Targeting of specific neuronal types in the non-human primate brain by using a murine CD25-specific recombinant immunotoxin

Abstract

Immunotoxin (ITX)-mediated cell targeting enables selective elimination of neuronal types of interest from a complex neural network. In this technology, human interleukin-2 receptor α-subunit or CD25 (hCD25) is expressed in specific cell types in transgenic rodents, and then the animals are treated with a recombinant ITX composed of monoclonal antibody variable regions for hCD25 fused to a Pseudomonas exotoxin fragment (PE38), resulting in the ablation of hCD25-expressing cells. However, there is a critical issue on the cross-reactivity of the recombinant ITX for endogenous CD25 in non-human primates (NHPs), leading to off-target effects. Here we generated a mouse CD25 (mCD25)-specific recombinant ITX, termed anti-mCD25-PE38, based on variable regions of a rabbit monoclonal antibody that specifically reacts to mCD25, but not to hCD25. Anti-mCD25-PE38 showed high-affinity binding to mCD25 and cytotoxic activity toward mCD25-expressing cells. Injection of anti-mCD25-PE38 into the ventral midbrain of common marmosets, in which the mCD25 transgene was expressed in dopamine neurons by a lentiviral vector for retrograde gene transfer, induced a significant loss of midbrain dopamine neurons. Therefore, anti-mCD25-PE38 provides a useful strategy for selective targeting of neuronal types to study the behavioral and neurological functions of these neurons in the NHP brain.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-39662-6.

Introduction

Non-human primates (NHPs) share a genetically close relationship with humans^1,2 and similarity of the structural and functional organization of the central nervous system to that of humans^3–5. NHPs provide a useful model system to study the pathogenesis of neurological and neuropsychiatric diseases and to search for new drugs and more effective therapeutic treatments. They also have advantages for investigations of the neural mechanisms underlying higher-order functions that control a variety of behaviors. For instance, chronic treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) causes a specific loss of nigrostriatal dopamine neurons accompanied by Parkinson’s disease-like motor and cognitive deficits in macaque monkeys and common marmosets^6–10. Administration of several drugs or deep brain stimulation restores these NHP models from the parkinsonian symptoms^11–16. Prenatal exposure to valproic acid (VPA) alters cortical synaptogenesis, synaptic organization, and gene expression patterns at the perinatal stage in marmosets, exhibiting their similarity to autism spectrum disorder^17,18. These alterations are ascribable to abnormalities in predictive coding and hormonal regulation in the VPA models^19,20. Recent advances in gene manipulation technology of target neural pathways have contributed to understanding the possible mechanisms underlying their functions and dysfunctions. Synaptic silencing with transient expression of tetanus neurotoxin reveals a key role of spinal interneurons in the corticospinal pathways that govern hand dexterity or its recovery after spinal cord injury in macaque monkeys^21,22. Application of optogenetic and chemogenetic tools offers new insights into brain circuit mechanisms at the normal and pathological states of the monkeys^22–28.

Immunotoxin (ITX)-mediated cell targeting is a genetic technology for the selective elimination of neuronal types of interest from an entire neural network^29–31. In this technology, human interleukin-2 receptor α-subunit or CD25 (hCD25) is expressed in target cell types of transgenic animals, and then the animals are treated with a single chain recombinant ITX termed anti-Tac(Fv)-PE38, which is composed of monoclonal antibody variable regions for hCD25 and a truncated form of Pseudomonas exotoxin (PE38)^32,33. Anti-Tac(Fv)-PE38 inhibits de novo protein synthesis, resulting in apoptosis of target cells having hCD25 protein. The cell targeting has been used for investigating a variety of central and peripheral nervous system functions in rodents^34–39. Anti-Tac(Fv)-PE38 exhibits high affinity for hCD25 protein, but not for mouse CD25 (mCD25) protein. This binding specificity is advantageous for suppressing off-target effects on the cells with endogenous CD25 in transgenic rodent models. By contrast, when the technology is applied to study the NHP brain, there is a pivotal issue on the cross-reactivity of the recombinant ITX with endogenous CD25 in NHPs that may lead to off-target effects.

In the present study, we aimed to produce an mCD25-specific recombinant ITX, termed anti-mCD25-PE38, which possesses high affinity for mCD25 protein and does not cross-react with hCD25 protein, for the elimination of specific cell types in the common marmoset brain. Rabbit monoclonal antibodies that specifically recognize the mCD25 extracellular domain were screened and the antibody variable regions were conducted to produce anti-mCD25-PE38 via a bacterial expression system. The purified protein exhibited specific binding activity to mCD25 protein and specific cytotoxic activity toward mCD25-expressing cells. Intracranial injection of the ITX into the ventral midbrain resulted in selective elimination of target cell types expressing mCD25 in the brain of marmosets. ITX-mediated cell targeting with anti-mCD25-PE38 provides a powerful strategy for selective removal of neuronal types of interest in the NHP brain.

