variety of mammalian cells to deliver different proteins or peptides [39,40]. We generated a series of fusion proteins containing a modified version of the TAT transduction domain; the PTD4 (protein transduction domain 4; YARAAARQARA) allows them to cross biological membranes [41]. Using this novel approach, we now describe the effects of PTD4-Profilin I (PTD4-Pfn I) on trabecular outflow regulation.

The present report shows that protein transduction technology is a valuable tool for studying aqueous humor outflow physiology, and an interesting delivery method to target TM cells. Our data indicate that the PTD4-Pfn I fusion protein effectively penetrates TM cells, increasing outflow facility in perfused anterior segments.

Methods

Expression vector generation and cloning

Oligonucleotides containing the transduction sequence PTD4 (YARAAARQARA) and several glycine residues in each side for a free rotation between domains, (pre-PTD4-MCS-F and pre-PTD4-MCS-R; Table 1, Figure 1), were hybridized by heating to 95 °C and slowly cooling down to room temperature for at least 2 h. The resulting pre-PTD4-MCS adaptor was cloned into the expression vector pRSET-A (Novagen, Madison, WI) between NheI and XhoI restriction sites to generate the pRSETA-PTD4-MCS vector, keeping both NheI and XhoI restriction sites (Figure 1). In order to keep fusion proteins free from unnecessary domains, the restriction procedure eliminated the tag-T7pol, the Tag Xpress^TM and the EK proteolysis site. In the new expression vector (pRSETA-PTD4-MCS), only the tag T7pol was rebuilt by the pre-PTD4-MCS-F and pre-PTD4-MCS-R oligonucleotides, while the multiple cloning site (MCS) was left intact (Figure 1). The pRSETA-mod vector, used to generate control proteins, was created missing the Tag Xpress^TM fragment, the EK proteolysis site, and the PTD4 sequence. For this purpose, pre-MCSnoPTD4-F and pre-MCSnoPTD4-R oligonucleotides were used following the procedure described (Table 1, Figure 1). The resulting pre-MCSnoPTD4 adaptor was cloned using the same strategy. Both vectors retained pRSET-A's resistance to ampicillin (Figure 1).

For β-galactosidase (β-Gal) cloning, two oligonucleotides, β-GalXC-F and β-GalSE-R (Table 1, Figure 1) were used to amplify the β-galactosidase encoding sequence from pCMV-β-Gal (GenBank U02451; Clontech, Mountain View, CA) kindly provided by Miguel Chillón (UAB, Spain). PCR reactions (50 μl volume in 0.2 ml PCR tube) were performed with 1 μM of each oligonucleotide, 1 mM MgCl₂, 2 mM each of dATP, dCTP, dGTP, and dTTP (dNTP mix), 0.2 μg pCMV-β-Gal vector, 2.5 U Accuzyme polymerase in 1X Accuzyme Reaction Buffer with 2% DMSO (Bioline, Teknovas, Spain). Reactions were carried out with an initial 5 min long denaturing step at 96 °C, 30 cycles at 96 °C for 1 min, 60 °C for 1 min and 72 °C for 7 min, and a final elongation step of 15 min at 72 °C. Amplification products were separated by electrophoresis on 0.8% agarose in 1X TAE buffer and then visualized by ethidium bromide staining. A fragment of 3,072 bp corresponding to the β-Gal coding sequence was cut and isolated (Perfect Prep Gel Cleanup kit; Eppendorf, Madrid, Spain). Cohesive ends were added to the purified fragment using the fragment as a template in a PCR reaction (50 μl) performed only with the dNTP mix and 2.5 U Taq polymerase in 1X Taq Reaction Buffer. The resulting fragment was isolated and cloned into the pGEM-T vector (Stratagene, La Jolla, CA). After sequence confirmation, β-Gal cDNA was subcloned into pRSETA-mod and pRSETA-PTD4-MCS vectors between XhoI and EcoRI sites, generating the expression vectors pRSETA-mod-β-Gal and pRSETA-PTD4-β-Gal (Figure 1).

Two oligonucleotides (XhoI-Profilin and Profilin-HindIII; Table 1) were used to amplify human PfnI cDNA (GenBank NM_005022) using the CMV-C1-PfnI-GFP vector as a template (kindly provided by Dr. Hitomi Mimuro, University of Tokyo, Japan) [42]. PCR reactions (50 μl) were performed with 1 μM of each oligonucleotide, 1 mM MgCl₂, 2 mM dNTP mix, 0.5 μg CMV-C1-PfnI-GFP vector, and 2.5 U Accuzyme polymerase in 1X Accuzyme Reaction Buffer. Reactions were carried out with an initial 5 min long denaturing step at 96 °C, 25 cycles at 96 °C for 3 min, 68 °C for 1 min with a temperature ramp of -0.2 °C each cycle and 72 °C for 1.5 min, and a final elongation step of 7 min at 72 °C. A fragment of 449 bp was isolated and cloned into the EcoRV restriction site of pBluescript SK^- (Stratagene, La Jolla, CA). After sequence confirmation, Pfn I cDNA was subcloned between XhoI and HindIII sites, generating the vectors pRSETA-mod-Pfn I and pRSETA-PTD4-Pfn I (Figure 1).

