Dual suppression of hemangiogenesis and lymphangiogenesis by splice-shifting morpholinos targeting vascular endothelial growth factor receptor 2 (KDR)

Abstract

The KDR gene, which participates in angiogenesis and lymphangiogenesis, produces two functionally distinct protein products, membrane-bound KDR (mbKDR) and its isoform, soluble KDR (sKDR). Since sKDR does not have a tyrosine kinase domain and does not dimerize, it is principally an antagonist of lymphangiogenesis by sequestering VEGF-C. Alternative polyadenylation of exon 30 or intron 13 leads to the production of mbKDR or sKDR, respectively, yet the regulatory mechanisms are unknown. Here we show that an antisense morpholino oligomer directed against the exon 13-intron 13 junction increases sKDR (suppressing lymphangiogenesis) and decreases mbKDR (inhibiting hemangiogenesis). The latent polyadenylation site in intron 13 of KDR is activated by blocking the upstream 5′ splicing site with an antisense morpholino oligomer. Intravitreal morpholino injection suppressed laser choroidal neovascularization while increasing sKDR. In the mouse cornea, subconjunctival injection of the morpholino-inhibited corneal angiogenesis and lymphangiogenesis, and suppressed graft rejection after transplantation. Thus, this morpholino can be used for concurrent suppression of hemangiogenesis and lymphangiogenesis. This study offers new insight into the mechanisms and potential therapeutic modulation of alternative polyadenylation.—Uehara, H., Cho, YK., Simonis, J., Cahoon, J., Archer, B., Luo, L., Das, S. K., Singh, N., Ambati, J., Ambati, B. K. Dual suppression of hemangiogenesis and lymphangiogenesis by splice-shifting morpholinos targeting vascular endothelial growth factor receptor 2 (KDR).

Blood vessel network formation (vasculogenesis and angiogenesis) are necessary for maintenance of the body in vertebrates (1). Many diseases (e.g., cancer, rheumatoid arthritis, macular degeneration, diabetic retinopathy) are due to uncontrolled neovascularization (2–6). Vascular endothelial cell growth factor A (VEGF-A) and KDR [also referred to as vascular endothelial growth factor receptor 2 (VEGFR2)] play central roles in physiological and pathological angiogenesis (7, 8). The KDR gene produces 2 functionally distinct protein products, membrane-bound KDR (mbKDR) and its isoform soluble KDR (sKDR) by alternative polyadenylation (9, 10). The mbKDR has an extracellular domain consisting of 7 immunoglobulin domains, a transmembrane domain, and tyrosine kinase domains (7, 8) and is the primary angiogenic receptor for VEGF-A. While mbKDR is composed of 30 exons in humans and mice, sKDR is produced by utilization of polyadenylation signals within intron 13 in mice. Since sKDR does not have tyrosine kinase domains and has much more affinity for VEGF-C than VEGF-A, it is an antagonist of VEGF-C, the key driver of lymphangiogenesis (9, 10). Thus, the membrane-bound isoform of KDR is prohemangiogenic, while the soluble isoform of KDR is antilymphangiogenic.

Here we report that a morpholino antisense oligomer can shift splicing of KDR pre-mRNA from the membrane to the soluble isoform in human umbilical vein endothelial cells (HUVECs). The induced sKDR requires utilization of a polyadenylation signal in intron 13, which is usually not activated in HUVECs. In addition, morpholino intravitreal injection suppressed laser choroidal neovascularization while increasing vitreous sKDR. Furthermore, in a mouse corneal suturing model, injection of the morpholino into the subconjunctival space suppressed corneal angiogenesis and lymphangiogenesis, and suppressed graft rejection in mouse corneal transplantation. Our results indicate that exon recognition by splicing factors affects subsequent polyadenylation signal activation and that by modifying it, latent polyadenylation signals can be activated, inducing alternative isoforms of proteins. We believe that this study elucidates alternative polyadenylation and that modification of this mechanism could potentially be a new drug target.

MATERIALS AND METHODS

Morpholino oligomer and primer sequences

Morpholino oligomers were purchased from Gene Tools (Philomath, OR, USA). Sequences of the morpholino oligomers and primers are listed in Table 1.

Table 1.

