Upregulation of multiple signaling pathways by Dock5 deletion in epithelial cells

Mice

The RLC mutation was first found in a male mouse from the CXSN/A strain [1], a recombinant inbred strain between BALB/cHeA and STS/A mice. The mutant was crossed once with a BALB/cHeA female and maintained by sister-brother mating over 20 generations to establish the mutant line RLC/Stm, which is homozygous for the mutant gene rlc. BALB/c mice were used as the wild-type (WT) control unless otherwise noted. The Dock5-knockout (KO) mouse, which is generated by disrupting the Dock5 locus by the retroviral insertion of a β-geo cassette in the intron flanked by exons 29 and 30, was a gift from Dr. Jean-Francoise Cote [10]. Genomic DNA from tail or finger biopsies was extracted using Genomic DNA extraction kits (Lyppo; Gene Modification R&D, Osaka, Japan), and the mouse genotypes were determined by PCR analyses, as described previously [3,10]. The animal protocols were reviewed and approved by the Animal Care and Use Committee of Kanazawa Medical University.

Total RNA preparation, reverse transcription, and quantitative PCR

After CO₂ anesthesia and dislocation of the neck, the mouse eyes were extracted. For LEC isolation, the eyes were soaked for several minutes in PBS (Nissui, Tokyo, Japan), and the layers of the basement membrane and LECs were stripped from the lens nucleus cores under a stereomicroscope. The LECs were prepared from 21-day-old RLC and WT mice. The total RNA was prepared from the pooled LECs from 5 to 8 mice for each experiment using an RNA purification kit (RNeasy Mini Kit; Qiagen, Valencia, CA), as described previously [14,15]. First, 500 ng of total RNA was reverse-transcribed with reverse transcriptase (ReverTraAce; Toyobo, Osaka, Japan) in 20 μl of reaction mixture. Real-time PCR was performed with a SyBr Green PCR system (SYBR® Green Realtime PCR Master Mix; Toyobo). Real-time PCR primers were designed using Primer3Plus as follows: C-type lectin domain containing 11A (CLEC11A; F: 5′-ACT GGG GTG GGA AAT GAG G-3′; R: 5′-AGG GTT GGG ACT AGA AGA AGG TG-3′), ubiquitin specific peptidase 18 (USP18; F: 5′-GAG CAG CAG GAG GAG CAA AC-3′; R: 5′-GGA CGA AAC ATC TCA AGG CAT C-3′), potassium voltage-gated channel subfamily E regulatory subunit 1 (KCNE1; F: 5′-GCT TCT TCG GCT TCT TCA CC-3′; R: 5′-AGA CGG CCT TGC CTT TCT C-3′), calpain 6 (CAPN6; F: 5′-CAA GGT GGA GAT GAA CCG AAG-3′; R: 5′-ACG GGT GTC GAT ATA AGT GGA AG-3′), fibroblast growth factor 1 (FGF1; F: 5′-GGA AAG GCT GGA AGA AAA CCA-3′; R: 5′-GGGAGGGGCAGAAACAAGA-3′), TMPRSS11D (forward: 5′-CGG CTT TGG TTT GTT GTG G-3′; R: 5′- ATC CAG TTG CGG TAG GCT GT-3′), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; F: 5′-AAT GAA GGG GTC GTT GAT GG-3′; R: 5′-AAA ATG GTG AAG GTC GGT GTG-3′).

The mRNA levels were normalized by the GAPDH mRNA level. Real-time PCR data were quantified using an Excel Macro designed for Gene Expression Analysis with an iCycler iQ real-time PCR detection system, which was provided by Bio-Rad (Hercules, CA).

Microarray analysis

For the Affymetrix microarray, labeled cRNAs were synthesized from sample RNA using a MessageAmp II-Biotin Enhanced Kit (Ambion, Carlsbad, CA). Target hybridizations were performed on a CanGene-1_0-st GeneChip (Affymetrix, High Wycombe, UK) according to the manufacturer’s protocol. The hybridized cRNAs were washed and stained in a GeneChip Fluidics Station 450, and signals were detected using a GeneChip Scanner 3000 (Affymetrix). Digitalized image data were processed using the GeneChip Operating Software (GCOS). Fold increases in the signal intensities of the selected genes were calculated using the GeneSpring GX software (Agilent Technologies, Santa Clara, CA). The RNA extraction and microarray analysis were independently performed twice.

