TNPO2 variants associate with human developmental delays, neurologic deficits, and dysmorphic features and alter TNPO2 activity in Drosophila

Abstract

Transportin-2 (TNPO2) mediates multiple pathways including non-classical nucleocytoplasmic shuttling of >60 cargoes, such as developmental and neuronal proteins. We identified 15 individuals carrying de novo coding variants in TNPO2 who presented with global developmental delay (GDD), dysmorphic features, ophthalmologic abnormalities, and neurological features. To assess the nature of these variants, functional studies were performed in Drosophila. We found that fly dTnpo (orthologous to TNPO2) is expressed in a subset of neurons. dTnpo is critical for neuronal maintenance and function as downregulating dTnpo in mature neurons using RNAi disrupts neuronal activity and survival. Altering the activity and expression of dTnpo using mutant alleles or RNAi causes developmental defects, including eye and wing deformities and lethality. These effects are dosage dependent as more severe phenotypes are associated with stronger dTnpo loss. Interestingly, similar phenotypes are observed with dTnpo upregulation and ectopic expression of TNPO2, showing that loss and gain of Transportin activity causes developmental defects. Further, proband-associated variants can cause more or less severe developmental abnormalities compared to wild-type TNPO2 when ectopically expressed. The impact of the variants tested seems to correlate with their position within the protein. Specifically, those that fall within the RAN binding domain cause more severe toxicity and those in the acidic loop are less toxic. Variants within the cargo binding domain show tissue-dependent effects. In summary, dTnpo is an essential gene in flies during development and in neurons. Further, proband-associated de novo variants within TNPO2 disrupt the function of the encoded protein. Hence, TNPO2 variants are causative for neurodevelopmental abnormalities.

Keywords: Drosophila; Importin-3; Karyopherin-β2b; TNPO1; TNPO2; Transportin; global developmental delays; intellectual disability; nucleocytoplasmic shuttling; rare disease.

Figure 1

TNPO2 variants are associated with varied dysmorphic features in individuals (A) Proband 4 at age 3 years with short philtrum, broad nasal bridge, large fleshy ears, and coarse facial features. (B) Proband 8 at age 8 years with strabismus, high nasal bridge, eversion of the lower lip, and clinodactyly. (C) Proband 11 at age 9 years has no clear dysmorphism. (D) Proband 13 at age 11 years with deep set eyes and large cupped ears.

Figure 2

Fly Transportin is essential for proper animal development and dTnpo loss in eyes and wings causes dysmorphisms (A) Protein sequence comparison of human TNPO2 (hTNPO2) and Drosophila Tnpo (dTnpo) shown as a diagram and a detailed amino acid alignment. All variants are at conserved amino acids (red) except p.Lys118Asn (orange). Symbols in the protein alignment: identical (|), similar (:), different (.), absent (_). (B) dTnpo mutants (red) created for loss-of-function (LoF) studies include dTnpo^Δ11 (an imprecise excision of the P-element, NP4408), dTnpo^Gly736Asp (an EMS-induced mutation), and a CRIMIC allele. Two independent RNAi lines, RNAi-1 and RNAi-2, were also obtained. (C) Animals homozygous for dTnpo mutant alleles demonstrate larval lethality due to dTnpo loss. None of the alleles or a large deficiency allele which lacks dTnpo, Df(3L)Exel8101, complement each other. Lethality caused by dTnpo^Δ11 and dTnpo^Gly736Asp can be rescued using a genomic rescue construct, GR^dTnpo. (D) The FRT/FLP system was used to make mosaic tissue in the fly eye during development. dTnpo^Gly736Asp causes a rough eye phenotype. No homozygous dTnpo^Δ11 mutant tissue is observed, indicating cell lethality. Scale bar = 100 μm. (E) The FRT/FLP system was used to make mosaic tissue in the developing wing. dTnpo^Gly736Asp causes notch and blister phenotypes. Scale bar = 200 μm. In (D) and (E), “Control” is yw;; FRT80B. Full fly genotypes for this and following figures are in Data S2. dTnpo-targeting RNAi produce consistent phenotypes (see Figure S1).

Figure 3

dTnpo is highly expressed in neurons, including mushroom body neurons The dTnpo CRIMIC (T2A-GAL4) allele was used to drive expression of UAS-fluorescent reporter transgenes. (A–L) UAS-mCherry.NLS (nuclear mCherry) was expressed and tissue were dissected from L3 larvae (CNS, includes central brain and VNC) or adults (brain). Shown is half of the adult brain. Tissue were counterstained with markers for neurons (Elav) or glia (Repo). Z stacked images showing dTnpo expression pattern compared to neurons (A–F) or glia (G–L). Dashed squares indicate regions used in (A′)–(L′). (A′–L′) Single slice images were used to better visualize cellular co-localization of mCherry.NLS signal with neurons or glia. White arrows highlight co-localized nuclei with most neurons and some glia. (M–R) dTnpo CRIMIC driven expression of UAS-mCD8::RFP (membrane-bound RFP) and FasII counter-staining confirmed overlap of dTnpo expression and mushroom body (MB) neurons in both larval and adult brains. (S and T) Schematics of the larval CNS (S) and adult brain (T) highlighting MB neurons (blue), the ventral nerve cord (VNC), the central brain, and optic lobes (OL). The adult OL includes the medulla and lamina. The adult brain also includes the subesophageal ganglion (not shown in the schematic).