Morpholino oligomer and primer sequences

Oligomer or primer	Sequence
Morpholino oligomer
KDR_MOe13 (human)	5′-gatccagaattgtctccctacCTAG-3′
moKDR_MOe13 (mouse)	5′-cacccagggatgcctccatacCTAG-3′
PCR primer for human
sKDR_F (exon13)	5′-TTCTTGGCTGTGCAAAAGTG-3′
sKDR_R (intron13)	5′-TCTTCAGTTCCCCTCCATTG-3′
mbKDR_F (exon15)	5′-GAGAGTTGCCCACACCTGTT-3′
mbKDR_R (exon17)	5′-CAACTGCCTCTGCACAATGA-3′
KDRexon10_F	5′-CCTACCAGTACGGCACCACT-3′
GAPDH_F	5′-CAGCCTCAAGATCATCAGCA-3′
GAPDH_R	5′-TGTGGTCATGAGTCCTTCCA-3′
PCR primer for mouse
Mouse sKDR_F	5′-ACCAAGGCGACTATGTTTGC-3′
Mouse sKDR_R	5′-CAATTCTGTCACCCAGGGAT-3′
Mouse mbKDR_F	5′-ACCATTGAAGTGACTTGCCC-3′
Mouse mbKDR_R	5′-CCGGTTCCCATCTCTCAGTA-3′
Mouse GAPDH_F	5′-AACTTTGGCATTGTGGAAGGGCTC-3′
Mouse GAPDH_R	5′-ACCAGTGGATGCAGGGATGATGTT-3′
3′RACE primer
Cloning_R1	5′-GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTTV-3′
Cloning_R2	5′-GGCCACGCGTCGACTAGTAC-3′
Cloning_F(1042-1061)	5′-CCAGCATCCTTCAAGTCACA-3′

Upper and lower case in morpholino oligomer sequences correspond to exon and intron, respectively. Morpholino oligomer sequences were synthesized.

Cell cultures, morpholino delivery, and total RNA extraction

HUVECs (Lonza, Walkersville, MD, USA) were cultured in endothelial basal medium (EBM) with endothelial growth medium SingleQuot Kit supplements and growth factors (Lonza) according to the manufacturer's instructions. To prevent loss of endothelial cell properties, cell cultures were limited to passages 4 to 7. As a mouse endothelial cell, MS-1 (American Type Culture Collection, Manassas, VA, USA) were cultured in 5% FBS/DMEM. Morpholinos were delivered by nucleofection (Amaxa, Gaithersburg, MD, USA) using Basic Nucleofector Kit for Primary Mammalian Endothelial Cells (Amaxa) program A-034 for HUVECs and MS-1 cells. For one nucleofection, 1 × 10⁶ cells were used and plated on a 6-well plate. Total RNA was extracted using the RNeasy mini kit (Qiagen, Valencia, CA, USA) with DNaseI treatment.

Complementary DNA (cDNA) synthesis and real-time PCR

cDNAs were synthesized from 400 ng total RNA using the Omniscript RT kit (Qiagen) and Oligo-dT primer (dT20) according to the manufacturer's instructions. Real-time PCR was performed using the QuantiTect SYBR Green PCR Kit (Qiagen) and 1 μl of cDNA. Real-time PCR conditions: 95°C for 10 min, followed by 40 cycles of 94°C for 15 s, 55°C for 30 s, and 72°C for 30 s.

3′ rapid amplification of cDNA ends (RACE)

cDNA was synthesized using cloning_R1. PCR was performed using the LongRange PCR Kit (Qiagen). PCR conditions were 93°C for 3 min, 35 cycles of 93°C for 15 s, 55°C for 30 s, and 68°C for 6 min using KDR_F and cloning_R2. Specific bands were subjected to DNA sequencing. To determine endogenous 3′ untranslated region (UTR) of sKDR mRNA, total RNA extracted from one human cornea was obtained from the Utah Lions Eye Bank (Salt Lake City, UT, USA). cDNA was synthesized with the same above method, and PCR was performed using cloning_F(1042-1061), which is designed in human KDR intron 13, and cloning_R2.