Ingenuity Pathway Analysis

The dataset contained gene identifiers and corresponding expression values. It was uploaded into the Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems, in the public domain), as described previously [16].

Immunohistochemistry

After CO₂ anesthesia and dislocation of the neck, the mouse eyes were extracted. For the immunostaining of activated extracellular signal-regulated kinase (Erk), paraffin-embedded eye sections were prepared and incubated with the antibody against phospho-142 Erk1/2 (pErk) Thr202/Tyr204 (Cell Signaling Technology, Beverly, MA). The bound antibody was visualized by SimpleStain Mouse MAX PO (Nichirei Biosciences, Tokyo, Japan) according to the manufacturer’s protocol.

Quantification of pErk in the lens

The LEC in the monolayer at the equator was designated as cell #1, and the next cell toward the anterior pole was designated as cell #2. The number of cells with or without positive staining in the nucleus, cytoplasm, or both was counted until 10 negative cells appeared. The positivity of the cells was expressed as “positive cells/positive + negative cells” from six equators (i.e., three eyes from three mice). Statistical analysis was performed using GraphPad Prism 5 software (La Jolla, CA).

Cell culture and N-(3,5-dichlorophenyl) benzenesulfonamide treatment

Madin–Darby canine kidney (MDCK) cells were purchased from RIKEN BioResource Center (Ibaraki, Japan, No RCB0995) and validated by the Cell Line Authentication Service (IDEXX BioResearch, Columbia, MO) based on the genetic profile using 14 species-specific single tandem repeat (STR) markers. MDCK cells were maintained in Dulbecco’s Modified Eagle Medium (D-MEM, WK0442976, Wako Pure Chemical Industries, Tokyo, Japan) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO), 100 units/ml of penicillin, and 100 μg/ml of streptomycin in a 5% CO₂ humidified incubator at 37 °C. Twenty-four hours after seeding to the plastic dish, MDCK cells were treated with or without 100 μM N-(3,5-dichlorophenyl) benzenesulfonamide (C21; Alfa Aesar, Ward Hill, MA) or dimethyl sulfoxide (DMSO) and further cultured for 48 h.

SDS–PAGE and western blotting

Cells were lysed in 1× SDS sample buffer [62.5 mM Tris-HCl (pH 6.8), 12% glycerol, 2% SDS, 0.004% bromophenol blue (BPB), and 10% 2-mercaptoethanol]. After sonication and incubation at 95 °C, the proteins were separated by SDS–PAGE on 10% Supersep precast gels (Wako Pure Chemical Industries), and transferred to polyvinylidene difluoride (PVDF) membranes (Immobilon; Millipore, Bedford, MA). After blocking with Odyssey blocking buffer (LI-COR, Lincoln, NE) for 30 min, the membranes were blotted with various antibodies diluted to 1:1,000 [phospho-Erk1/2, Erk1/2 (Cell Signaling Technology), phospho-nuclear factor-kappa B (NFκB) p65 (Ser536; Cell Signaling Technology), pan Akt (Cell Signaling Technology), and phospho-Akt (Ser473; Cell Signaling Technology)], 1:100 (NFκB p65; Santa Cruz Biotechnology, Santa Cruz, CA), or 1:5,000 (β-tubulin; Sigma-Aldrich) in a solution of Odyssey blocking buffer and Tris-buffered saline, followed by incubation with secondary antibodies. The membranes were then scanned with an Odyssey infrared (IR) scanner. Quantification was performed using Image Studio Lite Software (LI-COR). Statistical analysis was performed using GraphPad Prism 5 software.

