Gap junction protein connexin43 and tunneling nanotubes in human trabecular meshwork cells

Abstract

Intercellular communication via gap junctions and tunneling nanotubes (TNT) play pivotal roles for cell differentiation and proliferation and homeostasis. Intercellular communication has been found in human trabecular meshwork cells, which are key elements for allowing the direct passing aqueous humor and for maintaining the homeostasis of intraocular pressure. Here we showed the expression of gap junction protein connexin43 in human trabecular meshwork cells, and the presence of TNTs in these cells labeled with enhanced GFP generated using highly efficient lentiviral transduction. More importantly, the TNTs in human trabecular meshwork cells were significantly reduced after shRNA or CRISPR Cas9 mediated knockdown of Cx43 in these cells. The results indicated that gap junction protein connexin43 may play an important role in TNTs formation in human trabecular meshwork cells.

Introduction

Tunneling nanotubes (TNTs) are an especially unique form of intercellular communication, they are F-actin based long distance cellular protrusions that allows direct long-range cell-to-cell transfer of signaling molecules, including DNA, mitochondria, microRNAs, and others [1-3]. Another form of intercellular communication is gap junctions, which are intercellular channels that allow direct cell to cell exchange of ions and small molecules with molecular weight less than 1000 Dalton such as water, amino acids, nucleotides, ATP, calcium, cAMP, cGMP, IP3 [4,5]. To date, there are twenty mouse connexin genes and twenty-one human connexin genes in mouse and human genome, respectively [6]. TNTs and gap junctional intercellular communication (GJIC) play pivotal roles in maintaining physiological function of cells and tissues, which include cell differentiation, cell proliferation, apoptosis, cell development and tissue homeostasis. TNTs and gap junctions are widely distributed in different cells and tissues, and their functional involvement in cancer pathology is also extensively discussed and investigated [7-9]. One of the most ubiquitously expressed gap junction protein is connexin43 (Cx43), also called GJA1 [10]. Importantly, Cx43 has been shown to be localized in mitochondria, which can be transferred via TNTs, therefore raising the possibility that Cx43 and TNTs can be associated [11]. However, what functional role of Cx43 in the regulation of TNTs is not well determined.

Glaucoma is one of the leading causes of eye disease worldwide, and elevation in intraocular pressure is a key risk factor for glaucoma. Trabecular meshwork (TM) is one of the major aqueous humor (AH) drainage structure, it plays a very important role in the regulation of the aqueous flow resistance and intraocular pressure [12,13]. Recently, it has been reported that both the TNTS and gap junction protein Cx43 are presented in human trabecular meshwork cells [14,15]. Cx43 mutations are associated with many human disease conditions including glaucoma [16,17]. However, whether there is any functional interaction between Cx43 and TNTs in trabecular meshwork is totally unknown.

Here we use human TM cells with overexpressing enhanced GFP (eGFP) as a model cell system to define whether Cx43 can play a critical role in the formation of TNTs.

Materials and methods

Monoclonal anti-Cx43 antibody (Cat. No: MABT903) was purchased from EMD Millipore corporation. Cas9 antibody was obtained from Novus Biological Corp.

Primary human TM cells culture is based on the protocol as we describe previously and were performed in accordance with the tenets of the Declaration of Helsinki [13].

Lentiviral eGFP virus generation and transduction was conducted using the protocol as we described previously [18]. Briefly, HEK293 cells were transfected with Lipofectamine 3000 using expression plasmid plenti CMV GFP DEST and packaging vectors including pLP1, pLP2 and pLP/VSVG plasmids. After 48 hours of transfection, the lentivirus was harvested, centrifuged for 5 min at 1000 g centrifugal force and filtered, and then then equally distributed to HTM cells.