Flow cytometry

At 3 d after nucleofection, cells were treated with trypsin-EDTA and incubated in mouse anti-KDR antibody (ab9530, 1:1000; Abcam, Cambridge, MA, USA) with 10% FBS and 1% sodium azide/PBS for 60 min. After 3 washes, the cells were incubated in Alexa Fluor 647 conjugated anti-mouse IgG (Invitrogen, Carlsbad, CA, USA) for 30 min. The cells were washed 3 times, and fluorescence was detected by a FACScan Analyzer (BD Biosciences, San Jose, CA).

Western blot for sKDR and mbKDR from HUVECs

After nucleofection, cells were cultured in a 75-cm² flask for 3 d without changing the medium. The medium was collected, and cell debris was removed by centrifugation. Trichloroacetic acid (Fisher Scientific, Pittsburgh, PA, USA) was added to the supernatant; final concentration of trichloroacetic acid was 10%. After incubation for 30 min on ice, they were centrifuged at 12 000 g 4°C for 5 min. Supernatants were discarded, and cold acetone was added to the pellet. Centrifugation was repeated, the acetone was discarded, and 800 μl of RIPA buffer was added. Samples were sonicated and subjected to SDS-PAGE under reducing conditions. The same primary antibody in flow cytometry was used for immunobloting.

Deglycosylation of sKDR

Culture medium (2 ml) from HUVECs treated with morpholino oligomer targeting exon 13-intron 13 junction in human KDR (KDR_MOe13; 2 d culture) was freeze dried. Each sample was treated by 200 μl cold Dulbecco's phosphate-buffered saline (DPBS) or trifluoromethanesulfonic acid (TFMS) for 10 and 20 min. After adding 1 ml of cold 1 M Tris-Cl buffer (pH 8.8), the proteins were condensed with trichloroacetic acid precipitation. The pellet was dissolved with 100 μl of RIPA buffer. The same antibody for sKDR detection in Western blot was used.

Intravitreous injection of morpholino and Western blot for sKDR and mbKDR from mouse eye

On d 0 and 3, 2 μl of 100 ng/μl morpholino oligomer targeting exon 13-intron 13 junction in mouse KDR (moKDR_MOe13), standard morpholino (STD_MO), or DPBS was injected intravitreously. On d 4, retinal total RNA was extracted with the RNeasy mini kit with DNaseI treatment for real-time RT-PCR. For Western blot of mbKDR, on d 4, retina was dissolved in RIPA buffer. For Western blot of sKDR, on d 4, intraocular solution was obtained from 6 eyes by pipette. After centrifugation, supernatant was used for further experiments. For detection, biotin-conjugated anti-KDR (BAF644; R&D Systems, Minneapolis, MN, USA) was used at 1 μg/μl.

Laser-induced choroidal neovascularization (CNV)

Laser photocoagulation and CNV measurement were described previously (11). Briefly, laser photocoagulation (532 nm, 150 mV, 100 ms, 100 μm) was performed on both eyes (2–5 spots/eye). Laser CNVs were stained with 5 μg/ml Alexa488 -onjugated isolectin GS-IB4 (Invitrogen), and observed by laser confocal microscope (Olympus America, Center Valley, CA, USA). Supplemental Fig. S1 summarizes the condition of intravitreous injection in each experiment. After photocoagulation, on d 1 and 4, 2 μl of 100 ng/μl STD_MO or moKDR_MOe13, 500 ng/μl goat IgG (AB-108-C; R&D Systems), 500 ng/μl goat anti-mouse VEGF-A IgG (AF-493-NA; R&D Systems), or 2 ng/μl SU1498 (572888; EMD Chemicals, Gibbstown, NJ, USA) was injected intravitreously.

TUNEL assay

TUNEL assay was performed using Click-iT TUNEL Alexa Fluor 594 Imaging Assay (C10246; Invitrogen). DPBS, STD_MO, or moKDR_MOe13 was injected into normal mouse eyes or laser-photocoagulated eyes (1 d postlaser) as described above. After 2 d, the eyes were fixed with 4% paraformaldehyde for 2 h at 4°C, and incubated in 15% sucrose for 2 h and 30% sucrose overnight. After embedding in optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA, USA), they were cut into 12-μm sections. Staining for apoptosis was conducted with the manufacturer‘s protocol. DNase-I treated section was used for positive control.