Results

Screening of rabbit monoclonal antibodies that specifically react to mCD25

We immunized a rabbit with purified mCD25 extracellular domain fused to a His-tag (mCD25/His-tag) and isolated lymphocytes producing monoclonal antibodies that specifically react to mCD25 by immunospot array assay on a chip (ISAAC)^40,41 (Fig. 1a). In the ISAAC, a microarray chip was blocked with hCD25 extracellular domain fused to a His-tag (hCD25/His-tag), and then labelled with biotinylated mCD25/His-tag, followed by addition of Cy3-conjugated streptavidin, resulting in visualization of immuno-positive spots under a fluorescence microscope. We collected 78 individual cells and extracted their mRNA samples, which were subjected to reverse transcription-PCR to amplify cDNAs encoding immunoglobulin G (IgG) heavy and light chain variable (V_H and V_L, respectively) regions. We amplified 50 pairs of V_H and V_L cDNAs and inserted them into expression vectors containing cDNA for the rabbit IgG constant region (ϒ or K chain). These vectors were co-transfected into mammalian cells, leading to the production of 19 recombinant monoclonal antibodies that showed a higher affinity for mCD25/His-tag than for hCD25/His-tag. The binding properties of the recombinant antibodies were evaluated by enzyme-linked immuno-sorbent assay (ELISA). The results of ELISA obtained from a representative rabbit monoclonal antibody (RMAb-52) are shown in Fig. 1b. A competitive ELISA showed a K_D value of 1.10 × 10⁻⁸ M for RAMAb-52 (Fig. 1c).

Fig. 1.

Production of mCD25-specific recombinant monoclonal antibodies. (a) Strategy for screening rabbit monoclonal antibodies that specifically react to mCD25 by ISAAC technology. A rabbit was immunized with purified mCD25/His-tag protein, and IgG⁺ cells were isolated from peripheral blood lymphocytes, arrayed on the microarray chip, and cultured to trap secreted IgG. The chip was blocked with hCD25/His-tag protein, and then labelled with biotinylated mCD25/His-tag protein, followed by addition of Cy3-conjugated streptavidin. The cells producing mCD25-specific antibodies were visualized under a fluorescence microscope and collected from individual wells. The mRNA samples were extracted from single cell populations and subjected to reverse transcription (RT), and cDNA fragments encoding V_H and V_L regions were amplified by PCR, which were then cloned into the expression vectors containing rabbit immunoglobulin Y and K chains. The recombinant antibodies were produced in cultured cells and secreted into the medium. (b) Binding property of an mCD25-specific recombinant antibody to target proteins. Microplates were coated with either mCD25/His-tag or hCD25/His-tag, and then various concentrations of RMAb-52 were added to the wells. The binding of antibody was detected using a secondary antibody conjugated to alkaline phosphatase. (c) Competitive binding of antigen for the mCD25-specific recombinant antibody. RMAb-52 solution was incubated with various concentrations of mCD25/His-tag. Microplates were coated with mCD25/His-tag, and then the incubated mixture at equilibrium was added to the wells. The binding of free antibody was detected using a secondary antibody conjugated to alkaline phosphatase. The K_D value was determined by using Scatchard plots. Data are expressed as mean ± SEM of four independent experiments. Individual data are overlaid in (c).

Production and characterization of an mCD25-specific recombinant ITX

We used cDNA sequences encoding V_H and V_L regions of RMAb-52 to produce an mCD25-specific recombinant ITX (anti-mCD25-PE38), in which a single chain composed of V_H and V_L regions was fused to PE38 with KDEL sequence³³. We constructed a bacterial expression plasmid, pAnti-mCD25-PE38, to express anti-mCD25-PE38 under the control of T7 promoter (see Fig. 2a for the plasmid structure and Supplementary Fig. 1 for the nucleotide and deduced amino acid sequences). The recombinant proteins were expressed in E. coli through T7 expression system and accumulated in the inclusion body (Fig. 2b). The proteins were denatured, renatured, and applied for a line of ion-exchange chromatography with a Capto HiRes Q column and gel filtration liquid chromatography with a TSK G3000SW column (see Fig. 2c and e for elution profiles). Peak fractions were analyzed by SDS-polyacrylamid gel electrophoresis (SDS-PAGE) (Fig. 2d and f). The purified protein migrated as a single band with molecular weight of ~ 62 kDa on the gels (Fig. 2f).

Fig. 2.