The oligonucleotides used were purchased from Sigma-Genosys (Haverhill, UK). All constructions were transformed into competent Escherichia coli DH5α bacteria growing in LB medium. Prior to protein expression, positive clones were isolated and in-frame cloning was confirmed by sequencing.

Expression and purification of proteins

Chemically competent Escherichia coli BL21-PLys bacteria were transformed with pRSETA-PTD4-β-Gal, pRSETA-mod-β-Gal, pRSETA-PTD4-Pfn I, and pRSETA-mod-Pfn I. Protein expression was induced by adding 1 mM IPTG (Sigma, Madrid, Spain) at 37 °C for 6 h with continuous agitation. Bacterial pellets were isolated by centrifugation, resuspended in phosphate-buffered saline (PBS), and lysated by a freezing and unfreezing protocol in liquid N₂ followed by sonication on ice in the presence of DNase and a protein inhibitor cocktail (Sigma). Cellular lysates were resolved by centrifugation. The soluble fraction was loaded onto a 25 ml column packed with Ni-NTA resin (Quiagen, Hilden, GmgH) in purification buffer (0.5 M NaCl, 20 mM Na₂HPO₄, pH 7.4). The column was washed with 20 and 50 mM and eluted with 250 mM imidazole in purification buffer up to 20-25 ml. Purification efficiency was assessed by acrylamide gel electrophoresis and Coomassie staining. Buffer exchange and concentration of eluted proteins was performed by centrifugation in Amicon Ultra-15 10000 MWCO centrifugal filters (Millipore Iberica, Madrid, Spain) with PBS. Proteins were frozen in liquid N₂ and stored at -80 °C in 10-15% glycerol-PBS. After being defrosted, proteins were kept at 4 °C for up to one week.

Bacteria and proteins were handled following the Safety Guidance for Laboratory Personnel Working with TAT transduction domains [43].

Western blot analysis

Samples were electrophoresed in 8% sodium dodecyl sulfate-polyacrylamide gel and transferred to nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA) following the Laemmli method. Nonspecific protein binding sites were blocked with a solution containing 3% BSA and 0.1% Tween-20 in Tris-buffered saline (TBS; 20 mM Tris-HCl [pH 7.4] and 137 mM NaCl; TBT-BSA) for 30 min. Membranes were then incubated with rabbit anti-Profilin IgG (1:1000; Cytoskeleton, Denver, CO) and mouse anti-T7-tag IgG antibodies (1:10000; Novagen) in TBT-BSA for 1 h. Membranes were washed 6 times with TBT-BSA and incubated with horseradish peroxidase-conjugated anti-mouse IgG (1:5000) or horseradish peroxidase-conjugated anti-rabbit IgG antibodies (1:5000; Jackson ImmunoResearch Laboratories, Westgrove, PA) in TBS-T (0.1% Tween-20 in TBS) for 1 h. Finally, the membranes were washed 5 times in TBS-T and twice in TBS. Detection was performed by a chemiluminescence method with Immuno-HRP^TM Star Substrate (Bio-Rad).

Bovine trabecular meshwork cell culture

Bovine TM cells (BTM) were cultured using a modification of the technique described by Stamer et al. [44]. As previously described [5], bovine TM strips were digested with 2 mg/ml collagenase (Sigma) and 0.5 mg/ml bovine serum albumin (BSA; Sigma) at 37 °C for 2 h. After mechanical disruption, the supernatant was collected, centrifuged, resuspended, and seeded in culture flasks containing Dulbecco's Modified Eagle's Medium (DMEM; Bio-Whitaker, Barcelona, Spain) plus 10% fetal bovine serum, 100 mg/ml L-glutamine (Sigma), 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2.5 μg/ml amphotericin-B (Bio-Whitaker). Cell growth was observed 2-4 days after seeding, and culture medium was changeed three times a week. Confluence was reached after 12-15 days in culture. After confluence, cells were passaged using Trypsin-EDTA (Bio-Whitaker). Cells from this first passage to the third one were used.