Transmission electron microscopy

To examine the integrity of fenestration in choriocapillaris after intravitreous moKDR_MOe13 injection, we examined the choroid with a transmission electron microscope, following previously described methods (12, 13).

Mouse corneal injury and observation of CD31 and LYVE-1 in cornea flat mount

Experimental conditions are listed in Supplemental Fig. S1. Under anesthesia, 15 μl of moKDR_MOe13 (40 ng/μl), STD_MO (40 ng/μl), or DPBS was injected subconjunctivally into 2 different places. The corneas were fixed in acetone at room temperature for 20 min. After 4 washes in PBST (0.1% Tween20/PBS), the corneas were incubated in 3% BSA/PBS at 4°C for 3 d. Then corneas were incubated in 3%BSA/PBS with FITC-conjugated rat anti-CD31 antibody (553372, 1:500; BD Biosciences) and rabbit anti-LYVE-1 (ab14917, 1:200; Abcam) overnight at 4°C. After 3 washes, the corneas were incubated in 3% BSA/PBS with Alexa Fluor 546-conjugated goat anti-rabbit IgG (A11071, 1:2000; Invitrogen) for 1 h at room temperature. After 4 washes, corneas were mounted on glass slides with Fluoro-gel (Electron Microscopy Sciences, Hatfield, PA, USA). Fluorescence was observed by fluorescence microscope (Carl Zeiss MicroImaging, Thornwood, NY, USA). The data for each suture were calculated separately by ImageJ (U.S. National Institutes of Health, Bethesda, MD, USA).

Mouse corneal transplantation

Mouse corneal transplantation was described previously (9, 14). The donor cornea was marked with 2 mm trephine, the anterior chamber was penetrated using a knife (ClearCut; Alcon, Hünenberg, Switzerland), and the cornea was cut with Vannas scissors and then placed in balanced salt solution. The recipient mouse was anesthetized with ketamine (100 mg/kg body weight) and xylazine (20 mg/kg body weight). To dilate the pupil and anesthetize the cornea, 1% tropicamide ophthalmic solution and 0.5% proparacaine ophthalmic solution were used. The recipient's right cornea was marked with 1.5 mm trephine and removed by the same method as the donor cornea. Viscoelastic material (Healon,1% sodium hyaluronate; Abbott Medical Optics, North Chicago, IL, USA) was used during recipient cornea dissection. The donor graft was sutured into the recipient bed using interrupted sutures (11-0 nylon, CS160-6; Ethicon, Somerville, NJ, USA). After the transplantation, the eye was covered with 0.5% erythromycin ophthalmic ointment, and the lid was sutured with 8-0 coated vicryl (BV130-5; Ethicon). All sutures remained for the first postoperative week. We injected 15 μl moKDR_MOe13 (40 ng/μl), STD_MO (40 ng/μl), or DPBS subconjunctivally on the day of transplantation and postoperative 1, 2, 3, and 4 wk (Supplemental Fig. S1). The corneal opacity was examined weekly using an operating microscope by the end point (8 wk). The opacity was graded (from 0 to 5) to determine graft rejection (15). Opacity of grade 3 or more was considered to be graft rejection. At 8 wk, corneas were harvested and subjected for CD31 and LYVE1 stain using the method described above.

RESULTS

Blocking exon 13-intron 13 junction in KDR leads to increased sKDR and decreased mbKDR at the mRNA level

