Analysis of the intracellular traffic of IgG in the context of Down syndrome (trisomy 21)

Persons with Down syndrome (DS, trisomy 21) have widespread cellular protein trafficking defects. There is a paucity of data describing the intracellular transport of IgG in the context of endosomal-lysosomal alterations linked to trisomy 21. In this study, we analyzed the intracellular traffic of IgG mediated by the human neonatal Fc receptor (FcRn) in fibroblast cell lines with trisomy 21. Intracellular IgG trafficking studies in live cells showed that fibroblasts with trisomy 21 exhibit higher proportion of IgG in lysosomes (~ 10% increase), decreased IgG content in intracellular vesicles (~ 9% decrease), and a trend towards decreased IgG recycling (~ 55% decrease) in comparison to diploid cells. Amyloid-beta precursor protein (APP) overexpression in diploid fibroblasts replicated the increase in IgG sorting to the degradative pathway observed in cells with trisomy 21. The impact of APP on the expression of FCGRT (alpha chain component of FcRn) was investigated by APP knock down and overexpression of the APP protein. APP knock down increased the expression of FCGRT mRNA by ~ 60% in both diploid and trisomic cells. Overexpression of APP in diploid fibroblasts and HepG2 cells resulted in a decrease in FCGRT and FcRn expression. Our results indicate that the intracellular traffic of IgG is altered in cells with trisomy 21. This study lays the foundation for future investigations into the role of FcRn in the context of DS.

IgG intracellular trafficking assays. 1 × 10 4 cells per well (i.e., cell confluence ~ 70%) were seeded into 96-well plates suitable for fluorescence microscopy and cultured for 24 h in growth medium. Growth medium was replaced with MEM supplemented with 0.05 µM LysoTracker Red DND-99 (L7528, Life Technologies) and incubated for 30 min at 37 °C. After removal of LysoTracker, cells were supplemented with 0.25 mg/ml Alexa Fluor 633-human immunoglobulin G1 (hIgG1) and incubated for 60 min to allow IgG uptake. hIgG1 was removed, replaced with MEM supplemented with CellMask Green Plasma Membrane Stain (C37608, Invitrogen), and cells were maintained in the incubator for up to 15 min before imaging. hIgG1 (Sigma-Aldrich, Cat# 400120, RRID: AB_437947) was previously labeled using an Alexa Fluor 633 Protein Labeling Kit (A20170, Molecular Probes) following the manufacturer's instructions. Incubations were performed in serum-free prewarmed MEM without phenol red (51200038, Life Technologies).
Cell imaging was performed at 37 °C with 5% CO 2 in a humidified incubator using a Dragonfly spinning disk confocal microscope (Andor Technology Ltd.) attached to a DMi8 base (Leica Microsystems). Images (16 bits, 0.096 μ per pixel) were obtained in sequential mode with a Zyla 4.2 PLUS sCMOS camera using a PlanApo 40 × 1.10 NA water immersion objective. Images from multiple fields (~ 10 fields/well) were taken, and independent incubations were performed in at least 3 different wells for each cell line. Fields were randomly selected for the imaging of healthy-looking cells.
IgG recycling assay. The IgG recycling assay was adapted from Grevys et al. 30 . 2 × 10 5 cells per well were seeded into 24-well plates and cultured for 24 h in growth medium. Cells were washed twice with PBS, starved for 60 min at 37 °C in MEM, and incubated in pre-warmed MEM with 0.50 mg/ml hIgG1 (Sigma-Aldrich, Cat# 400120, RRID: AB_437947) for 60 min at 37 °C. The medium containing IgG was removed, and cells were washed three times with PBS. Pre-warmed MEM was added to the cells and incubated for 0 min (control), or 60 min at 37 °C to allow IgG recycling. The concentrations of IgG in cell lysates and supernatants after recycling were measured with an IgG Human ELISA Kit (88-50550, Invitrogen), following the manufacturer's instructions.
Images (8 bits) were obtained in sequential mode with a point scanning confocal microscope (Carl Zeiss, LSM 510 Meta) using a PlanApo 60 × 1.40 NA oil immersion objective. Images from multiple randomly selected fields (~ 10 fields/condition) were taken. Identical microscope configuration and camera settings were maintained during image acquisition for conditions from the same experiment.