Anterior segment perfusion

Eyes from 3- to 6-month-old cows were obtained at the local abattoir 0.5 to 2 h after killing, and kept in PBS at 4 °C for no more than 1.5 h. Isolation and perfusion of bovine anterior segments were performed as previously described [5,8,32]. Bovine anterior segments, located in their respective chambers together with force transducers (Letica, Barcelona, Spain) and the tubing system, were placed in an incubator (Selecta, Barcelona, Spain) at 37 °C and 5% CO₂. Perfusion was carried out with DMEM. The pressure of the artificial anterior chamber was monitored and recorded throughout the experiment with a pressure transducer (9162-0; Mallinckrodt, Northampton, UK) and was maintained with a suspended reservoir at a constant pressure of 10 mm Hg. Outflow facility (C), calculated as the ratio between flow and perfusion pressure (μl/min/mm Hg) was averaged in 15 min periods (mean of 450 data points for each period; sampling rate 0.5 Hz). A total of 20 periods of 15 min each were obtained for the whole protocol. Baseline facility (C₀) was calculated during the first 90 min period of stable C. Only anterior segments with baseline outflow facility values between 0.3-1.3 μl/min/mm Hg were used. Moreover, anterior segments that during the baseline period presented more than a 10% variability in outflow facility values were rejected as well [32]. Whenever a different experimental condition was implemented or drug added to the perfusion medium, the tubes and anterior chamber were flushed and refilled with the new medium. This change was made by rapidly replacing the contents of the artificial anterior chamber by opening the exit needle until 200% of the volume had been exchanged; this exchange was always made at a pressure below 10 mm Hg. Recording of C measurements started after flow stabilization. The perfusion procedure was carried out using a protocol with three different periods: perfusion with control DMEM for 90 min to establish C₀; perfusion with 2 μM Pfn I or 2 μM PTD4-Pfn I for 120 min to determine outflow facility changes produced by these proteins, and finally a 90 min period with DMEM to return to baseline conditions. After the last period, 2 μM latrunculin A (Lat A; Molecular Probes, Eugene, OR) was added to the perfusion medium as a positive control. Perfusions (60 or 90 min) with PTD4-β-Gal or β-Gal were done using the same experimental setup.

Immunocytochemistry

BTM cells were cultured on 12 mm diameter glass coverslips to a subconfluent monolayer. The recombinant purified proteins (PTD4-Pfn I and Pfn I) were added to the culture medium containing serum. After treatment for different time periods, cells were washed twice with PBS, fixed with 4% paraformaldehyde in PBS for 30 min, and then further washed (4 times). After blocking and permeabilization with a PBS solution containing 0.2% Triton X-100, 2% BSA, and 2% goat serum for 30 min, cells were incubated at room temperature with a mouse anti-T7-tag IgG antibody (1:1000; Novagen) in blocking solution for 1 h. Cells were next washed 4 times with PBS, followed by a second incubation with a goat Cy3 conjugated anti-mouse IgG antibody (1:5000; Jackson ImmunoResearch) and Oregon-green phalloidin (1:100; Molecular Probes). Finally, cells were washed 5 times with PBS and preserved with Mowiol mounting solution for fluorescence microscopy and confocal imaging. The secondary antibody was tested for cross-reactivity in BTM cells, incubating them in the absence of primary antibody. Confocal images were acquired using a Leica TCS NT laser scanning confocal microscope (Leica Microsystems Heidelberg GmbH, Mannheim, Germany) with Krypton-Argon laser attached to a Leica DMIRB upright microscope. All images were obtained using 63x oil immersion objective lens (NA 1.4) and the confocal pinhole set at 1 Airy unit. All optical sections were of 0.2 μm. Photographs were taken from representative fields. Pictures were analyzed using Leica software and Adobe Photoshop (Adobe Software, Mountain View, CA).

X-gal staining

After perfusion with PTD4-β-Gal or β-Gal proteins, bovine ocular anterior segments were washed with PBS and fixed with glutaraldehyde 2.5% for 1 h. Anterior segments were cut in quarters (about 2 cm wide), and each quadrant incubated with X-gal staining solution containing 1 mg/ml X-gal, 5 mM K₃Fe(CN)₆, 5 mM K₄Fe(CN)₆, 2 mM MgCl₂, 0.01% sodium deoxycholate, and 0.02% Nonidet-P40 (Sigma) [45] in PBS at 37 °C until development of blue staining (about 1 h). After X-gal staining, selected quadrants were immersed overnight in 4% paraformaldehyde in PBS, then embedded in paraffin, and sagittally oriented 5 μm sections of the chamber angle region were cut and counterstained with hematoxylin and eosin stain.

Data analysis

One-way ANOVA with Bonferroni post hoc tests were used to evaluate statistical differences between control perfusion with Pfn I and PTD4-Pfn I. An α level of 0.05 was selected for statistical significance.

Results

Construction of expression vectors, cloning, and protein preparation

To assess the efficiency of protein transduction as a tool to introduce active fusion proteins in TM cells, we constructed two bacterial expression vectors, pRSETA-mod and pRSETA-PTD4-MCS, which facilitate the production of fusion proteins. The expression vector omitting the PTD4 domain (pRSETA-mod) was used to express a control protein without the transduction domain. β-Gal and human Pfn I cDNAs were amplified and cloned in both expression vectors. Induction for 6 h of BL21-PLys bacteria transformed with the expression vectors resulted in overexpression of the corresponding recombinant proteins. The T7pol Tag antibody was used to recognize recombinant β-Gal and PTD4-β-Gal (Figure 1B). Fusion proteins Pfn I and PTD4-Pfn I were recognized as expected by antibodies against profilin and T7pol tag, as shown in Figure 1C,D.