Expression and purification of anti-mCD25-PE38 recombinant ITX. (a) Structure of the bacterial expression plasmid pAnti-mCD25-PE38. The plasmid contains a T7 promoter (PT7) and a gene cassette encoding a single chain composed of V_H and V_L for RMAb-52 fused to PE38 with a KDEL sequence. Amp^r: ampicillin-resistance gene. (b) Induction of protein expression by IPTG treatment. Proteins (~ 5 µg) from cell lysates (0, 90, and 120 min after treatment) and the inclusion body (IB) were separated on a 10% SDS-polyacrylamide gels. (c) Elution profile through ion-exchange liquid chromatography using a Capto HiRes Q column. (d) SDS-PAGE of proteins in fractions from the ion-exchange chromatography. Lane 1, fraction #40; and lane 2, fraction #41. (e) Elution profile through gel filtration liquid chromatography using a TSK G3000SW column. (f) SDS-PAGE of proteins in a fraction from the gel filtration chromatography. Lane 1, fraction #16. Molecular weight markers are indicated in lane M. Arrowheads show anti-mCD25-PE38 protein with molecular weight of ~ 62 kDa.

The purified protein was used for ELISA to evaluate the binding property to target proteins. The ELISA indicated a higher affinity of the protein for mCD25/His-tag compared to hCD25/His-tag (Fig. 3a). The affinity of the recombinant ITX was assessed by the competitive ELISA, and a K_D value was 4.64 × 10⁻⁸ M (Fig. 3b). A cell viability assay also showed selective cytotoxic activity of ITX toward mCD25-expressing cells with IC₅₀ of approximately 2.0 ng/mL, but not for hCD25-expressing cells (Fig. 3c).

Fig. 3.

Properties of anti-mCD25-PE38 ITX. (a) Binding property of the recombinant ITX for target proteins. Microplates were coated with either mCD25/His-tag or hCD25/His-tag, and then various concentrations of ITX protein were added to the wells. The binding of ITX was detected using a PE38-specific monoclonal antibody and a mouse IgG-specific secondary antibody conjugated to horseradish peroxidase. (b) Competitive binding of antigen for the recombinant ITX. ITX solution was incubated with various concentrations of mCD25/His-tag protein. Microplates were coated with mCD25/His-tag, and then the incubated mixture at equilibrium was added to the wells. The binding of free ITX was detected using a PE38-specific monoclonal antibody and a mouse IgG-specific secondary antibody conjugated to horseradish peroxidase. The K_D value was determined using Scatchard plots. (c) Cytotoxic activity of the recombinant ITX toward HT-2 and EL-4 cells expressing mCD25 and hCD25, respectively. Cells were incubated with various concentrations of the ITX. Viable cell number was evaluated by a cell viability assay using WST reagent. Relative ratios to the average of control cell number without the ITX were calculated. Data are expressed as mean ± SEM of four independent experiments. Individual data are overlaid in (b).

Selective targeting of neurons expressing mCD25 in the marmoset brain

To test the in vivo cytotoxic activity of anti-mCD25-PE38 in the NHP brain, we performed selective neural pathway targeting^42–44 using a lentiviral vector for neuron-specific retrograde gene transfer (NeuRet)^45–47. In this technique, the NeuRet vector encoding the mCD25 transgene is injected into the putamen (Pu) of common marmosets, resulting in transgene expression in striatal input pathways through retrograde axonal transport (Fig. 4a, left panel). Then, anti-mCD25-PE38 is injected into the substantia nigra pars compacta (SNc) in the ventral midbrain, leading to the elimination of nigrostriatal dopamine neurons (Fig. 4a, right panel). The NeuRet vector carrying a gene cassette for mCD25 fused to a V5-tag (1.90 × 10¹³ genome copies/mL) was bilaterally injected into the marmoset Pu (1.0 µL/site, 6 sites/hemisphere) (see Fig. 4b for the coronal planes 1 and 2 on a sagittal MR image; and Fig. 4c for the injection sites on two coronal images, which were obtained from a representative injected animal). Three weeks later, a solution containing anti-mCD25-PE38 (20 ng/µL) or phosphate-buffered saline (PBS) as a control was unilaterally injected into the SNc (0.5 µL/site, 6 sites/hemisphere) (see Fig. 4b for coronal planes 3 and 4 on the sagittal MR image; and Fig. 4c for the injection sites on the two coronal images from the same animal). The coordinates used for viral vector injections into the Pu and ITX/PBS injections into the SNc of four animals (termed MAR-1 to MAR-4) are summarized in Supplementary Tables 1 and 2. Our preliminary data obtained from intracranial injections with various doses of anti-mCD25-PE38 into the marmoset brain indicated that there was not any non-specific tissue damage with the doses less than 40 ng/µL. We thus selected 20 ng/µL of anti-mCD25-PE38 for the intracranial injection. Two weeks after the ITX/PBS injections, the animals were sacrificed, followed by immunohistochemical analysis. Coronal sections through the ventral midbrain were prepared and stained with the antibody against tyrosine hydroxylase (TH), a marker of dopamine neurons, or V5-tag. Some V5-positive signals were observed in SNc neurons containing TH-positive signals on the PBS-injected side (see Supplementary Fig. 2 for a typical microscopic image of double immunostaining). As compared to the PBS-injected side, the number of TH-positive cells in the corresponding SNc region was considerably reduced on the ITX-injected side (Fig. 4d, upper and middle images). Cresyl violet staining of the SNc sections excluded non-specific injury of the tissues after ITX treatment (Fig. 4d, lower images). Cell counts in the sections through the ventral midbrain indeed showed a significant reduction in the number of TH-positive cells to 66.7% on the ITX-injected side (155.83 ± 2.27/section), in comparison with the control side (233.47 ± 13.35/section; unpaired t-test, t₅ = 6.758, p = 0.001) (Fig. 4e). In addition, there was no apparent damage in the distribution of V5-positive cells in some cerebral cortical areas or the intralaminar thalamic nuclei on the ITX-injected side (Supplementary Fig. 3). The overall results indicate that anti-mCD25-PE38 treatment selectively and efficiently removes neurons expressing the mCD25 transgene from the marmoset brain.