At first, we examined whether modulation of splicing can lead to sKDR up-regulation and mbKDR down-regulation. To modulate splicing, we used antisense morpholino oligomers that bind mRNA or pre-mRNA with high specificity to inhibit translation and affect alternative splicing (16, 17). Antisense morpholino oligomers were designed corresponding to the junction of exon 13-intron 13 (KDR_MOe13; Fig. 1A and the sequences in Table 1). Although the human sKDR mRNA structure was not well characterized, we utilized the expressed sequence tag (EST) database and found the sequence of the initial 365 nt of intron 13, which include a stop codon at 48–50 nt. In addition, 2 polyadenylation sites (AAUAAA) are found in intron 13 (Fig. 1A). To introduce morpholino into cultured cells, we performed nucleofection. Fluorescent conjugated morpholino was used to confirm transfection into HUVECs (Fig. 1B). At 2 d after morpholino transfection, sKDR and mbKDR mRNA in HUVECs were examined by RT-PCR (Fig. 1C and Supplemental Fig. S2). mbKDR mRNA decreased in HUVECs transfected with KDR_MOe13 compared with DPBS- or STD_MO-transfected HUVECs. On the other hand, sKDR mRNA was increased by KDR_MOe13. To quantify these results, real-time PCR was performed (Fig. 1D, E). We found that KDR_MOe13 down-regulates mbKDR mRNA by 40% compared with STD_MO or DPBS (P<0.05). In contrast, sKDR mRNA showed a 17-fold increase (P<0.001) with KDR_MOe13.

Figure 1.

KDR_MOe13 decreases mbKDR mRNA and increases sKDR mRNA. A) Schematic design of KDR_MOe13. B) Fluorescent morpholinos can be observed after nucleofection. Indicated amounts of each morpholino were nucleofected into HUVECs. Scale bar = 100 μm. C) RT-PCR indicates mbKDR mRNA decrease and sKDR mRNA increase after KDR_MOe13. D, E) Quantitative real-time PCR shows 40% decrease of mbKDR and 17-fold increase of sKDR mRNA. All results were normalized by GAPDH and compared with DPBS-transfected HUVECs by 2-tailed Student‘s t test; n = 6/group. Bars represent means ± se. *P < 0.05, ***P < 0.001 vs. DPBS-transfected HUVECs.

KDR_MOe13 increased sKDR and decreased mbKDR protein

Next, we examined mbKDR protein expression by Western blot. KDR_MOe13 reduced mbKDR protein expression compared to DPBS and STD_MO (Fig. 2A). On flow cytometry, in DPBS- and STD_MO-transfected HUVECs, 83.3 and 81.0% were mbKDR positive, respectively, while KDR_MOe13 decreased mbKDR positive cells to 40.7% (Fig. 2B). To confirm sKDR protein expression, we performed Western blot from culture medium of KDR_MOe13-transfected HUVECs using an antibody recognizing KDR extracellular domains (Fig. 2C, D). Although the predicted molecular mass of human sKDR is ∼76 kD, 150-kD bands were detected from the culture medium of HUVECs transfected with KDR_MOe13. Based on previous studies, mbKDR can be glycosylated, increasing the molecular mass from 150 to 230 kD (18). In addition, sKDR has been detected at 160 kDa, although it was not reported whether this was derived from alternative polyadenylation or proteolytic cleavage from mbKDR (10). To confirm glycosylation of sKDR, we attempted deglycosylation of sKDR in medium from KDR_MOe13-treated HUVECs using TFMS (Fig. 2E). We found that the observed size of sKDR decreased from 150 to ∼70 kDa, confirming glycosylation of sKDR.

Figure 2.

KDR_MOe13 decreases mbKDR protein and increases sKDR protein. A) KDR_MOe13 decreased mbKDR protein expression, as detected on Western blot. B) Flow cytometry demonstrates that KDR_MOe13 decreases mbKDR cell surface expression. HUVECs stained with only secondary antibody were used as a negative control. C) Western blot from concentrated culture medium shows that KDR_MOe13 induced sKDR (150-kDa band). D) sKDR protein detection from conditioned culture medium without concentration. We exposed for a long period and enhanced the image. The 150-kDa band (arrow) was detected from KDR_MOe13-transfected HUVEC culture medium strongly. E) Deglycosylation of sKDR by TFMS. After deglycosylation, sKDR induced by KDR_MOe13 was observed at approximately calculated molecular weight (∼70 kDa).

Polyadenylation signal induced by KDR_MOe13 is the same polyadenylation site as in endogenous sKDR found in human cornea

Next, to determine the 3′ UTR of sKDR mRNA induced by morpholinos, 3′ RACE was performed (Fig. 3A). A strong band (∼1600 bp) was detected from KDR_MOe13-transfected HUVECs. Based on the sequence of this band, sKDR mRNA utilizes a polyadenylation site located in the 1403- to 1408-nt range of intron 13 (Fig. 3C). Based on this, we sought to identify the 3′ UTR of sKDR from human corneal total RNA (Fig. 3B). Corneal tissue is known to predominantly express sKDR (9). We found that in the human cornea, sKDR mRNA utilizes the same polyadenylation site that KDR_MOe13 induces (Fig. 3C).