Cx43 knockdown using shRNA and CRISPR Cas9 system

Lentiviral Cx43 shRNA transduction was carried out as we previously described [18]. Lentiviral Cx43 shRNAs were acquired from Santa Cruz Biotechnology (Dallas, TX, USA), the shRNA sequences used for targeting human Cx43 were consisted of four different sequences, including: A: 5’-GATCCCTGCGAACCTACATCATCATTCAAGAGATGATGATGTAGGTTCGCAGTTTTT-3’; B: 5’-GATCCGAACCTACATCATCAGTATTTCAAGAGAATACTGATGATGTAGGTTCTTTTT-3’; C: 5’-GATCCGTTGGGATGTCACTTAACATTCAAGAGATGTTAAGTGACATCCCAACTTTTT-3’; D: 5’-GATCCCCTACTTAATACACAGTAATTCAAGAGATTACTGTGTATTAAGTAGGTTTTT-3’. Scrambled lentiviral shRNA was used as control.

The sgRNA that targets human Cx43 in a CRISPR V2 lentiviral vector with sequence 5’-TGTGTTCTATGTGATGCGAA-3’ was generated from Genscript Biotechology (Piscataway, NJ). The HEK293 cells were cultured in Dulbecco’s Modified Eagle’s medium (DMEM) supplemented with low glucose, 10% fetal bovine serum with 1% antibiotics in a tissue culture incubator at 37°C with 5% CO₂. After reaching about 70% confluence, cells were transfected using Lipofectamine 3000 combined with lentiviral packaging vectors (pLP1, pLP2 and pLP/VSVG plasmids), a CRISPR Cas9 sgRNA or a plenti CMV GFP DEST plasmid. After 48 hours of transfection, the cell culture medium was harvested and filtered using a 0.45 um filter, and the fresh lentivirus for Cx43 CRISPR Cas9 sgRNA or Cx43 shRNA lentivirus were then equally distributed to HTM cells. After 3 days of viral transduction, cells were subjected to GFP monitoring, western blotting or immunofluorescence labeling.

Immunofluorescence (IF) labeling of Cx43. IF labeling of Cx43 was performed utilizing the protocol as we previously employed [19-21].

HTM cells grown on coverslips were fixed with 2% cold paraformaldehyde (PA) for 10 min, and were washed with PBS. For immunofluorescence labelling, cells were incubated in 50 mM Tris-HCl, pH 7.4, containing 1.5% sodium chloride (TBS) and 0.3% Triton X-100 (TBSTr) and 5% NGS overnight at 4°C with monoclonal anti-Cx43 primary antibody (0.5 ug/ml). Coverslips were then washed four times in TBSTr and incubated for 1 h at room temperature simultaneously with Alexa Fluor 594-conjugated goat anti-mouse IgG diluted at 1:1000 (Molecular Probes, Eugene, Oregon). The coverslips were then examined under an Olympus confocal microscope.

Western blotting

The western blotting procedure was described previously [22-24]. Briefly, HTM cells were rinsed in cold PBS buffer (50 mM sodium phosphate buffer, pH 7.4, 0.9% saline) and were lysed in an IP buffer as we previously used [25,26]. Homogenates were centrifuged at 20,000×g for 15 min at 4°C, and the protein concentration of the supernatants were determined with Bradford reagent (Bio-Rad Laboratories, Hercules, CA). Proteins were boiled for 3 min and then were separated using SDS-PAGE (10 µg of protein per lane) in a 12.5% gels and followed by transblotting to polyvinylidene difluoride (PVDF) membranes (Bio-Rad Laboratories) in standard Tris-glycine transfer buffer, pH 8.3, containing 0.5% SDS. Membranes were blocked for 1 h at room temperature in an Odyssey blocking buffer (Li-Cor Biosciences), rinsed 30 min in TBSTw (TBS + 0.05% tween 20), and incubated overnight at 4°C with monoclonal anti-Cx43 antibody diluted at 1:1000 in PBS buffer containing 0.05% Tween-20. Membranes were washed four times in TBSTw, incubated with specific secondary antibodies (Li-Cor Bio-sciences, Lincoln, NE). Membranes were then scanned on an imaging system (Odyssey Infrared; Li-Cor) with companion software (Odyssey 2.0; Li-Cor).