Figure 4

Fly Transportin is required in neurons for survival and eye function (A) The drug-inducible elav-GAL4^GS driver was used to express RNAi in adult fly neurons while avoiding RNAi expression during development. Expression of dTnpo RNAi-1 significantly impacts animal survival, indicating a progressive loss of neuron function due to dTnpo loss. (B and C) Rh1-GAL4 was used to express RNAi in mature photoreceptor neurons and electroretinograms (ERGs) were used to measure neuronal function at 7 days, 14 days, and 22 days. Blue annotation shows where amplitudes are measured. Orange bars indicate the light pulses. (D–F) dTnpo RNAi-1 nearly abolishes ON and OFF transients (D, E) and reduces the light coincident receptor potential (LCRP; F) compared to a control RNAi. Statistics: (A) log-rank, (D–F) 2-way ANOVAs with Sidak’s multiple comparisons test. ^∗p < 0.02, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. Each dot represents the mean of 5 recorded ERGs per animal. The mean from 5–6 animals is shown. Error bars denote SEM; “Control (Ctrl) RNAi” is UAS-Luciferase RNAi (TRiP.JF01355). UAS-dTnpo RNAi-1 is TRiP.HMJ23009.

Figure 5

Upregulation of dTnpo disrupts morphology of eyes and wings (A) Ubiquitous expression of UAS::dTnpo^GS11030 using da-GAL4 in flies causes a 25-fold increase in dTnpo mRNA levels by qPCR. L3 larvae were analyzed at 22°C. Unpaired t test, ^∗∗∗p = 0.0003. Each dot represents the mean from replicate wells per sample. The mean from 4 individual samples is shown. Error bars denote SD. (B) da-GAL4>UAS::dTnpo^GS11030 animals do not survive beyond pupariation at 25°C. (C and D) Upregulation of dTnpo during eye development, using either ey-GAL4 (early development) or GMR-GAL4 (late development) driven expression of UAS::dTnpo^GS11030, causes small eyes and rough eye phenotypes. Scale bar = 100 μm. (E) nub-GAL4 driven expression of UAS::dTnpo^GS11030 causes notch and blister phenotypes (arrows) in the fly wing. Scale bar = 200 μm. “Control (Ctrl)” is UAS-empty.

Figure 6

Variants in hTNPO2 cause different amounts of toxicity compared to wild-type hTNPO2 during fly development and in the eye (A) UAS-hTNPO2 fly lines were generated. Western immunoblots (WBs) confirmed hTNPO2 protein levels are similar between lines using a drug-inducible ubiquitous driver (da-GAL4^GS) to express transgenes and a human TNPO1/2 antibody. Normalized hTNPO2 band density from three independent westerns were quantified. Each dot represents one independent sample. The mean from 3 individual samples is shown. (B) da-GAL4 driven ectopic expression of UAS-hTNPO2:WT^HA reduces Mendelian ratios compared to UAS control flies, demonstrating toxicity during development. Variants p.Gln28Arg and p.Asp156Asn are more toxic than hTNPO2:WT whereas p.Trp727Cys is less toxic. Each dot represents one independent cross with >100 animals scored. The mean from three independent crosses is shown. (C and D) Ectopic expression of UAS-hTNPO2:WT^HA disrupts eye development using either ey-GAL4 (early development) or GMR-GAL4 (late development). Scale bars = 100 μm. (C) With ey-GAL4>hTNPO2:WT^HA, eyes are smaller than controls and have a rough eye phenotype. p.Trp370Cys and p.Trp370Arg are less toxic. (D) With GMR-GAL4>hTNPO2:WT^HA, eyes are moderately smaller and there is a mild rough eye phenotype compared to controls. p.Gln28Arg and p.Asp156Asn are more toxic. Statistics: 1-way ANOVAs with Dunnett’s (A) or Tukey’s (B) multiple comparisons test. no significance (n.s.) ≥ 0.05, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. Error bars denote SD “UAS Control” is UAS-(empty).

Figure 7

Variants alter hTNPO2-induced phenotypes and penetrance in the fly wing and impact of the variants corresponds with their location within the protein (A) Ectopic expression of UAS-hTNPO2:WT^HA using nub-GAL4 disrupts wing development, causing notching, blisters, and gain-of-vein phenotypes (arrows). p.Trp370Cys, p.Trp370Arg, and p.Ala546Val have less severe phenotypes whereas p.Gln28Arg and p.Asp156Asn are significantly more toxic. p.Trp727Cys has a moderately stronger gain-of-vein phenotype than hTNPO2:WT. Scale bar = 200 μm. (B) Blister and notch phenotypes caused by hTNPO2 expression in the wing occurs in 50% of wings, representing penetrance. Penetrance is significantly different for all variants except p.Trp727Cys. Statistics: 1-way ANOVAs with Dunnett’s multiple comparisons test. No significance (n.s.) ≥ 0.05, ^∗p < 0.05, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. Error bars denote SEM. Each dot represents the results from one cross with >50 animals scored. The mean from two independent experiments that included 2–3 individual crosses is shown. (A and B) “UAS Control” is UAS-(empty). (C) Table summarizing phenotype severity associated with variants when compared to hTNPO2:WT-associated phenotypes. Symbols: strong decrease in toxicity (green arrows), strong increase in toxicity (red arrows), mild increase in toxicity (orange arrows), no obvious difference in toxicity (dash). p.Trp727Cys strongly reduces toxicity in two situations and mildly increases toxicity in two situations, earning two green and one red arrow in the summary.