Overexpression of a mutant form of TGFBI/BIGH3 induces retinal degeneration in transgenic mice

Lentivirus generation

Recombinant lentiviral particles containing a vesicular stomatitis virus G protein (VSV-G) envelop were produced by transient transfection of 293T cells as previously described [6]. Briefly, 293T cells were transfected with the different plasmids using calcium phosphate, the medium harvested 36 h later and centrifuged at 70,000x g. Viral supernatants were concentrated by a second ultracentrifugation at 70,000x g for 90 min at 4 °C. The viral pellets were resuspended in a minimal of volume of PBS containing 10 mg/ml bovine serum albumin (BSA). Aliquots of 5 µl were then stored at −70 °C until further use. Total particle concentration of the viral stocks was estimated by quantification of the p24 capsid protein using Retro-TEK HIV-1 p24 Antigen ELISA kit (ZeptoMetrixCorporation, Buffalo, New York).

High transduction efficiency was observed in corneal endothelial cells in vivo, using an HIV-1-derived lentivirus with a cytomegalovirus immediate early promoter (CMV). A similar efficacy was reported for rat, ovine, and human cornea explants infected in vitro with a lentiviral vector carrying a simian virus (SV) promoter [7,8]. Previous results have shown similar efficacy in mouse cornea using HIV or equine infectious anemia virus (EIAV) vectors containing the vectors containing the VSV-G envelop (VSVG) envelop and the CMV vector [9]. In cell culture, EIAV has also shown a strong efficiency to transduce human endothelial cells using a CMV promoter and the VSV-G envelop [10].

A human TGFBI/BIGH3 cDNA with the R555W Groenouw mutation was ligated into BamHI sites of a replication-deficient, self-inactivating plasmid (pSIN) under the regulation of the mouse phosphoglycerate kinase promoter (PGK; Figure 1). The PGK promoter is a ubiquitous promoter that was successfully used with lentiviral vectors in many tissues [11,12]. Since the GC-rich sequence of the mouse PGK promoter contains only three ATG triplets, this promoter is particularly suitable for lentiviral constructions as compared to other ubiquitous promoters [13]. Presence or absence of mutation was confirmed by direct sequencing of the insert.

Lentiviral vector preparation

A replication-deficient, self-inactivating (SIN) backbone was obtained from Patrick Aebischer (Ecole Polytechnique Fédérale de Lausanne [EPFL], Lausanne, Switzerland). The transgene encoding the human TGFBI/BIGH3 cDNA was inserted into a backbone containing the SIN followed by the elongation factor-1 minimum promoter and the woodchuck hepatitis post-transcriptional regulatory elements (SIN-W-EFs) to generate the lentiviral vector LV_BIGH3. The lentiviral vector was produced as described above and previously by Naldini [14].

Generation of transgenic animals

All procedures implicating mice manipulation were performed in agreement with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the local Veterinary Committee Office on Use and Care of Animals in Research of Canton Vaud, Lausanne, Switzerland. Mice (120 total, 60 males and 60 females) were maintained on a 14:10 light-dark cycle under standard housing conditions. Water and standard high protein food (Kliba, Switzerland) were available at libitum.

Transgenesis was performed according to standard protocols [15]. Briefly, four-week-old National Marine Research Institute (NMRI) females were superovulated by intraperitoneal injection of 5U pregnant mare’s serum gonadotropin followed by injection of 5U human chorionic gonadotropin 48 h later. Superovulated females were mated with NMRI males, and embryos were collected at zygote stage on the following day. Zygotes were then immediately microinjected into the perivitelline space with 10-100 picoliters of about 109 transduction unit/ml lentiviral preparation and incubated in M16 medium (Sigma, St. Louis, MO) at 37 °C and 5% CO₂ atmosphere. On the following day, two-cell stage embryos were selected and transferred into the oviduct of 3 pseudopregnant NMRI females.

NMRI mice are widely used as experimental models in many fields of general biology. Albino NMRI mice are widely used in transgenesis since embryos derived from this strain are more resistant than inbred mouse embryos such as C57/BL6 and their use results in a higher transgenesis rate. Moreover, outbred mice embryos are more resistant than inbred mice embryos and lead to a higher transgenesis rate.

RNA extraction

Several organs including retina and cornea were removed from euthanized mice and directly placed in ice. Euthanasia was performed by neck dislocation. Total RNA was extracted using Trizol® reagent according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). RNA concentration and purity were measured using RiboGreen® (Molecular Probes, Eugene, OR) according to the manufacturer’s instructions.

cDNA synthesis

Equal amounts (1 µg) of RNA were used to synthesize cDNA using StrataScript Reverse Transcriptase according to the manufacturer’s instructions (Stratagene, La Jolla, CA).

Genomic DNA extraction

Genomic DNA was obtained from mouse tail biopsies as follows: cut tail was transferred into lysis buffer, which contained 100 mM TrisHCl, pH 8.5, 5 mM EDTA, 0.2% sodium dodecyl sulfate (SDS), 200 mM NaCl, and 100 µg proteinase K/ml, and incubated in a rotating shaker at 56 °C overnight. After complete lysis, samples were centrifuged for 2 min and the supernatant was transferred into isopropanol. After a second 2 min centrifugation, supernatant was removed and the pellet was dissolved in 150 µl 10 mM TrisHCl and 0.1 mM EDTA, pH 7.5, and incubated at 80 °C for 4 min. The obtained genomic DNA was directly used to perform PCRs.