Quantitative image analysis. Image analysis was performed with the ImageJ software 31 . Comparisons were performed by analyzing similar numbers of cells per condition with identical image processing parameters.
Images were pre-processed for noisy pixel elimination and background subtraction using Gaussian smoothing followed by the rolling ball method in Fiji 32 . Cellular regions of interest (ROIs) were created by segmentation from the differential interference contrast (DIC), bright field, or CellMask Plasma Membrane Stain channel ( Figure S2). Cellular ROIs were obtained from cells with well-defined edges, healthy-looking, and non-mitotic. Cells not crossing or with a minimal area crossing the edge of the field of view were selected for analysis. For cells stained with CellMask Plasma Membrane marker, binary masks were obtained using automated global thresholding. The resulting cell masks were inspected for detection of individual cells, manually separated using the pencil tool if necessary, and ROIs were added to the ROI manager in Fiji. For cells with borders detected with DIC or bright field, cellular ROIs were manually created with the freehand selection tool, and added to the ROI manager in Fiji. To quantify the average size and number of vesicles/mm 2 per ROI, binary masks for each channel were obtained using automated global thresholding and the Analyze Particles approach (particles size ≥ 0.019 µ 2 ) followed by watershed transform to enable separation of contiguous vesicles. For colocalization analysis, Pearson's correlation coefficient (PCC) was calculated for each ROI with the Colocalization test extension using Costes' automated thresholding method within the Fiji software 31,32 . Cell transfections. For FCGRT and APP knock down, cells were transfected using Dharmafect 4 transfection reagent (T-2004-02, Dharmacon) following the manufacturer's recommendations. Briefly, 2 × 10 4 cells per well were seeded into 24-well plates 24 h prior to transfection in antibiotic free media, and transfected with 5 nM siRNA. Cells were incubated at 37 °C in 5% CO 2 for a total of 96 h, with replacement of transfection medium 24 h post-transfection. Non-Targeting siRNA Control Pool (NS-siRNA, D-001206-13-05), siRNA Pool targeting APP (M-003731-00-0005), and siRNA against FCGRT were obtained from Dharmacon. The siRNA against FCGRT was designed using the web-based software OligoWalk (FCGRT target sequence NM_001136019.2, Table S2) 33 .
For expression of proteins with fluorescent tags, cells were transfected using ViaFect transfection reagent (E4981, Promega) following the manufacturer's recommendations. Briefly, 5 × 10 4 cells per well were seeded into 24-well plates 24 h prior to transfection in antibiotic free complete media, and transfected with 500 ng APP cDNA ORF Clone C-GFPSpark tag (HG10703-ACG, Sinobiological), or co-transfected with FCGRT cDNA ORF Clone C-GFPSpark tag (HG11604-ACG, Sinobiological) and B2M cDNA ORF Clone (HG11976-UT, Sinobiological). For control conditions, cells were transfected with an empty vector (PCMV6XL5, Origene). Cells were incubated at 37 °C in 5% CO 2 for a total of 48 h. Quantitative real-time polymerase chain reaction. Total RNA was isolated from cells using Trizol reagent following the manufacturer's instructions (Thermo Fisher). FCGRT (alpha chain component of FcRn) and APP mRNA expression was analyzed with specific primers (Table S2). Total RNA (12.5 ng) was reverse transcribed and amplified with the iTaq Universal SYBR Green One-Step Kit (Bio-Rad). FCGRT , APP and the reference gene ACTB were amplified in parallel in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad) with the following cycling parameters: 50 °C for 10 min (reverse transcription), 95 °C for 1 min, followed by 44 cycles of 95 °C for 10 s, 60.5 °C for 20 s. Calibration curves were prepared to analyze linearity and PCR efficiency. qRT-PCR data were analyzed with CFX manager Software (Bio-Rad). The ΔΔCt method was utilized for determining the relative abundance of FCGRT and APP mRNA.
Data processing and statistical analysis. Data processing was performed with Excel 2016 (Microsoft Office). Statistical analyses were performed with GraphPad Prism version 8. The D' Agostino & Pearson omnibus normality test was used to determine the normality of data sets. Comparisons between the means of two groups were performed with the Student's t-test or Mann-Whitney's U test for sets with normal and non-normal distributions, respectively. Spearman's rank-order test was used for correlation analyses.