Figure 3.

Polyadenylation signal induced by KDR_MOe13 was the same polyadenylation signal as in endogenous sKDR found in human cornea. A) 3′ RACE was performed from total RNA of DPBS-, STD_MO-, or KDR_MOe13-treated HUVECs. sKDR_F was used as a forward primer. A ∼1600-bp band was observed on administering KDR_MOe13. B) 3′ RACE was performed from human corneal total RNA. Cloning_F(1042-1061) was used as a forward primer. C) Sequence results of 3′ RACE products from sKDR induced by KDR_MOe13 (sequence 1) and sKDR of human cornea (sequence 2), along with genomic sequence of intron 13 in KDR gene (sequence 3). Arrow indicates cleavage site.

KDR_MOe13 suppressed experimental hemangiogenesis and lymphangiogenesis in vivo

Because KDR_MOe13 decreases mbKDR and increases sKDR, we predicted that KDR_MOe13 can inhibit hemangiogenesis and lymphangiogenesis in vivo. Therefore, we created moKDR_MOe13, which targets the exon 13-intron 13 junction of mouse KDR. By RT-PCR, we found that moKDR_MOe13 increased sKDR mRNA and decreased mbKDR mRNA in mouse MS-1 cells (Fig. 4A). To administer morpholinos in vivo, we developed morpholinos conjugated with dendrimer at the 3′ position (vivomorpholino; ref. 19). Vivomorpholino possesses guanidinium head groups, which are predicted to help morpholino transport into cells through binding with the phosphates of phospholipids. To determine moKDR_MOe13 function in vivo, each morpholino or DPBS was injected intravitreously, and the retinal total RNA was subjected to real-time RT-PCR for sKDR and mbKDR mRNA (Fig. 4B). moKDR_MOe13 significantly increased the sKDR/mbKDR mRNA ratio. Next, we examined mbKDR and sKDR retinal and vitreous protein by Western blot. Consistent with real-time PCR results, moKDR_MOe13 decreased retinal mbKDR protein (Fig. 4C) and increased vitreous sKDR protein (Fig. 4D). Next, we examined whether moKDR_MOe13 inhibits laser-induced CNV. Each morpholino or DPBS was injected intravitreously on d 1 and 4 after laser photocoagulation, and laser CNV volumes were examined on d 7. Figure 4E shows representative images of laser CNV with DPBS, STD_MO, and moKDR_MOe13. moKDR_MOe13 significantly suppressed laser CNV compared with STD_MO (P<0.05) and DPBS (P<0.01) (Fig. 4F). In addition, moKDR_MOe13 treatment was comparable to anti-VEGF-A IgG and KDR kinase inhibitor (SU1498) treatment (Fig. 4G). We also assessed apoptosis using TUNEL assay (Supplemental Fig. S3). Mild apoptosis was observed after treatment with moKDR_MOe13 and STD_MO. Furthermore, after morpholino intravitreous injection, fenestration in choriocapirallis was not affected by either moKDR_MOe13 or control (Supplemental Fig. S4). For a second model of angiogenesis, we used the corneal suture model. For the 1-wk subarm, each morpholino or DPBS was injected subconjunctivally 1 d before and 4 d after suturing; the corneas were harvested 7 d postsuturing. For the 2-wk subarm, each morpholino or DPBS was injected subconjunctivally 1 d prior and 4, 7, and 10 d after suturing; corneas were harvested at 14 d. CD31 and LYVE-1 were used as markers of neovascularization and lymphangiogenesis, respectively. Figure 4H, I displays representative images of CD31-stained corneas at 1 wk and LYVE-1-stained corneas at 2 wk. Figure 4J, K displays the mean area of neovascularization and lymphangiogenesis in each group. moKDR_MOe13 suppressed suture-induced neovascularization by 52.2% (1 wk) and 29.6% (2 wk) compared to DPBS (P<0.001 and 0.05, respectively). moKDR_MOe13 did not suppress lymphangiogenesis 1 wk postsuturing but suppressed lymphangiogenesis by 27.8% 2 wk postsuturing, compared to DPBS (P<0.05).