Transgene integration, expression, and monitoring

Transgene integration was determined from biopsies by PCR amplification of TGFBI/BIGH3. Genomic DNA (500 ng or 1 µg), obtained as described in the previous section, was amplified. If not otherwise stated, the conditions were as follows: a first incubation of 5 min at 94 °C, then 35 cycles of 1 min at 94 °C, 1 min at 56 °C and 3 min at 72 °C, and finally a last extension of 10 min at 72 °C. The following primers were used: Big-Big_F, Big-Big_R, PGK-BIG_F, and PGK-BIG_R (Table 1). Since the human transgene sequence contained a unique BamH1 site in a specific PCR-amplified region, transgene transcription was monitored by digestion of the PCR amplified cDNA produced by reverse-transcribed RNA. PCR was performed from 100 to 500 ng cDNA amplified by Hotstart Taq polymerase (Qiagen, Hilden, Germany), with the primers 227F and 649R (Table 1) and the following conditions: a first incubation of 5 min at 95 °C, then 30 cycles of 1 min at 95 °C, 1 min at 55 °C and 1 min at 72 °C, with a final extension of 10 min at 72 °C. BamHI digestion was performed from 0.1 to 1 µg gel purified PCR product incubated 8-16 h at 37 °C with 1 unit of enzyme in Roche B buffer. Digestion of endogenous mouse TGFBI/BIGH3 resulted in a product of about 500 bp whereas BamHI digestion of transgenic (human) TGFBI/BIGH3 resulted in two bands of about 300 bp and 200 bp.

The number of transgene integration sites was monitored by Southern blot. Genomic DNA obtained was digested with EcoRI overnight at 37 °C. The digested DNA was precipitated with 3M sodium acetate (pH 5.2)/EtOH on ice and resuspended in trishydroxymethylaminomethane – ethylene diamine tetraacetic acid (TE) buffer. A full-length TGFBI/BIGH3 cDNA was radiolabeled with the random primed technique and hybridized to the membrane. Hybridization was performed with a probe from the vector’s backbone (Figure 1). Integration sites were also monitored by linear amplification mediated polymerase chain reaction (LAM-PCR) as described [16]. Briefly, a linear PCR with a long-terminal repeat (LTR)-specific biotinylated primer (LTR1 in Table 1) was performed. The specific amplified PCR products were sorted by magnetic tag selection of extension primers. A second DNA strand of each target sequence was synthesized by random hexanucleotide priming. Double-stranded DNA was digested with CfoI, and an asymmetric ligation cassette was ligated to the end of CfoI-digested fragments. Nested exponential PCR amplifications were then performed with linker cassette (LC)-specific forward primers (LC primers, Table 1) and LTR-specific reverse primers (LTR primers Table 1). Primer pairs were LC1-LTR2, LC2-LTR3, and LC3-LTR4. Finally, LAM-PCR products were subcloned in pGEM-T easy vector, sequenced, and blasted against mouse genomic DNA.

Protein extraction

Retina and cornea were removed from dissected mouse eyes. Proteins were extracted and homogenized for 10 min in 100 µl ice-cold RIPA buffer, which consisted of 50 mM Tris-Cl pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate, and 0.1% SDS, and also contained protease (Roche, Mannheim, Germany) and 1% phosphatase (Sigma) inhibitor cocktails. Further homogenization through a needle was performed when necessary. Retinas or corneas were homogenized in ice-cold RIPA buffer with an 18 gauge needle. Protein concentrations were measured by bisinchoninic acid (BCA) protein assay (Pierce, Rockford, IL).

Western blot analysis

Equal amounts of protein samples were separated by 8% SDS–PAGE gel and transferred to a polyvinylidene difluoride membrane (Westran clear signal; Whatman Inc., Maidstone, UK). Membranes were then blocked for 1 h at room temperature with Tris-buffered saline, composed of 150 mM NaCl and 50 mM Tris, pH 7.5, and contained 0.1% Tween (TBS-T) and 5% nonfat dry milk. Next, membranes were probed with a polyclonal rabbit antiserum against human KE-2 (1/4,000) [17] for 2 h in TBS-T with 5% nonfat dry milk. A secondary horseradish peroxydase-coupled goat antirabbit IgG (1/25,000; Amersham Biosciences, Buckinghamshire, UK) was used for 1 h at room temperature in TBS-T with 5% nonfat dry milk. After three washes in TBS-T, the antigen-antibody complexes were detected by chemiluminescence using the Enhanced Chemi Luminescence Plus (ECL+) western blotting system (Amersham Biosciences). Results were quantified under conditions of linearity by integration of the density of total area of each band using a video densitometer and LabWorks 4.0 software (Media Cybernetics, Bethesda, MD). Results are expressed as a percentage of the control optical density.

