Methods of delivering a polypeptide molecule to Otx2 target cells using an Otx2 targeting peptide

Abstract

The disclosure provides a method of delivering a polypeptide molecule to an Otx2 target cell, including contacting the target cell with a chimeric polypeptide having (i) a targeting peptide consisting of SEQ ID NO: 2 and (ii) the polypeptide molecule.

Claims

1. A method of delivering a polypeptide molecule to an Orthodenticle Homeobox 2 (Otx2) target cell selected from retinal ganglion neurons, retinal bipolar neurons and parvalbumin neurons, comprising contacting the target cell with a chimeric polypeptide consisting of: (i) a targeting peptide consisting of the amino acid sequence SEQ ID NO: 2 bonded with (ii) a heterologous polypeptide molecule which is not an Otx2 protein, wherein the targeting peptide is located at the C-terminus of the chimeric polypeptide.

2. A method of delivering a polypeptide molecule to an Otx2 target cell selected from retinal ganglion neurons, retinal bipolar neurons and parvalbumin neurons in a subject comprising delivering a polypeptide molecule to the target cell according to the method of claim 1, wherein the delivering comprises administering the polypeptide molecule to the subject by infusion.

3. A method of delivering a polypeptide molecule to an Orthodenticle Homeobox 2 (Otx2) target cell selected from retinal ganglion neurons, retinal bipolar neurons and parvalbumin neurons, comprising contacting the target cell with a chimeric polypeptide consisting of: (i) a targeting peptide consisting of the amino acid sequence of SEQ ID NO: 2 bonded with (ii) a heterologous polypeptide molecule which is not an Otx2 protein, and (iii) a VP16 trans-activator domain of herpes virus, wherein the targeting peptide is located at the C-terminus of the chimeric polypeptide.

4. The method according to claim 3, wherein the VP16 trans-activator domain contains the amino acid sequence of SEQ ID NO: 4.

5. A method of delivering a polypeptide molecule to an Otx2 target cell selected from retinal ganglion neurons, retinal bipolar neurons and parvalbumin neurons in a subject comprising delivering a polypeptide molecule to the target cell according to the method of claim 3, wherein the delivering comprises administering the polypeptide molecule to the subject by infusion.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1 depicts, diagrammatically, various constructions carried out. signal seq.: signal peptide of alkaline phosphatase; Alkaline Phosphatase: alkaline phosphatase, Nt: N-terminal region of Otx2 (amino acids 1-37); Hd: homeodomain of Otx2 (amino acids 38-97); Ct: C-terminal region of Otx2 (amino acids 98-289); 6xHis: polyhistidine tag.

(2) FIG. 2 depicts results obtained with the AP-Nt-Otx2 and AP-Hd-Otx2 fusion proteins. A: Labeling obtained with the AP-Nt-Otx2 fusion protein; B: Labeling obtained with the AP-Hd-Otx2 fusion protein; CP: Cells of the retinal pigment epithelium; Cb: bipolar cells; RGC: ganglion cells.

(3) FIG. 3 depicts results obtained with mutant AA-Otx2. A: Labeling obtained with the RK-Otx2 peptide; B: Labeling obtained with the AA-Otx2 peptide; CP: Cells of the retinal pigment epithelium; Cb: bipolar cells; RGC: ganglion cells.

(4) FIG. 4 depicts results obtained with RK-Otx2 and AA-Otx2. A: Binding of AP-Nt-Otx2 in the presence of the RK-Otx2 peptide; B: Binding of AP-Nt-Otx2 in the presence of the AA-Otx2 peptide; CP: Cells of the retinal pigment epithelium; Cb: bipolar cells; RGC: ganglion cells.

(5) FIG. 5 depicts results for the level of expression of Brn3A mRNA.

(6) FIG. 6 depicts results obtained with cells of cerebral cortex. A-D, scale bar 500 m; (A) AP-Nt-Otx2; (B) AP (alkaline phosphatase alone); (C) AP-Nt-Otx2 in the presence of a whole Otx2; (D) AP-Hd-Otx2; E-H, scale bar 100 m; (E) AP-Nt-Otx2; (F) AP-HdAA-Otx2; (G) AP-Nt-Otx2 in the presence of RK-Otx2; (H) AP-Nt-Otx2 in the presence of AA-Otx2.

(7) FIG. 7 depicts results from experiments to determine whether the perineuronal net associated with parvalbumin neurons was involved in the binding between AP-Nt-Otx2 and its target cells. (A) and (B) labeling with WFA; (C) and (D) incubation in the presence of AP-Otx2; (A) and (C) untreated sections; (B) and (D) sections treated with chondroitinase ABC.