Microarray and IPA analysis

To understand the molecular mechanisms induced by Dock5 depletion in the LECs, we performed a microarray analysis for mRNA expression. LECs were isolated from 21-day-old WT mice (BALB/c) and 21-day-old RLC mice, and their RNAs were extracted. The genes that showed at least twofold and 1.5-fold changes in their expression in the RLC compared with the WT mice are listed in Table 1 and Appendix 1 (as an average from two independent experiments). The results from three analyses, that is, GeneSpring GX software, Affymetrix TAC software with analysis of variance (ANOVA), and the mean with standard deviation, are shown (Appendix 1).

To certify the accuracy of the microarray data, we performed a quantitative real time (RT)-PCR analysis of six genes: CLEC11A, USP18, KCNE1, CAPN6, FGF1, and TMPRSS11D. As shown in Appendix 2, the amounts of mRNA detected by RT–PCR showed patterns of increase/decrease that were comparable to those detected by the microarray. The relative expression of these six genes measured by RT–PCR and microarray showed parallel correlations (Appendix 2). Unfortunately, commercially available antibodies against CLEC11A, COL4A6, and CCBE1, which were upregulated in the RLC mice, did not detect the endogenous proteins either in paraffin-embedded eye sections or by western blotting (data not shown).

Using the data of the genes that showed greater up- or downregulation in the RLC mice, we performed an IPA to understand the biological relevance and function of the significantly altered genes (Table 2). IPA is a statistical gene-set-enrichment method in which differentially expressed genes intersect with sets of genes associated with a specific biological function or pathway [17]. The IPA database was manually constructed by trained PhD-level curators based on a review of the literature [17].

As a first step in the IPA, we extracted highly influenced Ingenuity canonical pathways from the genes that showed a >1.5-fold difference between the RLC and WT mice (Appendix 1, “FC1.5_GeneSpring” sheet). Overall, compared to the WT, 149 genes were differentially expressed in the RLC mice, and among them, 93 and 56 genes were upregulated and downregulated, respectively. Table 2 shows the pathways that were significantly (p<0.05) affected in the RLC mice. In the table, the term “Ratio” denotes the genes changed in the RLC and WT LECs over the predicted genes involved in the pathway listed on the left (as “Ingenuity Canonical Pathways”), and the genes involved in the individual pathway are listed on the right (as “Molecules”). We evaluated the significance of the association between the altered genes and canonical pathway using a right-tailed Fisher’s exact test to calculate a p value determining the probability that the association was explained by chance alone. In this analysis, three of the same genes, that is, OAS1, HLA-B, and IRF7, appeared in different pathways.

Next, an “Ontology analysis” was performed based on the genes’ categorization by their functions and related diseases. Figure 1 shows the pathways that were significantly (p<0.01) changed in the RLC mice. The graph expresses the -log (p value) for each “category of function,” revealing that the genes that were altered in the RLC mice showed the greatest enrichment for “Antimicrobial response,” “Inflammatory response,” and “Dermatological diseases and conditions.”

Prediction and confirmation of phosphorylated Erk in Dock5-depleted LECs

To further understand the biological relevance of the differentially expressed genes in the lenses of RLC mice, we used an IPA to explore their relationships and known annotations. The networks that were more affected in the LECs of RLC mice than those of WT mice are listed, and the top 10 biofunction networks affected are listed in Table 3. Notably, gene expression was the most affected network, with a high score of 42, which was for “Antimicrobial Response, Inflammatory Response, Dermatological Diseases and Conditions.”