Transduction into trabecular meshwork cells

To determine the fusion protein's ability to transduce into BTM cells, 2 μM PTD4-Pfn I or Pfn I were added directly to the culture medium containing 10% FBS. Cells were fixed at different time intervals (10, 30, and 60 min) and processed for immunocytochemistry. Single confocal microscopy images show intracellular localization at the same focal plane as actin stress fibers (Figure 2). A weak but consistent staining was clearly visible after only 10 min (Figure 2A). Longer incubations markedly increase the amount of intracellular protein (Figure 2B,C). After 60 min, a large amount of protein was present around the cellular edge or accumulated in a perinuclear position in most cells. Although not quantified, fluorescence intensity suggests that the amount of internalized protein is time-dependent. Pfn I was never found inside the cells in any of the intervals studied (Figure 2D). In both groups and after 1 h treatment, protein was also clearly present at the top focal plane, probably corresponding to the cell membrane. Confocal microscopy optical slices from the top and bottom of the cell confirmed that both proteins were present at the membrane level, but only PTD4-Pfn I was located intracellularly (Figure 3A). We noted that after long incubation times (over 30 min), a large amount of PTD4-Pfn I sticks to the coverslip. Nevertheless, the ratio of intracellular to background signal was relatively high. In contrast, Pfn I was easily washed out during the fixation process.

Endogenous Pfn I appears concentrated in regions with high actin dynamics, such as the leading lamellae and ruffles [28]. After 60 min incubation, PTD4-Pfn I immunolabeling appears to be more concentrated at the leading edge of the membrane lamellae where it colocalized with the actin cytoskeleton (Figure 3B).

PTD4-Pfn I increased outflow facility in perfused anterior segments

The trabecular meshwork is thought to regulate aqueous humor outflow resistance by cytoskeletal rearrangement of its cells. In order to study the ability of PTD4-Pfn I to transduce into the whole tissue and to modulate actin dynamics, we used a constant pressure perfusion system as a functional test. This system allowed us to measure trabecular outflow facility while perfusing control protein (Pfn I) or PTD4-Pfn I. The baseline outflow facility was 0.66±0.1 μl/min/mm Hg. Perfusion with 2 μM Pfn I (n=6) for 120 min did not significantly modify outflow facility (Figure 4A,B), which remained stable after returning to baseline conditions. To validate tissue functionality, 2 μM Lat A was used as a positive control at the end of the perfusion protocol (Figure 4B, bottom). In all eyes tested, Lat A significantly increased outflow facility, as previously reported [21]. In contrast, when 2 μM PTD4-Pfn I was added to the perfusion medium (n=5; Figure 4A), outflow facility significantly increased after 30 min (compared to control perfusion; p<0.05), reaching a maximum at 45 min (19%; p<0.001). Outflow facility remained high in the presence of PTD4-Pfn I, and returned to baseline values once the protein was removed from the perfusion medium. Again, to validate outflow facility modulation by PTD4-Pfn I, 2 μM Lat A was applied at the end of the experiment. A significant outflow facility increase was observed in all cases in the presence of this drug (Figure 4B, top).

Fusion protein localization in perfused anterior segments

To evaluate protein delivery to the trabecular meshwork tissue, PTD4-β-Gal or β-Gal at 2 μM concentration was tested in the perfusion system under the same experimental conditions as previously described. In Figure 5A, an anterior segment perfused with PTD4-β-Gal during 60 min was fixed and stained for β-Gal activity. Prominent staining was seen in the TM area as well as parts of the sclera and corneal endothelium. In a second set of experiments, after 90 min perfusion with PTD4-β-Gal or β-Gal, anterior segments were perfused for 90 min with control media to wash extracellular protein (Figure 5B,C). Then, anterior segments were removed and stained for β-Gal activity. Paraffin sections of the angle tissues showed a prominent blue staining at the trabecular meshwork area (Figure 5B) compared with the control (Figure 5C). The most intense staining was present in external cellular layers of the trabecular meshwork (Figure 5A,B), confirming positive delivery of the transduction proteins into this tissue. As a control for possible effects due to the transduction technique, outflow facility was monitored; neither PTD4-β-Gal nor β-Gal modified outflow facility significantly (Figure 5D).

Discussion

This study shows that protein transduction techniques are useful for the study of TM cell physiology and the evaluation of outflow pathway mechanisms. Manipulation of cells has previously been achieved by several techniques such as expression vectors, microinjection, gene deletion, and viral infection. A different kind of approach has taken advantage of the transduction properties of HIV TAT transcription factor [38]. Since 1988, several transduction domains have been identified and fused to different proteins, allowing them to cross biological membranes in a time- and concentration-dependent manner. The process is based on a receptor-independent mechanism whose limits in terms of protein size and function are not known [43]. Transduction of fusion proteins can be used to introduce inhibitory peptides or mutated exogenous proteins that may compete with the cell's endogenous proteins, allowing study of their cellular functions in vivo, while circumventing the problems derived from deleting a gene in the whole animal [46].