(8) FIG. 8 depicts results obtained with RK-Otx2 peptide. A-E: Infusion of RK-Otx2 (scale bar=100 m); A, B: Detection of the RK-Otx2 peptide; A: control hemisphere; B: infused hemisphere; C, D: Detection of Otx2; C: control hemisphere; D: infused hemisphere; E: Quantification of the cells expressing Otx2; black bars: control hemisphere; gray bars: infused hemisphere; F-J: Infusion of AA-Otx2 (scale bar=100 m); F, G: Detection of the AA-Otx2 peptide; A: control hemisphere; B: infused hemisphere; H, I: Detection of Otx2; H: control hemisphere; I: infused hemisphere; J: Quantification of the cells expressing Otx2; black bars: control hemisphere; gray bars: infused hemisphere; *p<0.005, paired Student's t-test; the error bars represent the standard error of the mean.

(9) FIG. 9 depicts results obtained from the experiments of Example 5. A: Contralateral bias index after infusion of the RK-Otx2 peptide (RK) or of the AA-Otx2 peptide (AA), or after injection of chondroitinase ABC (chABC) or of injection buffer alone (Veh); B-E: Labeling with WFA (B and C) and expression of parvalbumin (D and E) in the supragranular region of the visual cortex of the hemisphere infused with the RK peptide (C and E), and of the noninfused hemisphere (scale bar=100 m); F: Quantification of the cells labeled with WFA and of the cells expressing parvalbumin after infusion of the RK-Otx2 peptide (RK) or of the AA-Otx2 peptide (AA); black bars: control hemisphere; gray bars: infused hemisphere; *p<0.05, paired Student's t-test; the error bars represent the standard error of the mean.

EXAMPLE 1: IDENTIFICATION OF A SEQUENCE WHICH TARGETS OTX2 TO RETINAL GANGLION AND RETINAL BIPOLAR NEURONS

(10) During previous experiments (cf. application PCT/FR 2009/000031 of Jan. 9, 2009), it was noted that Otx2 injected into the eye was concentrated essentially in the retinal ganglion neurons (RGCs).

(11) In order to investigate whether a cell targeting domain was present in the sequence of Otx2, chimeric proteins comprising an alkaline phosphatase domain coupled to the whole Otx2 protein or to various fragments of this protein were constructed.

(12) The following fusion proteins were constructed:

(13) Alkaline Phosphatase-whole Otx2 (AP-Otx2)

(14) Alkaline Phosphatase-C-terminal region+homeodomain of Otx2 (AP-Ct-Otx2)

(15) Alkaline Phosphatase-N-terminal region+homeodomain of Otx2 (AP-Nt-Otx2)

(16) Alkaline Phosphatase-homeodomain of Otx2 (AP-Hd-Otx2).

(17) The sequences encoding the human Otx2 protein or encoding the fragments tested were cloned into the vector pAPtag-5 (GenHunter), in frame with the sequence encoding alkaline phosphatase. The various constructions carried out are represented diagrammatically in FIG. 1. HEK 293 cells cultured in culture dishes 10 cm in diameter were transfected with 10 g of each of the vectors constructed, purified beforehand, using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. The transfected cells were incubated for 48 hours in DMEM/F12, supplemented with 10% of fetal bovine serum (FCS). The supernatant was collected, centrifuged at 100g for 5 minutes and stored at 80 C. This supernatant is used, without purification, for testing the binding of the chimeric proteins on sections of retina.

(18) To carry out the binding test, cryostat sections (20 m) of frozen fresh retinas are fixed for 8 minutes in 100% cooled methanol, and then washed 3 times 10 in phosphate buffer (PBS) with 4 mM of MgCl.sub.2.

(19) The sections are incubated in PBS buffer, 4 mM MgCl.sub.2, plus 10% FCS for 1 hour at room temperature (RT).

(20) For the binding, the supernatants containing the fusion proteins tested are diluted to 1/20 in PBS and incubated for 2 hours at room temperature.

(21) The sections are then washed 5 times in PBS, 4 mM MgCl.sub.2, and the bound ligands are then fixed for 2 minutes (60% acetone, 4% PFA, 20 mM Hepes, pH 7).

(22) After 3 washes in PBS, the sections are heated at 65 C. for 2 hours in PBS in order to inactivate the endogenous phosphatases. The sections are then washed twice in PBS before visualization of the alkaline phosphatase activity (preincubation of the sections for 5 minutes in 100 mM Tris, pH 9.5, 100 mM NaCl, 5 mM MgCl.sub.2, followed by the addition of NBT/BCIP (Promega)).