TNTs counts

The counting of TNTs in living TM cells grown in 4 well glass chamber was conducted using an Olympus microscope, the number of TNTs was counted from 100 eGFP labeled cells in 10 randomly chosen fields, and the counting was repeated in at least three independent experiments.

Statistics analyses

Graphpad Prism software and ANOVA was performed to determine the differences in protein expression level and TNT numbers in different groups. Data was showed as the average ± standard error of the mean, and P value < 0.05 was considered statistically significant.

Results

Formation of TNTS in human trabecular meshwork cells

We first investigated the formation of TNTs in human trabecular meshwork cells using highly efficient lentiviral transduction of eGFP in HTM cells. As shown in Figure 1, it appears that almost all the HTM cells express eGFP, indicating the method we used can achieve high level of eGFP expression. Importantly, the presence of TNTs can be easily seen (with arrow) between HTM cells, with different size, different length, indicating the diversity distribution of TNTs in HTM cells and the feasibility of detecting TNTs in eGFP labeled HTM cells.

Figure 1.

Detection of TNTs in eGFP labeled HTM cells. Representative image showing the presence of TNTs (arrows) between HTM cells, with different size, different length, indicating the diversity distribution of TNTs in HTM cells. Scale bar: 10 µm.

Cx43 in human TM cells and TNTs

We further investigated gap junction protein Cx43 immunofluorescence labeling in primary cultured HTM cells and its morphological localization in TNTs. As shown in Figure 2, IF labeling of punctate Cx43 showed the presence of strong gap junction Cx43 in HTM cells, further, it appears that some of Cx43 is localized at TNTS (arrows), indicating some Cx43 is localized at TNTS.

Figure 2.

Immunofluorescence labeling of Cx43 (red) in HTM cells. Representative image showing the detection of punctate labeling of Cx43 with different size in HTM cells, and it appears that some Cx43 labeling is localized at TNTS (arrows). Scale bar: 10 µm.

Knockdown of Cx43 using lentiviral Cx43 shRNA and CRISP-Cas9 system

We further investigated whether there is some functional relationship between Cx43 and TNTs formed in HTM cells, so we first performed lentiviral Cx43 shRNA mediated Cx43 knockdown in HTM cells using the same Cx43 shRNA as we previously reported [18]. Then we also used highly efficient CRISPR Cas9 system to knockdown Cx43 in HTM cells. As shown in Figure 3, WB results showed that Cx43 expression level is significantly reduced in shRNA Cx43 as well as in CRISPR Cas9 sgRNA treated HTM cell lysates, indicating the high efficiency of Cx43 knockdown using lentiviral Cx43 shRNA and CRISPR-Cas9 sgRNA techniques (Figure 3A). Three separate experiments showed the same results and the statistics analysis was summarized in Figure 3B. In addition, WB also showed the presence of 170 kDa Cas9 protein in CRISPR Cas9 treated HTM cell lysates, but no detection in control cell lysates (Figure 3C).

Figure 3.

WB showing significantly knockdown of Cx43 using lentiviral Cx43 shRNA or CRISPR-Cas9 sgRNA, and detection of Cas9 protein in Cas9 sgRNA lentiviral transduced HTM cell lysates. (A) Representative WB images showed reduced Cx43 expression in lysates of lentiviral Cx43 shRNA transduced HTM cells (lane 1), as well as in lysates of CRISPR Cas9 sgRNA transduced cells (lane 2), while scrambled shRNA (lane 3) and sgRNA control (lane 4) were used as controls. Three separate experiments showed the same results and the statistics analysis was summarized in (B), (C) Representative image showed the detection of Cas9 protein in lysates of CRISPR Cas9 sgRNA transduced cells.

Reduction of TNTs in HTM cells after knockdown of Cx43 by lentiviral Cx43 shRNA or CRISPR Cas9 sgRNA

Further, we determined whether there is any changes in TNTs after knockdown of Cx43 using lentiviral Cx43 shRNA or CRISPR Cas9 system. As shown in Figure 4, the TNTs in HTM cells were significantly reduced after lentiviral Cx43 shRNA or CRISPR Cas9 sgRNA mediated Cx43 knockdown in HTM cells, indicating that Cx43 plays an essential role in the formation of TNTs in HTM cells.