Figure 4.

moVEGFR2_MOe13 suppresses experimental neovascularization and lymphangiogenesis in mice. A) RT-PCR from MS-1 cells shows that moKDR_MOe13 decreases mbKDR mRNA and increases sKDR. B) Real-time PCR indicates that intravitreous injection of moKDR_MOe13 increases sKDR/mbKDR mRNA ratio in mouse retina (n=4). C, D) moKDR_MOe13 decreased mbKDR protein and increased sKDR protein in vivo, as demonstrated by Western blots of mouse retina and vitreous. E) Representative images of laser-induced CNV. F) moKDR_MOe13 suppressed laser-induced CNV relative to PBS or standard nonspecific morpholino (n=11–17). G) moKDR_MOe13 suppressed laser-induced CNV volume by comparable amounts relative to anti-mouse VEGF-A antibody and the KDR kinase inhibitor SU1498 (n=14–18). ANOVA: P = 0.000255. H, I) Representative images of corneal neovascularization 1 wk after suture (H) and corneal lymphangiogenesis 2 wk after sutures (I). J, K) moKDR_MOe13 suppressed neovascularization at 1 and 2 wk (J), and decreased lymphangiogenesis at 2 wk (K) compared with the controls. Scale bars = 100 μm. P values were calculated by 2-tailed Student's t test (n=13–16). Bars represent means ± se. *P < 0.05, **P < 0.01, ***P < 0.001.

KDR_MOe13 reduced graft rejection after mouse corneal transplantation

Finally, we examined whether moKDR_MOe13 suppressed murine corneal transplant rejection, which is pathologically dependent on both hemangiogenesis and lymphangiogenesis. After cornea transplantation, moKDR_MOe13, STD_MO, or DPBS was injected subconjunctivally. The corneas were observed with the stereomicroscope (Fig. 5A). We found that moKDR_MOe13 increased graft survival compared with DPBS and STD_MO (Fig. 5B, log rank test: P=0.0186 and 0.0610, respectively). Figure 5C shows the representative images of CD31 and LYVE-1 stained cornea at the end point (8 wk). Concordantly, moKDR_MOe13 significantly reduced neovascularization and lymphangiogenesis (Fig. 5D, E).

Figure 5.

moKDR_MOe13 suppresses graft rejection in mouse corneal transplantation. A) Representative stereomicroscope images 3 wk after corneal transplantation show that corneal transplants treated with moKDR_MOe13 were clear and had less inflammation than the others. B) Cumulative graft survival rate. moKDR_MOe13 increased graft survival rate compared with DPBS and STD_MO (log rank test: P=0.0186 and 0.0610, respectively). Arrow indicates censored data. C) Representative images of corneal neovascularization and lymphangiogenesis in mouse corneal transplants at 8 wk. Scale bar = 1 mm. D, E) moKDR_MOe13 suppressed neovascularization (D) and lymphangiogenesis (E) in mouse corneal transplants at 8 wk. P values were calculated by 2-tailed Student‘s t test (n=11–17). Bars represent means ± se. *P < 0.05, **P < 0.01).

DISCUSSION

Generally, alternative splicing produces a variety of isoforms from a single gene and contributes diversity. For example, the VEGF-A gene produces VEGF121, VEGF165, VEGF189, and VEGF206 by alternative splicing, and the products may contribute to forming the gradient of VEGF-A (20–22). Compared with alternative splicing, alternative polyadenylation is not yet well understood despite its importance. In humans and mice, EST database analysis indicated that ∼60% of mRNA may result from alternative polyadenylation (23). We can divide alternative polyadenylation into 2 classes. In the first class, the 3′ UTR has ≥2 polyadenylation sites. Depending on which polyadenylation site is used, the length of 3′ UTR changes. The difference of 3′ UTR length affects stability and localization of mRNA (24–26). In the second class, the active polyadenylation site exists in a different exon or intron. The example of this type of alternative polyadenylation is the immunoglobulin system in B cells (27, 28). This type of polyadenylation is believed to be associated with an alternative splicing event (29). The KDR gene would represent this type of alternative polyadenylation.