Fig. 4.

In vivo cytotoxic activity of anti-mCD25-PE38 ITX against mCD25-expressing cells in the common marmoset brain. (a) Strategy for selective targeting of the nigrostriatal dopaminergic pathway of common marmosets. The marmosets received bilateral injections of the NeuRet vector encoding mCD25/V5-tag transgene into the Pu, and the recombinant ITX was unilaterally injected into the SNc, resulting in the elimination of nigrostriatal dopamine neurons. (b, c) Representative MR images obtained from marmoset MAR-3. Sagittal MR image showing the coronal planes 1 to 4 (b). Coronal images showing injection sites corresponding to planes 1 and 2 (vector injections into the Pu) and planes 3 and 4 (ITX/PBS injections into the SNc) (c). Injections sites are indicated by yellow spots. (d) Typical microscopic images of TH immunostaining of sections through the ventral midbrain of the injected marmosets. Middle images are magnified views of the rectangles in the upper image. Sections stained with cresyl violet are shown in the lower images. Arrows in the middle images represent the locations of cell loss. VTA, ventral tegmental area. (e) Schematic illustrations of the sections used for cell counts. The SNc regions are indicated by yellow. Anteroposterior coordinates (mm) from the interaural line are shown. (f) Cell counts of TH-positive neurons in the SNc. Data are presented as mean ± SEM and individual data are overlaid (n = 4 for each of the PBS- and ITX-injected sides). *p < 0.05 (unpaired t-test). Scale bars: 5 mm (b, c), 2 mm (d), 500 μm (e).

Discussion

In the present study, we screened rabbit monoclonal antibodies that react to the mCD25 protein, but do not cross-react to the hCD25 protein by using the ISAAC procedure. We then employed the sequences for V_H and V_L regions of the monoclonal antibody to produce anti-mCD25-PE38 in a bacterial expression system. The recombinant ITX was characterized by the in vitro binding to CD25 protein and cytotoxic activity against CD25-expressing cells. The ITX was further tested for the in vivo cytotoxic activity toward nigrostriatal dopamine neurons in the ventral midbrain of common marmosets, in which the mCD25 transgene was expressed via a lentiviral vector allowing retrograde gene transfer after its injection into the striatum.

Anti-mCD25-PE38 exhibited high-affinity binding for mCD25 protein, but not to hCD25 protein, together with the cytotoxic activity toward mCD25-expressing cells, but not toward hCD25-expressing cells. However, we did not test the binding activity to the CD25 protein derived from monkeys and marmosets or the cytotoxic activity toward their CD25-expressing cells. Comparison of deduced amino acid sequences among species indicates the homology of 60% in average between the hCD25 and the mCD25 sequence including some conserved regions localized in the transmembrane and cytoplasmic domains^48,49. The CD25 sequences of macaque monkeys and common marmosets share the homology of 92% and 81% in average with the hCD25 sequence, respectively. This confirms strong similarity of the extracellular domain in addition to the conserved regions in the transmembrane and cytoplasmic domains. By contrast, the macaque and marmoset CD25 sequences possess the homology of only 59% and 57% in average with the mCD25 sequence, respectively^50,51. The lower homology of such NHP sequences with the murine orthologue suggests no or low cross-reactivity of anti-mCD25-PE38 to the macaque or marmoset CD25 protein. In addition, intracranial injection of the recombinant ITX into the viral vector-injected marmoset brain did not cause any non-specific damage to the tissues around the injection sites, except for the dopamine neurons expressing the mCD25 protein.