To generate transduction fusion proteins, we modified a commercial expression vector with the addition of a transduction domain (PTD4). The PTD4 is an optimized and more efficient version of the wild type TAT domain, where the putative α-helix structure of the TAT domain has been strengthened by substituting several residues with alanines [41]. The exact mechanism of protein transduction across cellular membranes remains unclear. Some authors have suggested that they cross the membrane directly by an unknown receptor-independent mechanism [47]. Recently, it has been proposed that TAT fusion proteins require binding or interaction with membrane lipid rafts (directly with cholesterol or proteoglycans) followed by rapid internalization by macropinocytosis [48]. Following internalization, most of the protein is thought to be released into the cytosol after a fall in pH inside the macropinosome, which destabilizes its membrane integrity [48]. The PTD4 domain presumably has the same transduction mechanism.

In agreement with the afore described model, the intracellular location of the recombinant protein indicates that the protein has been released into the cytoplasm. The early effect on outflow facility described in this report (15-30 min) suggests that the PTD4-PfnI must be released rapidly from the macropinosomes, although we cannot rule out the possibility that some of the protein may transduce directly by crossing the membrane. Confocal microscopy fluorescence images show that both control and transduced proteins are located on the cellular membrane, but only the PTD4-Pfn I has a clear intracellular localization in the same focal plane as actin stress fibers. It is possible that trabecular meshwork cells internalize some of the protein due to their phagocytic activity [49,50], which would explain the presence of both proteins (Pfn I and PTD4-Pfn I) in submembranous compartments. This pattern was commonly found at longer incubation times (Figure 3B), while fluorescence close to the membrane was absent at shorter time periods (data not shown). We also noticed that the PTD4-Pfn I construct is stickier than the control, and protein fluorescence is sometimes present as background in immunocytochemistry images (Figure 2). Histological sections of the chamber angle tissues from anterior segments perfused with PTD4-β-Gal show β-Gal activity concentrated in the cellular layers of the TM, as was expected due to the AH flow through this tissue. PTD4-Pfn I is likely to follow the same pathway. Some β-Gal activity is also present in scleral tissues surrounding the artificial anterior chamber, but not in deep tissue layers, supporting the view that AH flow predominantly delivers the fusion protein to the outflow pathway.

After internalization, protein staining overlaps with actin polymerization areas close to the leading edge of the lamellae. It has been reported that endogenous profilin is highly concentrated in these cellular areas [51]. Since the expected localization of the fusion protein is a good criterion for its functionality, we can hypothesize that PTD4-Pfn I is a functional protein and may be involved in actin dynamics. The effects on trabecular outflow, similar to those induced by actin-depolymerizing drugs, support this argument [21,22]. In fact, the control protein, which was unable to transduce the cells, had no effect on outflow facility. This indicates that the effect of PTD4-Pfn I on outflow facility is not an extracellular function of Pfn I and is not an effect of the PTD4 peptide either, since perfusion of PTD4-β-Gal did not induce any outflow facility changes.

Studies in cell lines have shown that microinjection of high concentrations of Pfn I (50 μM to 1 mM) produces depolymerization of actin stress fibers [52,53] and increases the off rate of actin monomers in vitro [54]. These results are in agreement with the avidity of Pfn I for actin monomers. In addition, cells injected with lower concentrations of profilin/actin complexes displayed alterations in actin filament formation, as seen by a change in the morphology of membrane lamellae [55]. In contrast, in the presence of preformed actin filaments, lower concentrations (5 to 10 μM) of Pfn I increase the rate of actin polymerization in vitro [54].

An extensive depolymerization of the actin cytoskeleton was found after treatment of TM cells with Lat A or other depolymerizing agents (unpublished and [56]). In spite of the fact that Pfn I sequesters actin monomers, we did not detect substantial changes in actin stress fiber morphology in cultured cells after PTD4-Pfn I addition at any time interval analyzed. Because Pfn I affinity for actin monomers is much lower than Lat A (Kd of 40 nM and 0.6 μM, respectively), it is possible that no evident structural changes can be detected at the concentration tested. Experiments using higher concentrations and longer incubation times are required to clarify this point.

It has been proposed that a population of TM cells have smooth-muscle contractile properties, and that modulation of their contractile state may influence outflow facility [1]. Agonists known to decrease trabecular outflow facility and contract TM tissue (e.g., endothelin-1, carbachol) [1,57] activate the Rho/Rho-kinase pathway through Gα_12/13 proteins [58]. In addition, inhibition of cellular contractility by Rho kinase (ROCK) inhibitors increases outflow facility [59]. Moreover, drugs that directly affect the cytoskeleton network, such as cytochalasin B or latrunculin A or B, have also been shown to increase outflow facility by direct disassembly of the actin cytoskeleton and associated cellular adhesions in the trabecular meshwork [13,56]. In nonmuscle cells, the intracellular concentration of actin/profilin has been estimated to be around 3-20 μM [55]. Assuming a free equilibrium for PTD4-Pfn I, a maximum internal PTD4-PfnI concentration of 2 μM could be achieved, although this concentration may actually be reduced by a limited diffusion through the TM tissue. Since in culture PTD4-Pfn I appears to be mainly concentrated in the cellular edges, it is possible that a similar intracellular distribution may occur in the tissue.