In this study, we demonstrate that the latent polyadenylation site in intron 13 of KDR can be activated by blocking the upstream 5′ splicing site (exon 13-intron 13 junction) using KDR_MOe13, which decreased mbKDR and increased sKDR at mRNA and protein levels. The polyadenylation site induced by KDR_MOe13 is normally inactive in HUVECs, preferentially excluding intron 13 during physiological splicing. KDR_MOe13 likely competes with U1snRNPs at the exon 13-intron 13 junction. U1snRNPs may inhibit downstream polyadenylation and are among the key components for the splicing event, although U1snRNP-independent RNA splicing has been demonstrated (30–32). It is probable that KDR_MOe13 activates the latent polyadenylation signal by inhibiting U1snRNP binding to the exon 13-intron 13 junction.

Interestingly, the polyadenylation site induced by KDR_MOe13 is the same polyadenylation site as in endogenous sKDR found in human cornea. The sequence around the polyadenylation site indicates that a cleavage site (CA dinucleotides) and a GU-rich region are located 26 and 71 nt downstream of AAUAAA, respectively (Fig. 3C). These sequence components are similar to typical polyadenylation signals (29, 33). Without any direct modification of the polyadenylation mechanism, KDR_MOe13 activates the latent polyadenylation signal. Thus, mbKDR-preferred cells, such as HUVECs, are still able to recognize the latent polyadenylation signal. This indicates that endogenous sKDR may be produced by splicing factor control, rather than polyadenylation mechanism modification. To understand altenative polyadenylation events, splicing factors should be focused rather than polyadenylation mechanisms themselves.

sKDR protein induced by KDR_MOe13 was processed (glycosylation) and exported to culture medium in vitro. These results indicated that sKDR has similar binding capacity of mbKDR to VEGF-C and inhibits it in extracellularly. We have shown the effectiveness of morpholinos in vivo by using moKDR_MOe13 to suppress corneal neovascularization and lymphangiogenesis in a suture injury model in mice. These results suggested also that sKDR induced by KDR_MOe13 has binding capacity to VEGF-C.

siRNA therapies to block angiogenesis are currently under development. We propose the use of morpholinos to shift the functionality of proangiogenic mbKDR to antiangiogenic sKDR as an alternative avenue for developing antiangiogenic therapies. siRNAs can knock down target mRNA but lacks the distinct advantage that morpholinos have in being able to not only knock down an undesirable protein, but also produce a desirable protein at the same time. The dual effect of knocking down one protein, while concurrently increasing the translation of another, could offer therapeutic advantages in the ability to more precisely regulate protein levels. In addition, unlike siRNA, morpholinos are of neutral charge and induce a lesser inflammatory response and less off target binding.

We used a vivomorpholino in vivo study, although the dendrimer that enables cell entry has been reported to have mild toxicity (34), and we found some toxicity (Supplemental Fig. S3). To advance this further toward clinical translation, in future studies, we will test cell-penetrating peptides such as polyarginine or cholesterol conjugation (to avoid use of dendrimers). Future investigations will also explore the distribution and duration of morpholinos in vivo using novel but currently unavailable vivomorpholinos with fluorophores. To target delivery of morpholinos, nanoparticles may also be employed.

Another potential application for our morpholino-targeting KDR would be the area of oncology. Solid tumors rely on both hemangiogenesis and lymphangiogenesis for growth and metastasis. The approach described herein holds great potential for solid tumor therapy as it simultaneously suppresses formation of both blood and lymphatic vessels.

A recent report showed that polyadenylation activation could induce soluble isoforms of several receptor tyrosine kinases, including KDR (35). We confirm the concept of activating a latent polyadenylation signal using morpholino oligomers and demonstrate in vivo utility in suppressing hemangiogenesis and lymphangiogenesis with therapeutic effect in 3 models of blinding disorders, including macular degeneration, corneal injury, and corneal transplant rejection. This has applications not only for antiangiogenesis or eye disease by targeting KDR but also in other conditions where regulatory manipulation of splicing and polyadenylation could have therapeutic valence.