A previous study has reported selective removal of a particular neuronal pathway in the NHP brain by using ITX-mediated tract targeting⁵². A lentiviral vector encoding the hCD25 transgene fused to GFP was injected into the subthalamic nucleus of macaque monkeys to express the transgene in the cortico-subthalamic pathway originating from motor-related areas of the frontal lobe. Injection of anti-Tac(Fv)-PE38 into the supplementary motor area resulted in a reduction in cortically-evoked response activity in the basal ganglia output nuclei⁵², thereby validating the role of the cortico-subthalamic pathway in the relay of motor information processing through the cortico-basal ganglia circuitry⁵³. In the present study, appropriate doses of the recombinant ITX did not exert any side effects on the tissues around the injection sites, suggesting low-level expression of CD25 in the supplementary motor area of the monkeys. However, prominent expression of endogenous CD25 is reported in some brain regions of rodents, including the olfactory bulb and hippocampus^54,55, although we did not examine the expression pattern of endogenous CD25 in the NHP brain. The CD25 immunoreactivities are also upregulated in some brain regions in central nervous system diseases, such as Alzheimer’s disease and multiple sclerosis^56–58. The use of mCD25-specific ITX is particularly advantageous for studying NHP brain circuits in which endogenous CD25 expression is observed or under certain pathological conditions.

Immunohistochemical analysis with sections through the SNc of the injected marmosets indicated a significant decrease in TH⁺ cell number in the ITX-injected side as compared to the PBS-injected side, although the decrease was 66.7% of the control. A partial loss of dopamine neurons by the ITX injection can be explained by some experimental conditions. In this study, we performed the injection of the lentiviral vector (1.90 × 10¹³ genome copies/mL) into 6 sites in the posterior Pu per hemisphere of the marmoset brain. The diffusion extent of the viral vector injected restricts the region of vector entry, from which the particles are transported to the cell bodies, resulting in gene expression in transduced neurons. Increasing the number of the vector injection sites within a wider area of the striatum enables to increase the frequency of SNc neurons expressing the transgene. In addition, the vector titer is an important factor to determine the transduction efficiency and gene expression level in a single cell. The titer is also affected by the length and sequence of the transgene used for vector production. Increasing the titer of the vector encoding the mCD25 transgene leads to the enhanced frequency of mCD25-expressing neurons and its intracellular expression level.

ITX-mediated cell targeting combined with the transgene expression system using lentiviral vectors for retrograde gene transfer enables us to achieve selective targeting of given neural pathways in the brains of both rodents and NHPs^42–44,52. In the present study, we developed anti-mCD25-PE38 to suppress off-target effects of ITX on the NHP brain tissues. Although we focused on histologically confirming the ablation of target neuronal cells in the marmoset brain, the application of this technique to various behavioral and physiological studies on the NHPs may yield a great advantage in investigating their neural network functions. The technology will provide a general and powerful strategy to explore not only the specific roles of individual pathways or neuronal types constituting a particular neural network, but also the central mechanisms for comprehensively integrating diverse networks.

Methods

Animals

All animal experiments were carried out in accordance with the guidelines established by Fukushima Medical University and the University of Toyama, and the Laboratory Animal Welfare established by CLEA Japan, Inc. (Tokyo, Japan) and by the Central Research Center for Experimental Animals (Kawasaki, Japan). All care and handling procedures were approved by their Institutional Animal Care and Use Committees. A rabbit (New Zealand White, aged 12 weeks, Sankyo Lab) was used for immunization and cell preparation. Four common marmosets (Callithrix jacchus, males, aged 40 months, 305–344 g in body weight; CLEA Japan, Inc.) were named MAR-1 to MAR-4 and used for the brain imaging, intracranial surgery, and histological analysis. The marmosets were housed individually in home cages measuring 39 × 55 × 70 cm (W × D × H) in size, and maintained at 28 ± 1 °C and 50 ± 10% humidity in a 12-h light/12-h dark cycle (7:00 to 19:00 for light).

Immunization and lymphocyte preparation

A rabbit was immunized subcutaneously with 500 µg of purified mCD25/His-tag (Sino Biological Inc., Cat#50292-M08H) in Freund’s complete adjuvant (Sigma-Aldrich, Cat#AR001). Two, 4, and 6 weeks after the primary immunization, the rabbit was boosted subcutaneously with 500 µg of the same material as used for the immunization in Freund’s incomplete adjuvant (Sigma-Aldrich, Cat#AR002). One week after the final boost, peripheral blood lymphocytes and spleen were collected, and rabbit IgG⁺ cells were isolated using rabbit IgG-specific antibody-conjugated microbeads with an autoMACS Pro Separator (Miltenyi Biotec).