(23) The results obtained with the AP-Nt-Otx2 and AP-Hd-Otx2 fusion proteins are illustrated by FIG. 2.

(24) These results show that the fusion protein comprising the N-terminal domain of Otx2 and its homeodomain binds specifically to the RGCs and the bipolar cells, whereas the fusion protein containing only the homeodomain does not bind to any of the retinal cells.

(25) Biotinylated peptides corresponding to various fragments of the N-terminal domain+homeodomain portion of Otx2 were synthesized and the binding thereof to sections of retina, prepared as described above, was tested. The incubation of the peptides with the sections of retina was carried out under the same conditions as those described above, and then the sections bearing the bound peptides were incubated with alkaline phosphatase-labeled streptavidin, and the alkaline phosphatase activity was detected as described above.

(26) One of the peptides tested (RK-Otx2), corresponding to the sequence RKQRRERTTFTRAQL (SEQ ID NO: 2), has the same binding specificity as the AP-Otx2 and AP-Nt-Otx2 fusion proteins.

(27) Mutations were then created in this polypeptide. One of the mutants (AA-Otx2), in which two basic amino acids (RK) are exchanged for two neutral amino acids (AA) and which therefore corresponds to the sequence AAQRRERTTFTRAQL (SEQ ID NO: 3), does not inhibit any binding to the retinal cells.

(28) These results are illustrated by FIG. 3.

(29) These results show that the RK-Otx2 peptide binds, like AP-Nt-Otx2, to the ganglion and bipolar cells. On the other hand, no binding is observed with the AA-Otx2 peptide. Another peptide (RA-Otx2), in which the RK dipeptide is replaced with RA, binds only very weakly to the ganglion and bipolar cells (results not shown).

(30) In order to verify that the RK-Otx2 peptide indeed had the same binding specificity as AP-Nt-Otx2, the ability of the RK-Otx2 and AA-Otx2 polypeptides to antagonize the binding of AP-Nt-Otx2 was tested. The test for binding of AP-Nt-Otx2 to sections of retinas was carried out as described above, with the exception that the incubation of the supernatant containing AP-Nt-Otx2 was carried out in the presence of 2 g/ml of the RK-Otx2 peptide or of the AA-Otx2 peptide.

(31) The results are illustrated by FIG. 4.

(32) These results show that the RK-Otx2 peptide blocks the binding of AP-Nt-Otx2 to the bipolar cells and to the ganglion cells, whereas the AA-Otx2 peptide has no effect on this binding.

EXAMPLE 2: EFFECT OF A CHIMERIC POLYPEPTIDE ASSOCIATING AN OTX2 FRAGMENT CONTAINING THE CELL TARGETING SEQUENCE WITH A HETEROLOGOUS TRANSCRIPTION-ACTIVATING DOMAIN, ON THE SURVIVAL OF RETINAL GANGLION NEURONS

(33) It has previously been shown (cf. application PCT/FR 2009/000031 of Jan. 9, 2009) that Otx2 protects retinal ganglion neurons against the toxic effects of N-methyl-D-aspartate (NMDA).

(34) A chimeric polypeptide was constructed genetically and produced by bacterial synthesis, by fusing the N-terminal domain of Otx2 and its homeodomain (amino acids 1-97 of Otx2), with the VP16 trans-activator domain of the herpes virus (MLGDGDSPGPGFTPHDSAPYGALDMADFEFEQMFTDALGIDEYGG, SEQ ID NO: 4).

(35) C57 B16 mice received, in the right eye, 1 l of injection buffer (PBS or 9 NaCl) containing either 1 mM of NMDA, or 1 mM of NMDA supplemented with 30 ng of the chimeric polypeptide, and in the left eye, the same volume of injection buffer, without additive.

(36) The survival of the ganglion neurons was determined by measuring the level of expression of Brain 3A (Brn3A), a transcription factor which, in the retina, is specifically expressed in the ganglion neurons (Xiang et al., J. Neurosci., 15, 4762-4785, 1995).

(37) After 4 days, the animals are sacrificed, the retinas are removed, and the mRNA is extracted therefrom.

(38) The level of expression of Bm3A mRNA was determined by quantitative RT-PCR using the hypoxanthine phosphoribosyltransferase (HPRT) gene as reference gene, and the ratio between the expression of the Bm3A mRNA in the right eye and in the left eye was calculated.

(39) The results are illustrated by FIG. 5. The additives used are indicated along the x-axis; the ratio between the amounts of Brn3A mRNA (standardized relative to the HPRT mRNA) in the right eye and in the left eye is indicated along the y-axis.