Hematoxylin-eosin and DAPI staining of mouse eyes

Mouse eyes (8 transgenic and 3 wild type) were dissected and fixed in 4% paraformaldehyde for 1 h. Fixed eyes were then embedded in 30% albumin egg, 3% gelatin (Yazzula buffer) and frozen in liquid nitrogen to be sectioned at 14 µm with a cryostat. Slides were rehydrated and stained with hematoxylin for 10 min. After two washes, slides were stained with eosin for 1 min and fixed in 50% glycerol-50% PBS.

Different slides were stained for 15 min with DAPI safe from the light and then fixed with FluorSave. In order to analyze similar fields, we evaluated slides on which the optical nerve was visible on a BX51 Olympus microscope at 100X or 400X magnification. Pictures were taken using Cell^D imaging software (Carl Zeiss, Jena, Germany).

Autopsy

Eight-month-old transgenic mice (two males and two females) and wild-type (one male and one female) living mice were sent to Frimorfo Inc. (Fribourg, Switzerland) for an in depth analysis of all organs. This analysis includes macroscopic inspection, necroscopy and histopathological evaluation completed by full X-ray analysis and blood analysis (blood chemistry and hematology).

Generation and validation of the transgenic model

In total, 25 independent PGK-Groenouw transgenic founders were generated, with a transgene integration averaged ratio of 37%±18 of the newborns, depending on the lentiviral batch and titer. In this study, we observed that only batches of lentiviral particles titrated above 1x10⁸ TU/ml were effective in generating transgenic animals. We routinely screened the mice by PCR for transgene integration, using primers specific for the transgene and RT–PCR for transcript expression (Figure 2). As described in the previous section, digestion of endogenous mouse TGFBI/BIGH3 resulted in only one band whereas BamHI digestion of transgenic (human) TGFBI/BIGH3 resulted in two bands.

Southern blot analyses were used to evaluate the number of integration sites in the F1 generation of transgenic mice. Generally one to four sites were observed (Figure 3A). Based on the size of the fragments, and assuming no recombination in the region of the probe, five integration sites were seen in the Southern blot analysis of transgenic mice. It is assumed that, with successive generations, the number of integration sites will diminish. Moreover, localization of integration sites were monitored in some transgenic animals using linear amplification mediated polymerase chain reaction (LAM-PCR) method. Integration of lentiviral vectors were seen in chromosomes 7, 8, 14, 19, and Y. Table 2 indicates the exact positions of the identified integration sites.

We additionally monitored tissue expression in glycerol-embedded sectioned tissues (not shown) and by western blot using an anti-KE2 antibody in transgenic mice expressing the R555W mutation under the control of the PGK promoter. KE expression was slightly increased in the retina of transgenic mice compared to wild-type mice (Figure 3B, left panel). We also observed a clear expression of KE in the brain of transgenic mice, whereas no expression was found in wild-type mice (Figure 3B). Overall, it did not seem to affect endogenous expression in other tissues (Figure 3B, left and right panels). The figure also showed smaller bands corresponding to degradation products of the protein.

Organ comparisons between phosphoglycerate kinase promoter-Groenouw transgenic and wild-type mice

The weight of the organs of eight-month-old transgenic mice (two males and two females) was compared to age- and sex-matched control mice. Transgenic mice appeared to be healthy and did not display particular abnormal behavioral signs for up to 18 months.

Transgenic mice tended to have increased organ size and weight, except for the thymus that was reduced in size compared to age-matched wild-type NMRI mice (Table 3). Transgenic female mice showed follicular hyperplasia (Figure 4, lower panel), increased amount of extramedullary hematopoiesis, and higher spleen weights (Table 3). The weight of the heart and kidneys was higher in transgenic females (Table 3), but morphology was normal.

We did not observe other differences between males and females for relative weight or histology of other organs, compared to non-transgenic animals. Overall, no statistical differences were observed in protein concentrations (Table 4A) or in cell composition (Table 4B) of blood between transgenic mice and wild-type mice. One transgenic female presented lipogenic pigment accumulation in the adrenal gland.

Transgenic-dependent retinal degeneration

Transgenic expression of KE was observed in the retina but not in the cornea of transgenic animals. We therefore investigated the effect of TGFBI/BIGH3 overexpression in retinal tissue. While electroretinogram (ERG) response to increasing intensities flashlights were normal in 2-month-old transgenic mice, we observed a reduction in a- and b-waves in more than 50% of the older mice (Figure 5). In scotopic conditions, at an illumination of 100 mcds/s, 2-month-old mice showed a mean b-wave of 346 mV whereas this mean measure falls to 211 mV at 6 month. Four out six (66.7%) mice showed b-waves even lower than the mean measure. In comparison and in the same conditions, NMRI wild type mice showed no difference in mean values from 2 to 6.5 month of age. Mice with flat ERG also showed a diminution of the retina thickness and a global disorganization of the cell disposition in the retina of transgenic mice expressing the Groenouw mutation compared to control mice (Figure 6). Quantifications of the outer and inner nuclear layers thickness showed a significant diminution of each layer corresponding to a global diminution in transgenic mice (Figure 6B).