Results
Intracellular transport of IgG in the context of trisomy 21. First, optimal conditions for IgG uptake and intracellular trafficking were determined for fibroblast cell lines derived from donors with and without DS using cell imaging and ELISA ( Figures S3 and S4 (Fig. 4A,B). The subcellular distribution of FcRn in early endosomes and lysosomes was similar in cells with and without trisomy 21 (Fig. 4A,C). In cells with and without trisomy 21, the extent of colocalization of FcRn with the endosomal marker EEA1 was ~ 40% (PCC DS: 0.40 ± 0.10, PCC NDS: 0.43 ± 0.12), and the extent of colocalization of FcRn with LAMP1 was ~ 10% (PCC DS: 0.10 ± 0.05, PCC NDS: 0.09 ± 0.03).

Morphological features of compartments involved in the traffic of IgG in cells with
FCGRT and APP expression in the context of trisomy 21. Trisomic fibroblasts showed higher FCGRT mRNA and FcRn protein expression in comparison to diploid cell lines derived from age-and sexmatched donors (Fig. 5, Table S1). FCGRT mRNA expression was ~ 1.5-to 5-fold higher in trisomic cells, and on average, FcRn protein levels were ~ 3-fold higher in trisomic cells than in diploid cells (Fig. 5A). Fibroblasts with trisomy 21 showed ~ 2-fold higher APP mRNA expression than diploid cells (DS: 2.14 ± 0.64 relative fold, NDS: 1.00 ± 0.34 relative fold) (Table S3). There was a significant positive correlation between the endogenous expressions of APP and FCGRT mRNA in diploid and trisomic cells (R 2 = 0.97, Pearson r = 0.99, P < 0.001) (Fig. 5B).
Impact of APP on the expression of FcRn. The impact of APP on the expression of FCGRT was investigated with two strategies. First, fibroblasts were transfected with an anti-APP siRNA pool (Fig. 5C). APP knock down (APP mRNA expression < 10%) resulted in a ~ 60% increase in the expression of FCGRT mRNA in the diploid cell line AG07095 (NDS control siRNA : 99.7 ± 6.6, NDS APP siRNA : 167.5 ± 22.6% relative fold) and in the trisomy 21 cell line AG06922 (DS control siRNA : 99.7 ± 16.4, DS APP siRNA : 160.0 ± 6.22% relative fold). The second strategy involved the overexpression of the APP protein in diploid cells. APP-GFP expression was confirmed by fluorescence microscopy and immunoblotting (Fig. 5D,E). The specificity of APP and FcRn antibodies was examined by immunoblotting in lysates from non-transfected and transfected cells (Figures S5 and 6). Rab11 expression assessed by immunoblotting was similar between cells overexpressing APP and controls transfected with empty vector ( Figure S7). Diploid AG07095 fibroblasts overexpressing APP-GFP showed a ~ 86% decrease in the expression of FCGRT mRNA (NDS control : 100.0 ± 25.6, NDS APP-GFP : 14.3 ± 2.7% relative fold), and a ~ 7% decrease in FcRn protein expression (Fig. 5D). These observations were extended by examining the interplays between APP and FcRn expression in HepG2 cells. HepG2 cells express relatively high basal levels of FcRn and APP as evidenced by immunoblotting ( Figure S6). HepG2 cells overexpressing APP-GFP showed a ~ 40% reduction in the expression of FCGRT mRNA (HepG2 control : 100.0 ± 13.3, HepG2 APP-GFP : 60.8 ± 12.6% relative fold), and a ~ 23% decrease in FcRn protein expression (Fig. 5E).

Discussion
With the exception of neuronal phenotypes in DS, the impact of endosomal-lysosomal dysfunction on prominent receptor-mediated protein trafficking pathways has been scarcely documented 28,34,35 . In this study, we analyzed the FcRn mediated-intracellular traffic of IgG in human fibroblasts with trisomy 21, a cellular model that recapitulates protein trafficking defects linked to DS 27,36 .