(40) These results show that NMDA, administered alone, significantly decreases (by approximately 60%) the amount of ganglion neurons, and that the addition of 30 ng of the chimeric polypeptide effectively protects the ganglion neurons against the toxic effects of the NMDA.

EXAMPLE 3: BINDING OF OTX2 TO TARGET CELLS OF THE CEREBRAL CORTEX

(41) The interaction, with the cells of the cerebral cortex, of the AP-Otx2, AP-Nt-Otx2, AP-Hd-Otx2 and AP-HdAA-Otx2 fusion proteins (AP-Nt-Otx2 being a variant of AP-Nt-Otx2 in which the RK amino acid doublet has been replaced with the AA doublet), alone (culture supernatant diluted to 1/20.sup.th) or in the presence of a whole Otx2 (1 g/ml) or of the RK-Otx2 or AA-Otx2 peptides (2 g/ml), was tested on cryostat sections of adult mouse brains, using the protocol described in example 1 above.

(42) The results are illustrated by FIG. 6.

(43) These results show that AP-Nt-Otx2 binds to cortical cells, comprising those of the visual cortex; on the other hand, no binding is observed with AP, AP-Hd-Otx2, or AP-HdAA-Otx2. In addition, whole Otx2 and also RK-Otx2, but not AA-Otx2, block the binding of AP-Nt-Otx2 to its target cells.

(44) Glycosaminoglycans (GAGs), and in particular chondroitin sulfate proteoglycans, are an essential constituent of the extracellular matrix (perineuronal net) which surrounds the parvalbumin neurons of the visual cortex. The putting into place of this perineuronal net coincides with the end of the critical period of plasticity; it constitutes a major factor in the loss of cortical plasticity that occurs at the end of this critical period, and it has been shown that the destruction of this perineuronal net by treating with chondroitinase-ABC makes it possible to restore this plasticity (Pizzorusso et al., Science, 298, 1248-51, 2002).

(45) In order to determine whether the perineuronal net associated with parvalbumin neurons was involved in the binding between AP-Nt-Otx2 and its target cells, the binding of AP-Otx2 was tested on cryostat sections of adult mouse brains fixed with methanol, and then incubated for 24 hours in the presence of chondroitinase ABC (2 U/ml) in a buffer containing 50 mM Tris [pH 8.0], 40 mM of sodium acetate, 0.1% BSA and protease inhibitors. In parallel, the sections, untreated or treated with chondroitinase ABC, were incubated with 0.01 mg/ml of Wisteria floribunda agglutinin lectin (WFA; Sigma-Aldrich), which binds to the GAGs of the perineuronal net, and which is labeled with FITC.

(46) The results are illustrated by FIG. 7.

(47) These results show that the treatment with chondroitinase ABC, which destroys the GAGs of the perineuronal net, also abolishes the binding of AP-Nt-Otx2.

(48) It therefore appears that it is the GAGs of the perineuronal net associated with the parvalbumin neurons which bear the AP-Nt-Otx2 binding site.

EXAMPLE 4: IN VIVO BLOCKING OF THE ENDOGENOUS TRANSFER OF OTX2 BY THE RK-OTX2 PEPTIDE

(49) As shown above, the RK-Otx2 peptide can block the binding of Otx2 to its target cells, in vitro. It was tested whether this effect also occurred in vivo.

(50) For this purpose, the RK-Otx2 peptide (0.25 mg/ml), the AA-Otx2 peptide (0.25 mg/ml), or PBS buffer, combined with polysialic acid (0.25 mg/ml, in order to enable diffusion of the peptides and to avoid nonspecific binding thereof to neurons expressing polysialic acid at their surface), were slowly infused (1 l/h) for 7 days into the right visual cortex of adult mice, using osmotic minipumps (Alzet 1003D, Alza) connected to stereotaxically implanted cannulas (Hensch et al., Science, 282, 1504-8, 1998; Fagiolini & Hensch, Nature, 404, 183-6, 2000). At the end of the infusion, the mice are perfused with 4% PFA, and brain sections (25 m) are cut in order to study the localization of Otx2 and that of the RK-Otx2 peptide. Otx2 is visualized using a rat anti-Otx2 monoclonal antibody diluted to 1/200 (Sugiyama et al., Cell, 134, 508-20, 2008), followed by a donkey anti-rat antibody labeled with Alexa 488 (Molecular Probes), diluted to 1/2000. The RK-Otx2 peptide is visualized using streptavidin labeled with Cy5.

(51) The cells expressing Otx2 were counted over a surface area of 700350 m encompassing layers I/II and IV of the binocular zone of the visual cortex.