Based on the collected information and data from the literature, the IPA draws a map by connecting the genes changed in RLC mice (Figure 2A). In the present case, this map included nine predicted genes (open objects). Among them, we focused on Erk, which was connected by 10 lines with positive regulation. Erk plays a central role in cell growth [18]. It is a serine/threonine kinase, the activity of which is regulated by phosphorylation [19]. We first examined Erk phosphorylation (pErk) in isolated mouse LECs (mLECs) by western blotting, which revealed that the Erk phosphorylation was below the detectable level (data not shown). Next, we performed immunohistochemistry and observed Erk phosphorylation in the LECs of the WT mice at the equator, but not in the anterior pole (Figure 2B). The higher magnification of the LECs at the equator indicated that the RLC mice showed stronger signals compared with the WT ones (Figure 2C). Moreover, the numbers of positive cells with phosphorylated Erk were higher in the RLC mice compared with the WT mice. During the experiments, we obtained Dock5-KO mice [10] and found that they had a similar eye phenotype to that of the RLC mice at the age of 40–62 weeks (Appendix 3). Similar Erk phosphorylation results were obtained from the Dock5-KO mice (Figure 2D). We quantified the levels of phosphorylated Erk in BALB/c and RLC and C57BL/6 and Dock5-KO mice. As shown in Figure 2E, the number of pErk-positive cells in the RLC lens was significantly higher than that in the BALB/c lens. The level of phosphorylated Erk in the C57BL/6 lens was higher than that in the BALB/c lens. In the Dock5-KO lens, pErk was slightly increased, but not significantly (p = 0.1459). Since we sacrificed the paired mice of BALB/c and RLC, or those of C57BL/6 and Dock5-KO, in one experiment and independently repeated it three times, the average of pErk positivity from each pair of mice was compared. As shown in Figure 2F, in the three BALB/c and RLC pairs, pErk was upregulated in RLC, while in the C57BL/6 and Dock5-KO pairs, pErk upregulation was observed in the two Dock5-KO mice. From these results, we concluded that although phosphorylation of Erk may differ among mouse strains, depletion of Dock5 in LECs induced Erk activation. These results also suggested that, upon Dock5 depletion, the predicted pathway “Antimicrobial Response, Inflammatory Response, Dermatological Diseases and Conditions” is changed to induce Erk activation.

It was previously reported that C21 efficiently inhibits the Rac1 activation induced by Dock5 DHR-2 [11]. To confirm the contribution of Dock5 to Erk activation, we first treated the cultured cell line derived from mLECs with C21 and found that Erk was not activated upon Dock5 inhibition (Appendix 4). The cells exhibited a spindle shape with reduced expression of E-cadherin (Appendix 5), the expression of which is necessary for cell–cell adhesion and the apico-basal polarity of the epithelial cells. Since Rac1 is important for epithelial cell polarity [20], we suspected that altered cell properties, including cell adhesion, polarity, and morphology, may influence gene expression in the mLEC line. We therefore examined the effect of Dock5 inhibition in the MDCK cell line, which has been widely used for epithelial cell research. As shown in Figure 2G, Erk1/2 was phosphorylated in the MDCK cells treated with C21 for 48 h. Quantification confirmed that Dock5 inhibition significantly upregulated Erk phosphorylation (Figure 2H).

To confirm the IPA prediction, we examined two other highly scored networks. In the network with the second highest score, “Lipid Metabolism, Molecular Transport, Small Molecule Biochemistry” (Figure 3A), we found that Akt was predicted to be activated in RLC with eight arrows. Akt (also called protein kinase B), a Pleckstrin homology (PH) domain-containing serine/threonine kinase, has been implicated in the control of diverse cellular functions, including glucose metabolism, gene transcription, cell proliferation, and apoptosis [21]. Upon growth factor stimulation, Akt is phosphorylated at Thr308, then at Ser473, which can be catalyzed by various kinases and can transmit signals to the downstream factors. Therefore, we examined the Ser473 phosphorylation of Akt in MDCK cells treated with C21 for 48 h. As shown in Figure 3B, Akt was phosphorylated in the MDCK by Dock5 inhibition.

Finally, from the map of the third highest network, “Endocrine System Disorders, Gastrointestinal Disease, Metabolic Disease” (Figure 4A), we examined the NFκB activation. NFκB is activated by a wide variety of stimuli. Previously, it was reported that phosphorylation of NFκB p65 (RelA) Ser536 was physiologically induced in response to a variety of proinflammatory stimuli [22]. As shown in Figure 4B, NFκB was significantly phosphorylated in the MDCK by Dock5 inhibition.