At this point, we can only speculate about the precise mechanism by which Pfn I increases outflow facility. The functional properties of Pfn I and the effects of actin-depolymerizing drugs on outflow facility suggest that Pfn I may exert its effects through a similar mechanism. It might be hypothesized that an increase in the intracellular concentration of Pfn I would modify the actin polymerization/depolymerization equilibrium and interfere in the normal function of the actomyosin network. This may induce a relaxation of the tissue and therefore, an increase in outflow facility.

Further experiments are required to achieve a better understanding of Pfn I effects on TM cell actin dynamics and outflow facility. Beside highlighting the interesting effects of Pfn I on outflow facility, the present report demonstrates that protein transduction is a useful technique for inserting functional proteins into TM cells and for the study of AH outflow regulation.

Acknowledgements

We thank Dr. Imanol Martinez-Padrón for his critical review of the manuscript. We also thank Dr. J. Freeman for English grammar revisions. Work supported by SAF V2002-PN03517-O, BFI2002-01190, FIS 031495, Spain.

References

1. Wiederholt M, Thieme H, Stumpff F. The regulation of trabecular meshwork and ciliary muscle contractility. Prog Retin Eye Res 2000; 19:271-95.

2. Wiederholt M, Stumpff, F. The trabecular meshwork and aqueous humor reabsorption. In: Civan MM, editor. The eye's aqueous humor. From secretion to glaucoma. New York: Academic Press; 1998. p. 163-201.

3. Llobet A, Gasull X, Gual A. Understanding trabecular meshwork physiology: a key to the control of intraocular pressure? News Physiol Sci 2003; 18:205-9.

4. Lutjen-Drecoll E. Functional morphology of the trabecular meshwork in primate eyes. Prog Retin Eye Res 1999; 18:91-119.

5. Llobet A, Gual A, Pales J, Barraquer R, Tobias E, Nicolas JM. Bradykinin decreases outflow facility in perfused anterior segments and induces shape changes in passaged BTM cells in vitro. Invest Ophthalmol Vis Sci 1999; 40:113-25.

6. O'Donnell ME, Brandt JD, Curry FR. Na-K-Cl cotransport regulates intracellular volume and monolayer permeability of trabecular meshwork cells. Am J Physiol 1995; 268:C1067-74.

7. Al-Aswad LA, Gong H, Lee D, O'Donnell ME, Brandt JD, Ryan WJ, Schroeder A, Erickson KA. Effects of Na-K-2Cl cotransport regulators on outflow facility in calf and human eyes in vitro. Invest Ophthalmol Vis Sci 1999; 40:1695-701.

8. Soto D, Comes N, Ferrer E, Morales M, Escalada A, Pales J, Solsona C, Gual A, Gasull X. Modulation of aqueous humor outflow by ionic mechanisms involved in trabecular meshwork cell volume regulation. Invest Ophthalmol Vis Sci 2004; 45:3650-61.

9. Mitchell CH, Fleischhauer JC, Stamer WD, Peterson-Yantorno K, Civan MM. Human trabecular meshwork cell volume regulation. Am J Physiol Cell Physiol 2002; 283:C315-26.

10. Gasull X, Ferrer E, Llobet A, Castellano A, Nicolas JM, Pales J, Gual A. Cell membrane stretch modulates the high-conductance Ca2+-activated K+ channel in bovine trabecular meshwork cells. Invest Ophthalmol Vis Sci 2003; 44:706-14.

11. Gills JP, Roberts BC, Epstein DL. Microtubule disruption leads to cellular contraction in human trabecular meshwork cells. Invest Ophthalmol Vis Sci 1998; 39:653-8.

12. Tumminia SJ, Mitton KP, Arora J, Zelenka P, Epstein DL, Russell P. Mechanical stretch alters the actin cytoskeletal network and signal transduction in human trabecular meshwork cells. Invest Ophthalmol Vis Sci 1998; 39:1361-71.

13. Tian B, Geiger B, Epstein DL, Kaufman PL. Cytoskeletal involvement in the regulation of aqueous humor outflow. Invest Ophthalmol Vis Sci 2000; 41:619-23.

14. Johnson M, Erickson KA. Mechanisms and routes of aqueous humor drainage. In: Albert DM, Jakobiec FA, editors. Principles and practice of ophthalmology. Philadelphia: WB Saunders; 2000. p. 2577-95.

15. Lutjen-Drecoll E, Rohen JW. Morphology of aqueous outflow pathways in normal and glaucomatous eyes. In: Ritch R, Shields MB, Krupin T, editors. The glaucomas. St. Louis: Mosby; 1996. p. 89-123.