(52) The results are illustrated by FIG. 8.

(53) These results show that the infusion of the RK-Otx2 peptide into the visual cortex for 7 days significantly reduces the number of cells expressing Otx2. On the other hand, in the case of the AA-Otx2 peptide, only a small, insignificant reduction is observed in the number of cells expressing Otx2.

(54) It was verified, by Sytox green labeling, that the infusion had no effect in itself on the number of cells. In order to be sure that the decrease in the number of cells expressing Otx2 was not due to cell death, an infusion of this peptide was carried out according to the protocol described above, and the number of cells expressing Otx2 was measured 8 days after the end of the infusion. Under these conditions, only very small amounts of RK-Otx2 peptide are detected 8 days after the end of infusion, and the number of cells expressing Otx2 in the treated hemisphere is restored to the level of that of the control hemisphere.

EXAMPLE 5: RESTORATION OF CORTICAL PLASTICITY BY THE RK-OTX2 PEPTIDE

(55) The effects of the RK-Otx2 peptide on the plasticity of the ocular cortex were compared with those of chondroitinase-ABC, which is known to enable the restoration of this plasticity (Pizzorusso et al., Science, 298, 1248-51, 2002).

(56) The RK-Otx2 peptide or the AA-Otx2 peptide was infused in adult mice (therefore after closing of the critical period of plasticity), as described in example 4 above. Chondroitinase-ABC, or the injection buffer (deionized water+0.1% BSA), were injected (0.4 l for each injection) at 3 sites surrounding the visual cortex (AP lambda, LM 1.5 mm; AP lambda, 4.0 mm; AP+1.5 mm; LM 2.5 mm) at two different depths (300 and 500 m).

(57) After infusion of the peptides or injection of chondroitinase ABC, the mice are subjected to monocular deprivation for 4 days, and the responses to visual stimuli are then measured by single-unit extracellular electrophysiology. The electrophysiology recordings are carried out under nembutal/chlorprothixene anesthesia using standard techniques (Gordon & Stryker, J Neurosci, 16, 3274-86, 1996; Mataga et al., Neuron, 44, 1031-41, 2004). 5 to 7 single-unit recordings were carried out for each mouse, on both sides of the medial-lateral axis of the primary visual cortex, in order to cover the monocular zone and the binocular zone, and to avoid sampling biases. Cell dominance scores were assigned to the cell responses, using a 7-point classification system (Wiesel & Hubel, J Neurophysiol, 26, 978-93, 1963) (Gordon & Stryker, J Neurosci, 16, 3274-86, 1996). The ocular dominance in the binocular zone was calculated for each mouse according to a contralateral bias index (CBI), determined as follows:

(58) (CBI): [(n.sub.1n.sub.7)+(n.sub.2n.sub.6)+(n.sub.3n.sub.5)+N]/2N, where N=total number of cells and nx=number of cells corresponding to an ocular dominance score of x.

(59) This weighted mean of the bias in favor of one or other eye can range from 0, for complete ipsilateral dominance, to 1, for complete contralateral dominance.

(60) In addition, treated mouse brain frontal sections were prepared, as described in example 4 above, in order to determine the influence of the RK-Otx2 peptide on parvalbumin neurons. The sections were labeled either with WFA, as described in example 3, or using a mouse anti-parvalbumin monoclonal antibody ( 1/500, Sigma-Aldrich), which was visualized using a donkey anti-mouse antibody labeled with Cy3. The labeled cells were quantified as described in example 4.

(61) The results are illustrated by FIG. 9.

(62) These results show that the monocular deprivation induces an ocular dominance (decrease in the contralateral bias index from 0.7 to 0.57) in the adult mice treated with the RK-Otx2 peptide, as in those treated with the chondroitinase ABC (chABC), unlike the mice treated with the injection buffer or the AA-Otx2 peptide.

(63) In parallel, the infusion of the RK-Otx2 peptide (but not that of the AA-Otx2 peptide) decreases the expression of the sites for the binding of WFA, and also that for parvalbumin. The number of parvalbumin-positive cells decreases by 56.2%, and that of cells surrounded by sites for the binding of WFA decreases by 51.3%.

(64) It emerges from these results that the blocking of Otx2 transfer by the RK-Otx2 peptide causes the inhibition of parvalbumin expression, and also a destruction of the perineuronal net similar to that caused by chondroitinase ABC. This results in a return of the parvalbumin neurons to an immature state, similar to that normally observed during the critical period, and this immature state enables reopening of the critical period, and restoring of the plasticity that is associated therewith.