FcRn is the only high affinity and pH specific receptor of IgG 23 . Monomeric IgG internalized by fluid phase pinocytosis is protected from lysosomal degradation by pH-dependent binding to FcRn at endosomes (pH 5.5-6.0). FcRn then transports IgG in Rab11 + recycling endosomes back to the extracellular media or across cell layers in polarized cells 21,37 . Our intracellular IgG trafficking studies showed that fibroblasts with trisomy 21 exhibit a higher proportion of IgG in lysosomes, decreased IgG content in intracellular vesicles, and a trend towards decreased IgG recycling in comparison to diploid cells (Figs. 1 and 2). In general, recycling endosomes are tubulovesicular structures, and recent studies showed FcRn-mediated recycling of IgG in tubulovesicular transport carriers 21,38,39 . Here, we documented ~ 10% decreases in size and number of IgG-containing intracellular vesicles in fibroblasts from donors with trisomy 21. In diploid cells, many of the IgG + vesicles displayed a tubular-like shape with low overlapping to late endosomes/lysosomes. In contrast, IgG + vesicles in trisomic cells were more puncta-like shaped, smaller than the tubular structures, and overlapped more frequently with late endosomes/lysosomes (Fig. 1A). The increased expression of FcRn in trisomic cells did not result in increased   (Figs. 1, 2 and 5A,B). These observations suggest that increased FcRn expression in cells with trisomy 21 is a compensatory response for decreased FcRn activity. Fibroblast cell lines expressed relatively low levels of endogenous FcRn, and detection of FcRn expression by immunofluorescence was difficult (Figures S5 and 6). For this reason, an FcRn-GFP construct was used to analyze the subcellular distribution of FcRn 21,32 . FcRn-GFP localized mainly in EEA1 + endosomes, and the distribution in sorting endosomes and LAMP1 + lysosomes was similar between trisomic and diploid cells (Fig. 4A,C). FcRn + vesicles in fibroblasts with trisomy 21 were more numerous (~ 23% increase) and slightly bigger (~ 6% increase) than vesicles in diploid cells (Fig. 4B). The subcellular localization of FcRn in fibroblasts was in line with previous studies documenting the distribution of FcRn in other cell types (transfected and non-transfected cells) 21,40,41 . Our observations suggest that the sorting of IgG to recycling endosomes is decreased in cells with trisomy 21 with a consequent increase in lysosomal degradation. Abnormal targeting of FcRn to lysosomes appears not to be a cause for increased IgG detection in lysosomes.
Alterations in the morphology of endosomes have been described in primary fibroblasts, neurons, peripheral blood mononuclear cells, and lymphoblastoid cell lines from individuals with DS 27,36,[42][43][44] . Our comparative morphological analysis of IgG + vesicles and subcellular compartments in fibroblasts with and without trisomy 21 revealed enlargement of subcellular organelles from the endosomal pathway, resembling the "traffic jam" previously described for DS and Alzheimer's disease 27,44,45 . We found that trisomic cells have larger EEA1 + endosomes (~ 12%), Rab11 + endosomes (~ 24%), and Lysotracker-stained lysosomes (28%) than diploid cells (Fig. 2). In agreement, other reports have noted a ~ 18% increase in the size of EEA1 + vesicles without significant differences in the number of this type of vesicles. Enlargement of lysosomes in fibroblasts from individuals with trisomy 21 has also been reported 44 . Of note, the average size of LAMP1 + vesicles in fixed cells with trisomy 21 was smaller. www.nature.com/scientificreports/ It remains to be determined whether the observed differences reflect vesicular heterogeneity or result from microscopy approaches (i.e., fixed cells vs live cells), or a combination of both factors 46 . Although the magnitudes of some of the observed cellular changes may appear modest at first glance, effect sizes are in line with other cellular phenotypes in DS (e.g., mitochondrial dysfunction, endosomal defects) 35,47,48 . Several studies implemented receptor-mediated uptake approaches to examine endosomal defects in trisomic cells 36,44,49 . There is limited information describing IgG traffic following internalization by fluid phase pinocytosis in the context of trisomy 21 21,50 . Cataldo et al. documented increased endocytic fluid phase uptake of horseradish peroxidase in trisomic fibroblasts 36 . We did not detect increases in intracellular IgG content or number and size of IgG + vesicles during the IgG trafficking assays (Figs. 1C and 2A). It is possible that under our experimental conditions, there is uptake of IgG followed by cellular recycling and re-uptake with no IgG accumulation. An Evidence indicates that the increased expression of APP is linked to widespread cellular endosomal-lysosomal abnormalities in DS 28,29 . The APP gene triplication contributes to the development of early-onset AD in DS 