16. Johnson M, Chan D, Read AT, Christensen C, Sit A, Ethier CR. The pore density in the inner wall endothelium of Schlemm's canal of glaucomatous eyes. Invest Ophthalmol Vis Sci 2002; 43:2950-5.

17. Johnson M, Shapiro A, Ethier CR, Kamm RD. Modulation of outflow resistance by the pores of the inner wall endothelium. Invest Ophthalmol Vis Sci 1992; 33:1670-5.

18. Borras T. Gene expression in the trabecular meshwork and the influence of intraocular pressure. Prog Retin Eye Res 2003; 22:435-63.

19. Gonzalez P, Epstein DL, Borras T. Genes upregulated in the human trabecular meshwork in response to elevated intraocular pressure. Invest Ophthalmol Vis Sci 2000; 41:352-61.

20. Vittitow JL, Garg R, Rowlette LL, Epstein DL, O'Brien ET, Borras T. Gene transfer of dominant-negative RhoA increases outflow facility in perfused human anterior segment cultures. Mol Vis 2002; 8:32-44 <http://www.molvis.org/molvis/v8/a5/>.

21. Peterson JA, Tian B, Bershadsky AD, Volberg T, Gangnon RE, Spector I, Geiger B, Kaufman PL. Latrunculin-A increases outflow facility in the monkey. Invest Ophthalmol Vis Sci 1999; 40:931-41.

22. Peterson JA, Tian B, Geiger B, Kaufman PL. Effect of latrunculin-B on outflow facility in monkeys. Exp Eye Res 2000; 70:307-13.

23. Peterson JA, Tian B, McLaren JW, Hubbard WC, Geiger B, Kaufman PL. Latrunculins' effects on intraocular pressure, aqueous humor flow, and corneal endothelium. Invest Ophthalmol Vis Sci 2000; 41:1749-58.

24. Rao PV, Deng P, Sasaki Y, Epstein DL. Regulation of myosin light chain phosphorylation in the trabecular meshwork: role in aqueous humour outflow facility. Exp Eye Res 2005; 80:197-206.

25. Rao PV, Deng PF, Kumar J, Epstein DL. Modulation of aqueous humor outflow facility by the Rho kinase-specific inhibitor Y-27632. Invest Ophthalmol Vis Sci 2001; 42:1029-37. Erratum in: Invest Ophthalmol Vis Sci 2001; 42:1690.

26. Thieme H, Nuskovski M, Nass JU, Pleyer U, Strauss O, Wiederholt M. Mediation of calcium-independent contraction in trabecular meshwork through protein kinase C and rho-A. Invest Ophthalmol Vis Sci 2000; 41:4240-6.

27. Witke W. The role of profilin complexes in cell motility and other cellular processes. Trends Cell Biol 2004; 14:461-9.

28. dos Remedios CG, Chhabra D, Kekic M, Dedova IV, Tsubakihara M, Berry DA, Nosworthy NJ. Actin binding proteins: regulation of cytoskeletal microfilaments. Physiol Rev 2003; 83:433-73.

29. Kaufman PL, Barany EH. Loss of acute pilocarpine effect on outflow facility following surgical disinsertion and retrodisplacement of the ciliary muscle from the scleral spur in the cynomolgus monkey. Invest Ophthalmol 1976; 15:793-807.

30. Barany EH. Simultaneous measurement of changing intraocular pressure and outflow facility in the vervet monkey by constant pressure infusion. Invest Ophthalmol 1964; 31:135-43.

31. Johnson DH, Tschumper RC. Human trabecular meshwork organ culture. A new method. Invest Ophthalmol Vis Sci 1987; 28:945-53.

32. Gual A, Llobet A, Gilabert R, Borras M, Pales J, Bergamini MV, Belmonte C. Effects of time of storage, albumin, and osmolality changes on outflow facility (C) of bovine anterior segment in vitro. Invest Ophthalmol Vis Sci 1997; 38:2165-71.

33. Erickson-Lamy K, Rohen JW, Grant WM. Outflow facility studies in the perfused human ocular anterior segment. Exp Eye Res 1991; 52:723-31.

34. Borras T, Matsumoto Y, Epstein DL, Johnson DH. Gene transfer to the human trabecular meshwork by anterior segment perfusion. Invest Ophthalmol Vis Sci 1998; 39:1503-7. Erratum in: Invest Ophthalmol Vis Sci 1998; 39:1984.

35. Borras T, Rowlette LL, Erzurum SC, Epstein DL. Adenoviral reporter gene transfer to the human trabecular meshwork does not alter aqueous humor outflow. Relevance for potential gene therapy of glaucoma. Gene Ther 1999; 6:515-24.

36. Borras T, Tamm ER, Zigler JS Jr. Ocular adenovirus gene transfer varies in efficiency and inflammatory response. Invest Ophthalmol Vis Sci 1996; 37:1282-93.