ISAAC and production of recombinant monoclonal antibodies

The ISAAC method was conducted as previously described^40,41 with some modifications. The surface of the microarray chip was coated with 10 µg/mL rabbit IgG-specific antibody (MP Biomedicals, Cat#0855641) overnight at 4 ℃. After removing the antibody, we blocked the chip with 0.01% Biolipidure (NOF Corporation) for 15 min at room temperature and subsequently washed it with culture medium. The rabbit IgG⁺ cells were arrayed on the chip and cultured for 3 h to trap secreted IgG. The chip was incubated with 10 µg/mL hCD25/His-tag (Sino Biological Inc., Cat#10165-H08H) for 30 min as a blocking step, and then with 10 µg/mL biotinylated mCD25/His-tag for 30 min, followed by addition of Cy3-conjugated streptavidin (Sigma-Aldrich, Cat#S6402) for 30 min. Cells were stained with 1 mM Oregon Green (Molecular Probes, Cat#C34555) for 5 min, and the cells secreting mCD25-specific antibodies were visualized under a fluorescence microscope (BX51W1, Olympus) and collected from individual wells using a micromanipulator (TransferMan NK2, Eppendorf). Extraction of mRNA, amplification and cloning of cDNA fragments for IgG V_H and V_L regions by reverse transcription-PCR, and expression of recombinant IgG in Expi293F cells (Thermo Fisher Scientific) were performed as previously described^40,41. Plasmid DNAs encoding variable regions of IgG were prepared using an automated DNA extraction machine (GENE PREP STAR PI-480, Kurabo Industries), and their sequences were determined using a capillary electrophoresis system (3500 Genetic Analyzer, Thermo Fisher Scientific).

Expression and purification of a recombinant ITX

ITX protein was prepared as previously described⁵⁹ with some modifications. The V_H and V_L cDNA sequences of the RMAb-52 monoclonal antibody were exchanged by the corresponding parts of pRTK749K, which contained a T7 promoter and a gene cassette for anti-Tac(Fv)-PE38 encoding a single chain of variable regions for anti-Tac monoclonal antibody against hCD25, fused to PE38, containing translocation and catalytic domains of Pseudomonas exotoxin and a KDEL sequence³³, resulting in a bacterial expression plasmid pAnti-mCD25-PE38 (see Fig. 2a for the plasmid structure and Supplementary Fig. 1 for the nucleotide and deduced amino acid sequences). E. coli BL21(λDE3) cells were transformed with the plasmid and cultured at 37 ℃ in Superbroth (MP Biomedicals, Cat#3010-032) supplemented with 100 µg/mL ampicillin. At an OD₆₀₀ of ~ 0.8, isopropyl-1-thio-β-D-galactopyranoside (IPTG) was added to a final concentration of 1 mM, and culture was further incubated for 120 min. Cells were harvested and sonicated, and cell lysates were treated with repeated cycles of sonication and centrifugation to prepare the inclusion body. Proteins contained in the inclusion body were denatured in 6 M guanidine HCl in 0.1 M Tris-HCl buffer (pH8.0) containing 2 mM ethylenediaminetetraacetic acid (EDTA), renatured in 0.1 M Tris-HCl buffer (pH8.0) containing 0.5 M L-arginine, 2 mM EDTA, and 0.9 mM glutathione oxide, and dialyzed against 20 mM Tris-HCl buffer (pH8.0) containing 100 mM urea. The renatured proteins were loaded onto ion-exchange liquid chromatography using a HiTrap QFF column (Cytiva) in 20 mM Tris-HCl buffer (pH7.4) containing 1 mM EDTA and eluted with a step gradient from 0.1 to 0.5 M NaCl in the same buffer. Elutes were separated into 4-mL fractions at a flow rate of 4 mL/min. Peak fractions were pooled and diluted with the buffer, and the diluted solution was loaded onto ion-exchange high-performance liquid chromatography using a Capto HiRes Q column (Cytiva) in 20 mM Tris-HCl buffer (pH7.4) containing 1 mM EDTA and eluted through a linear gradient 0 to 0.5 M NaCl in the same buffer. The elutes were separated into 1.5-mL fractions at a flow rate of 1.5 mL/min. Peak fractions were pooled and concentrated using Ultracentrifuge filters (Merck, Cat#UFC503024), and applied for gel filtration high-performance liquid chromatography using a TSK G3000SW column (Tosoh Co.) in PBS. Elutes were divided into 1-mL fractions at a flow rate of 0.5 mL/min. The purified protein was aliquoted and stored at −80 ℃; they were analyzed by 10% SDS-PAGE and stained with Coomassie brilliant blue. Protein concentrations were measured using a Bradford protein assay kit (Bio-Rad Laboratories, Cat#5000001) with bovine serum albumin (BSA) as the standard.