28 . APP is processed in endosomes by the β-site APP cleaving enzyme 1 (BACE1) to form APP β-carboxyl-terminal fragment (APP-βCTF). In DS and AD, APP-βCTF over-activates Rab5 by recruiting APPL1, resulting in endosomal trafficking defects and abnormal endosome-mediated signaling 27,28 . We observed that APP overexpression www.nature.com/scientificreports/ in diploid fibroblasts replicated the increase in IgG sorting to the degradative pathway observed for cells with trisomy 21 (i.e., ~ 20% increase in IgG into lysosomes and ~ 16% reduction in IgG content. Figures 1 and 6). In diploid fibroblasts overexpressing APP, the number and size of intracellular IgG vesicles remained unaltered, which suggests that other factors contribute to morphological alterations in trisomic cells in addition to increased APP expression 43,44 . It remains to be defined whether the increased sorting of IgG to lysosomes is a direct consequence of APP overexpression, or is mediated by other proteins regulated by APP, or by other genes altered in trisomy 21. Examples of other trisomy 21 genes with roles in endosomal function and protein transport include SYNJ1 and DOPEY2. Elevated SYNJ1 expression alters the metabolism of the signaling phospholipid phosphatidylinositol-4,5-biphosphate (PI [4,5]P2), and results in enlargement of early endosomes and protein sorting defects in DS 28,43 . DOPEY2 is part of an endosome-associated membrane re-modelling complex that is involved in SNX3-retromer function 51 . Retromer complex proteins have also been shown to be reduced in cells from subjects with DS 52 . It is probable that the traffic of IgG in the context of DS is modified by diverse mediators which may or may not act similarly in different cell types. It is now recognized that the complex and variable architecture of DS phenotypes is modeled by the combined effects of relatively subtle spatiotemporal changes in multiple cellular pathways 34,53 . Analysis of endogenous expression of FCGRT and APP mRNA in fibroblasts derived from donors with and without trisomy 21 revealed substantial "inter-individual" variability in gene expression (APP range: ~ 1-to 5-fold, FCGRT range: ~ 1-to 9-fold). There was a positive correlation between APP and FCGRT mRNA expression levels in the group of diploid and trisomic cell lines (Fig. 5B). This observation prompted us to further examine potential interplays between APP and FcRn expression. Notably, APP knock down increased the expression of FCGRT mRNA by ~ 60% in both diploid and trisomic cells (Fig. 5C). Furthermore, overexpression of APP in diploid fibroblasts and HepG2 cells also resulted in a decrease in FCGRT and FcRn expression (Fig. 5D,E). It is known that APP plays many roles and a growing amount of evidence suggests that APP may also act as a transcriptional regulator [54][55][56] . Cleavage of APP-βCTF by γ-secretase results in the generation of Amyloid-β (Aβ) and APP intracellular domain (AICD) 57 . Studies suggest that AICD regulates gene transcription, although the exact role that this fragment plays is still controversial because detection of endogenous AICD is difficult 56, 58-60 . It will be of interest to test whether the AICD fragment regulates the expression of FcRn. Points that merit further consideration are (1) whether the observed FCGRT -APP gene expression correlation extends to "inter-individual" comparisons involving various cell types and tissues from multiple donors, and (2) additional regulatory mechanisms that may impact FCGRT and FcRn expression in trisomic cells 61,62 .
Our results suggest that the intracellular traffic of IgG is altered in cells with trisomy 21. This study lays the foundation for future investigations into the role of FcRn in the context of DS. For example, children with DS are susceptible to infections (e.g., recurrent respiratory infections) and many suffer from chronic immune dysregulation 63,64 . FcRn plays an important role during the regulation of innate immune responses and immunosurveillance at mucosal sites 24 . It is possible that alterations in IgG traffic extend to other cell types (e.g., cells of hematopoietic origin, epithelial cells, and endothelial cells) and could eventually limit the availability of IgG in specific tissues in individuals with DS. Polarized epithelial cells play a fundamental role in the transport of immunoglobulins across the mucosa 24 . It would be of interest to analyze the traffic of IgG, therapeutic mAbs (e.g., palivizumab), and other types of immunoglobulins (e.g., IgA and IgM) in polarized cells with trisomy 21. In addition, it is licit to hypothesize that augmented transport of IgG to the degradative pathway in dendritic cells with trisomy 21 could impact antigen presentation of peptides derived from IgG immune complexes (IgG IC) 25 .

Data availability
The datasets generated during the current study are available from the corresponding author upon request.