Figure 4.

Reduced TNTs in HTM cells transduced with lentiviral Cx43 shRNA (A) or CRISPR-Cas9 sgRNA (B), three separate experiments showed the same results and the statistics analysis was summarized in (C).

Discussion

Our results showed the presence of TNTs in human trabecular meshwork cells, and the TNTs in human trabecular meshwork cells were significantly reduced after lentiviral Cx43 shRNA and CRISPR Cas9 mediated knockdown of CX43. These results demonstrated that gap junction protein connexin43 plays an important role in TNTs formation in human trabecular meshwork cells.

In order to identify TNTs in human TM cells, we used highly efficient lentiviral transduction of enhanced GFP into human TM cells, and the long-range direct Cell to Cell connected TNTs are easily found in these cells. From our knowledge, this is the first evidence to show the presence of TNTs in human TM cells utilizing only GFP labeled TM cells without tagging with other fusion proteins. It raises the possibility that our method employed in this report may be considered as a useful tool for the investigating TNTs in other type of cells.

The presence of Cx43 in TNTs of human TMs is line with other groups finding in other cells [27-29]. Our data clearly showed that Cx43 plays an important role in the formation of TNTs in human TM cells, which is demonstrated by the knockdown of Cx43 and accompany with reduced TNTs. The mechanisms involving Cx43 in the regulation of the formation of TNTs is not clear, but may be related to the following possibilities. 1. Cx43 is associated with cell cytoskeleton network including actin and microtubules [30,31], while F-actin and microtubules mediated protrusion is related to TNTs [2]; 2. Cx43 can enhance the formation of filopodia and cell migration, which are related to actin network [32]; 3. Cx43 has been shown to be expressed in mitochondria, and the transfer of mitochondria via TNTs is well established in a variety of cells or tissues [2,33]. Recently, it also has been reported that the transfer of mitochondria from TNTs formed by HMSCs to MScs is mediated by Cx43 [11]. Another report also showed that the TNTS in HIV infected macrophages mediate long-range gap junctional communication [27]. In addition, it appears that communication of Ca(2+) signals via TNTS tunneling membrane nanotubes is mediated by transmission of inositol trisphosphate through gap junctions [34]. More importantly, it has been reported the presents of TNTs in neuronal populations, which are absence of Cx43 but expressing Cx36 and Cx45, it raises the possibility whether neuronal gap junctional proteins including Cx36 and Cx45 may also play a functional role in the formation of TNTs in neurons [35-37].

In addition to knockdown Cx43 with Cx43 shRNA in this report, as we previously reported in human umbilical vein endothelial cells, we also employed the highly efficient CRISPR-Cas9 sgRNA mediated knockout of Cx43 in human TM cells, this is achieved with its non-homologus end joining (NHEJ) effect, which creates random insertions or deletions during the process of double stranded break repair, which is initiated by Cas9 [38]. Our western blotting results also showed detection of Cas9 protein in lysates of CRISPR-Cas9 transduced cells, indicating the Cas9 is expressed in these cells. More importantly, WB results showed significant loss of Cx43 in these cells, indicating the efficiency of knockdown Cx43 utilizing CRISPR-Cas9 system is high, which has been verified in other cells, thereby providing a new way to manipulate gene expression in human TM cells. Considering the fact that whether there is off-targets effect of CRISPR-Cas9 sgRNA, further studies, such as employing deep sequencing combined with off-targets prediction methods are needed to validate the results. Further, by administrating mutant Cas9 system or using the ribonucleoprotein (RNP) will dramatically reduce the off-target effect [38].

From our experimental data with eliminating Cx43 and subsequently displaying reduced TNTs in human TM cells, we conclude that Cx43 may play a key role in the formation of TNTs in human TM cells.

Disclosure of conflict of interest

None.