ELISA

For the analysis of binding property, 96-well microplates were coated with 50 µL per well of 2 µg/mL of mCD25/His-tag or hCD25/His-tag protein in PBS, and then blocked with 3% BSA in PBS. After washing, various concentrations of monoclonal antibodies or recombinant ITX were added to the wells and incubated for 1 h at room temperature. For the binding of monoclonal antibodies, a rabbit IgG-specific secondary antibody conjugated to alkaline phosphatase (donkey, 1:2000; Sigma-Aldrich, Cat#A3687) was added to the wells, and the immunoreactive signals were visualized with p-nitrophenylphosphate. For the binding of ITX, a PE38-specific monoclonal antibody (mouse, 1:120) and a mouse IgG-specific secondary antibody conjugated to horse radish peroxidase (donkey, 1:2000; Jackson ImmunoResearch Laboratories, Cat#715-035-151) were added to the wells, and the signals were stained with ELISA substrate solutions (Thermo Fisher Scientific, Cat#34024). Optical absorbance at 405/450 nm was measured using a microplate reader (Varioskan LUX, Thermo Fisher Scientific).

The affinity of the recombinant monoclonal antibody or ITX was determined by competitive ELISA^40,41. A constant concentration of recombinant proteins (2 nM) was incubated with various concentrations of mCD25/His-tag protein overnight at 4 ℃ until equilibrium was reached. To measure the concentrations of free antibody or ITX protein that remained unsaturated with antigen, 96-well microplates were coated with 50 µL per well of 2 µg/mL of mCD25/His-tag in PBS, and then blocked with 3% BSA in PBS. After washing, the incubation mixture at equilibrium was added to the wells and incubated for 1 h at room temperature. Free antibody or ITX protein was detected using the same methods as used for the binding analysis. The data were used for determination of K_D value by using Scatchard plots.

Cell viability assay

The specific cytotoxic activity of the recombinant ITX was evaluated using a cell viability assay, as previously described⁵⁹. HT-2 and EL-4 cells (ATCC, CRL-1841 and TIB-39, respectively), which express mCD25 and hCD25, respectively, were seeded in 96-well microplates overnight and treated with various concentrations of recombinant ITX for 72 h at 37 ℃. Twenty µl of WST-8 reagent (Dojindo Molecular Technologies, Cat#W209) was added to the wells and incubated for 1 to 3 h at 37 ℃. Optical absorbance at 450 nm was measured using a microplate reader.

Viral vector preparation

NeuRet vectors were prepared and purified as previously described^45,46. The transfer plasmid contained the cDNA encoding mCD25 fused to a V5-tag (GKPIPNPLLGLDST) downstream of the murine stem cell virus promoter (pCL20c-MSCV-mCD25/V5-tag). HEK293T cells (ATCC, CRL-3519) were transfected with the transfer, envelope containing fusion glycoprotein type E (FuG-E)⁴⁶ cDNA, and packaging plasmids using the calcium phosphate precipitation method. After viral collection and purification, vector titer (genome copies per mL) was determined using quantitative PCR with the StepOne Real-Time PCR System (Applied Biosystems) and the Lenti-X qRT-PCR Titration Kit (Takara Bio Inc., Cat#631235).

Brain imaging and intracranial surgery

MR imaging and X-ray computed tomography (CT) were conducted as previously described⁶⁰, with some modifications. Common marmosets were intramuscularly administered 0.1 mg/kg atropine sulfate and 12 mg/kg alfaxalone, and the anesthetized state was maintained using 1 to 2% isoflurane inhalation. The marmosets were placed in an acrylic imaging cradle (Takashima Seisakusho Ltd.), and an acrylic head holder was attached to the head by inserting ear bars into the external auditory canals. Pulse oxygen and skin/rectal temperature were monitored regularly during the imaging.

MR image data were obtained using a 7.0 T Biospec 70/16 MR Image Scanner System (Bruker BioSpin GmbH) equipped with actively shielded gradients at a maximum strength of 700 mT/m and an imaging coil (inner diameter 60 mm; Bruker Biospin GmbH). T2-weighted images were acquired using a rapid acquisition with relaxation enhancement (RARE) sequence with the following parameters: repetition time 4,500 ms; echo time 20 ms; field of view (FOV) 48 × 48 mm; matrix 240 × 240; slice thickness 0.35 mm; RARE factor 4; number of averages 11; and scan time 42 min.

CT image data were then obtained at the same scanning position as the MR imaging using a cone-beam CT system (Cosmo Scan Fx, Rigaku Corp.), which was operated under the following conditions: tube voltage 90 kV; tube current 88 µA; exposure time 2 min; FOV 61.44 × 61.44 × 61.44 mm; and voxel size 120 µm³. The skull was semi-automatically extracted from the CT data using the “Segmentation Editor” in Amira software version 7.0 (Visage Imaging, Inc.).