37. Ethier CR, Wada S, Chan D, Stamer WD. Experimental and numerical studies of adenovirus delivery to outflow tissues of perfused human anterior segments. Invest Ophthalmol Vis Sci 2004; 45:1863-70.

38. Green M, Loewenstein PM. Autonomous functional domains of chemically synthesized human immunodeficiency virus tat trans-activator protein. Cell 1988; 55:1179-88.

39. Schwarze SR, Ho A, Vocero-Akbani A, Dowdy SF. In vivo protein transduction: delivery of a biologically active protein into the mouse. Science 1999; 285:1569-72.

40. Zhao M, Weissleder R. Intracellular cargo delivery using tat peptide and derivatives. Med Res Rev 2004; 24:1-12.

41. Ho A, Schwarze SR, Mermelstein SJ, Waksman G, Dowdy SF. Synthetic protein transduction domains: enhanced transduction potential in vitro and in vivo. Cancer Res 2001; 61:474-7.

42. Mimuro H, Suzuki T, Suetsugu S, Miki H, Takenawa T, Sasakawa C. Profilin is required for sustaining efficient intra- and intercellular spreading of Shigella flexneri. J Biol Chem 2000; 275:28893-901.

43. Becker-Hapak M, McAllister SS, Dowdy SF. TAT-mediated protein transduction into mammalian cells. Methods 2001; 24:247-56.

44. Stamer WD, Seftor RE, Williams SK, Samaha HA, Snyder RW. Isolation and culture of human trabecular meshwork cells by extracellular matrix digestion. Curr Eye Res 1995; 14:611-7.

45. Ma W, Rogers K, Zbar B, Schmidt L. Effects of different fixatives on beta-galactosidase activity. J Histochem Cytochem 2002; 50:1421-4.

46. Bernatchez PN, Bauer PM, Yu J, Prendergast JS, He P, Sessa WC. Dissecting the molecular control of endothelial NO synthase by caveolin-1 using cell-permeable peptides. Proc Natl Acad Sci U S A 2005; 102:761-6.

47. Joliot A, Prochiantz A. Transduction peptides: from technology to physiology. Nat Cell Biol 2004; 6:189-96.

48. Wadia JS, Stan RV, Dowdy SF. Transducible TAT-HA fusogenic peptide enhances escape of TAT-fusion proteins after lipid raft macropinocytosis. Nat Med 2004; 10:310-5.

49. Grierson I, Day J, Unger WG, Ahmed A. Phagocytosis of latex microspheres by bovine meshwork cells in culture. Graefes Arch Clin Exp Ophthalmol 1986; 224:536-44.

50. Zhou L, Li Y, Yue BY. Alteration of cytoskeletal structure, integrin distribution, and migratory activity by phagocytic challenge in cells from an ocular tissue--the trabecular meshwork. In Vitro Cell Dev Biol Anim 1999; 35:144-9.

51. Buss F, Temm-Grove C, Henning S, Jockusch BM. Distribution of profilin in fibroblasts correlates with the presence of highly dynamic actin filaments. Cell Motil Cytoskeleton 1992; 22:51-61.

52. Hajkova L, Bjorkegren Sjogren C, Korenbaum E, Nordberg P, Karlsson R. Characterization of a mutant profilin with reduced actin-binding capacity: effects in vitro and in vivo. Exp Cell Res 1997; 234:66-77.

53. Cao LG, Babcock GG, Rubenstein PA, Wang YL. Effects of profilin and profilactin on actin structure and function in living cells. J Cell Biol 1992; 117:1023-9.

54. Bubb MR, Yarmola EG, Gibson BG, Southwick FS. Depolymerization of actin filaments by profilin. Effects of profilin on capping protein function. J Biol Chem 2003; 278:24629-35.

55. Hajkova L, Nyman T, Lindberg U, Karlsson R. Effects of cross-linked profilin:beta/gamma-actin on the dynamics of the microfilament system in cultured cells. Exp Cell Res 2000; 256:112-21.

56. Cai S, Liu X, Glasser A, Volberg T, Filla M, Geiger B, Polansky JR, Kaufman PL. Effect of latrunculin-A on morphology and actin-associated adhesions of cultured human trabecular meshwork cells. Mol Vis 2000; 6:132-43 <http://www.molvis.org/molvis/v6/a18/>.

57. Wiederholt M, Bielka S, Schweig F, Lutjen-Drecoll E, Lepple-Wienhues A. Regulation of outflow rate and resistance in the perfused anterior segment of the bovine eye. Exp Eye Res 1995; 61:223-34.

58. Somlyo AP, Somlyo AV. Signal transduction by G-proteins, rho-kinase and protein phosphatase to smooth muscle and non-muscle myosin II. J Physiol 2000; 522:177-85.

59. Tian B, Kaufman PL. Effects of the Rho kinase inhibitor Y-27632 and the phosphatase inhibitor calyculin A on outflow facility in monkeys. Exp Eye Res 2005; 80:215-25.