The MR and CT images were manually overlaid based on the position of the ear bars as stereotaxic landmarks. CT images were resliced to resample isotropic MR images (0.2 mm³). MR and CT images used for injection coordinates were viewed with PMOD image analysis software version 3.7 (PMOD Technology Ltd.).

The marmosets were placed in a stereotaxic instrument (SR-6 C-HT, Narishige) under anesthesia as described above. They were administered viccillin (2.5–10 mg/kg, intramuscular injection) as an antibiotic drug and meloxicam (0.5 mg/kg, subcutaneously) as an anti-inflammatory drug, together with Ringer’s solution (5 mL, subcutaneously) containing DL-methionine (4.5 mg), thiamine chloride hydrochloride (300 µg), sodium riboflavin phosphate (60 µg), pyridoxine hydrochloride (150 µg), nicotinic acid amide (750 µg), and sodium L-ascorbate (3.0 mg). After hair removal and sterilization with povidone iodine, burr holes were opened on the skull. A glass microinjection capillary was inserted into the brain, and moved slowly to the target positions: into the posterior part of the Pu (1.0 µL/site, 6 sites/hemisphere) for the lentiviral vector (1.90 × 10¹³ genome copies/mL); and into the SNc for solution of anti-mCD25-PE38 protein (20 ng/µL in PBS containing 0.1 mg/mL monkey serum albumin) or PBS (0.5 µL/site, 6 sites/hemisphere) (see Supplementary Tables 1 and 2 for the coordinates of the injections into the Pu and SNc, respectively). Injections were performed at a constant flow rate of 0.2 µL/min for viral vector and 0.1 µL/min for ITX/PBS solution using a microinfusion pump (ESP-32, Eicom).

Histology

Perfusion fixation was performed under anesthesia with a mixture of medetomidine (0.04 mg/kg), midazolam (0.4 mg/kg), and butorphanol (0.4 mg/kg) for induction and isoflurane (4%) for terminal care. The marmosets were perfused transcardially with PBS, followed by fixation with 4% paraformaldehyde in 0.1 M PBS (pH 7.4). Brains were removed from the skull, postfixed for 1 to 2 days, and saturated with 15% and 30% sucrose in PBS at 4 °C. Fixed brains were cut into Sect. (30 μm thick) through a coronal plane using a cryostat.

For immunohistochemistry, the sections were incubated overnight at 4 °C with either an anti-TH monoclonal antibody (mouse, 1:400; Chemicon, Cat#MAB5280) or an anti-V5 monoclonal antibody (mouse, 1:100; Thermo Fisher Scientific, Cat#R960-25). The sections were then incubated with an anti-mouse IgG secondary antibody conjugated either to biotin (goat, 1:500; Jackson ImmunoResearch Laboratories, Cat#715-065-151) or to Cy3 (goat, 1:500; Jackson ImmunoResearch Laboratories, Cat#715-165-150) for 2 h at room temperature. Biotinylated signals were visualized with an avidin-biotin-peroxidase complex kit (Vector Laboratories, Cat#PK-6101). The sections were also stained with 0.1% cresyl violet solution (Muto Pure Chemicals, Cat#41021). These sections were mounted on gelatin-coated glass slides and coverslipped. Images were acquired using a fluorescent microscope (BZ-X810, Keyence).

Cell counts

The number of immunopositive cells was counted in each of 5 sections through the SNc between the anteroposterior coordinates 4.3 and 5.8 mm from the interaural line, and the values for individual animals were averaged.

Statistical analysis

All values were expressed as mean ± standard errors of the mean (SEM) of the data. For statistical comparison, the unpaired t-test was used with significance set at p < 0.05.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

T. K., M.T., and K.K. conceived the study, designed the experiments, and directed the project. T.O., K.I., and K.K. designed the antibody screening, and T.O. performed ISAAC and recombinant antibody production. M.O., I.P., and T.O. generated the recombinant ITX. S.K., S.K., M.S., and K.I. performed the intracranial injections and histological/biochemical examinations. T.O., M.T., and K.K. wrote the manuscript. All authors discussed the results and their implications and commented on the manuscript at all stages.

Funding

This work was supported by grants-in-aid for Brain/MINDS 2.0 under Grant Number 25wm0625103h0002 (K.K.) and for Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research: BINDS) under Grant Number JP21am0101077 (T.O.) from the Japan Agency for Medical Research and Development; and a grant-in-aid for Scientific Research on Transformative Research Areas (A) Adaptive Circuit Census (21H05244) (K.K.) from the Ministry of Education, Science, Sports, and Culture of Japan.

Data availability

The datasets generated and/or analyzed during the current study are available in the Mendeley Data [https://data.mendeley.com/datasets/9zz9x8b3rs/2].

Declarations

Competing interests

The authors declare no competing interests.