Cholinotrophic basal forebrain connectome dysfunction in Down syndrome with and without dementia

Summary

Cholinotrophic basal forebrain (CTBF) neurons depend upon the complex interaction of both upstream and downstream nerve growth factor (NGF) signaling pathways for survival and function. Although dysfunction of the NGF system occurs in both Down syndrome (DS), not all individuals with DS develop dementia. Whether NGF system dysregulation differs between demented individuals with DS (DSD+) versus those without dementia (DSD-) is unknown. Here, we report a significant reduction in neurons positive for the p75^NTR within the nucleus basalis of Meynert in DSD+, but not DSD-, compared to aged-matched controls (AMC). ChAT positive cells were significantly lower in DSD+ compared to DSD- and AMC cases. FC levels of p75^NTR and proNGF were increased, while ChAT levels decreased in DSD+ compared to AMC. A greater number of AT8 tau positive neuropil threads and Thioflavin-S labeled neurofibrillary tangles were found in DSD+. These findings suggest a greater role for p75^NTR/proNGF in demented in individuals with DS compared to those with dementia. The factors that underlie cognitive resilience in DSD- remains to be determined.

Graphical abstract

Highlights

• •p75^NTR and ChAT positive neurons in nbM are reduced in individuals with DS with dementia
• •Tau pathology in the nbM was greater in DS individuals with dementia than in those without
• •Cortical proNGF levels are higher in individuals with DS and dementia than in controls
• •Cortical p75^NTR/proNGF stability may underline resilience to dementia in individuals with DS

Neuroscience; Cell biology

Introduction

Down syndrome (DS), a genetic disorder caused by the presence of an extra copy of chromosome 21 (HSA21), is one of the most common genetic conditions associated with intellectual disability.¹ Due to great advances in medical care, the life expectancy for a person with DS has increased into the seventh decade of life.^2,3 Trisomy of HSA21 also increases the risk of developing the classic pathological hallmarks of AD, amyloid-β (Aβ) plaques and tau-containing neurofibrillary tangles (NFTs) in the 4th or 5th decades of life.^4,5,6 Interestingly, there is a disconnect between the prevalence of Aβ plaque and NFT pathology and dementia in DS.^7,8,9,10,11 While people with DS are at a higher risk for developing dementia decades prior to patients with sporadic AD,^12,13,14,15 there is a wide age range for the onset of the AD type dementia in DS. Many individuals exhibit cognitive decline only after the age of 55¹⁶ but only two-thirds of individuals with DS eventually develop dementia.^17,18,19,20 Although dementia may be a contributing factor to the death of individuals with DS, the most frequently reported causes of death in this population include respiratory infections (such as pneumonia), congenital heart defects, and various circulatory system diseases.^21,22 Since not all people with full or partial trisomy develop dementia,²³ it is critical to compare the neuropathobiology between individuals with DS with or without dementia to better understand possible factors that impart brain/cognitive resilience in DS.

Another pathognomonic hallmark in DS is a reduction in the volume and degeneration of cholinergic neurons located in the basal forebrain, that innervate the entire cortical mantle, hippocampus, amygdala, and contribute to cognitive impairment in both DS and AD.^{24,25,26,27,28,29,30,31,32} The survival of these cholinergic neurons is dependent upon the anterograde transport of the high affinity TrkA and the low affinity p75^NTR cognate receptor for nerve growth factor (NGF)³³ to their cortical projection sites, that contain NGF and its precursor protein, proNGF and related NGF metabolites.^{34,35,36,37,38,39,40,41} Based upon these associations, we refer to these neurons as cholinotrophic basal forebrain (CTBFs) neurons.

Following endocytosis, mature NGF (mNGF) and/or proNGF bind with their cognate receptors and are retrogradely transported to CTBF consumer neurons.⁴² Moreover, CTBF neurons are vulnerable to microtubule dysfunction in the form of tau intraneuronal tangles and neuropil threads in DS¹⁵ and AD,⁴³ which are formed in the absence of NGF.⁴⁴ Overall, these neurons are dependent on a complex interaction of the NGF system for their maintenance and survival in both DS^35,45,46,47 and AD.^{36,48,49,50,51} Interestingly, there is an upregulation in choline acetyltransferase (ChAT) activity in the frontal cortex (FC) and hippocampus⁵² in prodromal AD, which is driven by a biochemical increase in this protein in nbM neurons and not structural changes in the cortex.⁵³ This cholinotrophic pathway may represent a potential correlate of brain resilience in AD but whether a similar compensation occurs in trisomic individuals who remain dementia-free later in life remains unknown.

Here, we evaluated changes in NGF/proNGF pathways, ChAT levels, and tau pathology in cholinotrophic neurons within the nbM and cortical projection sites using both immunohistochemistry and western blotting in postmortem tissue collected from individuals with DS with and without dementia provided by the Down Syndrome Biorepository Consortium (DSBC).⁵⁴ Findings derived from these studies will provide novel information about the differential involvement of the cholinotrophic cortical system with translation to brain resilience and drug targets to treat people with DS.

Results

Demographics

There was no significant difference in sex, age at death, postmortem interval (PMI), and APOEε4 carrier status between DSD+, DSD− cases and AMC cases (p > 0.05, see Table 1). Brain weight was significantly lower in DSD+ compared to the DSD− and AMC cases (Mann-Whitney test: AMCs = DSD+<DSD-, p = 0.0004, see Table 1). Braak tau scores were significantly higher in DSD+ compared to DSD− cases (Mann-Whitney test: DSD+>DSD-, p < 0.0001). A significant difference in the distribution of neuropathological ABC score profiles was observed among DSD−, DSD+, and AMC (Fisher’s exact test, p = 0.007). Table 1 contains the average demographic data for these variables for each group (AMC, DSD- and DSD+).

Table 1.

Demographic and neuropathological data of DSD-, DSD+ and AMC

	AMC (n = 8)	DSD- (n = 8)	DSD+ (n = 18)	p value
Age (y) [range]	65.13 ± 5 [53,64]	53.25 ± 2 [36,72]	53.85 ± 2 [46,70]	0.3277a
PMI (h) [range]	5.82 ± 2 [2.7,18.3]	8.245 ± 1 [2.2,6.5]	3.647 ± 1 [3.1,11.5]	0.3722a
Male/Female	4/4	3/5	10/8	0.5956b
Brain weight (g)c [range]	1250 ± 80 [1015,1456]	1115 ± 32 [961.6,1217]	952.2 ± 20 [823,1093]	0.0004a
ApoE genotypec				0.999b
23	1	1	1
33	4	2	7
24	0	0	1
34	0	1	1
Braakc				<0.0001a
I-II	3	1	0
III-IV	0	6	0
V-VI	0	1	18
ABCc				0.007b
A score	0–2	2–3	3
B score	1	1–2	3
C score	0–2	2–3	2–3

Data are presented as mean ± SEM.

^aMann-Whitney rank-sum test.

^bFisher exact test.

^cWe do not have information about all the cases.

Quantification of CTBF neurons within the nbM in DSD+, DSD- and age matched controls

CTBF neuron density within the nbM was quantified in sections immunolabeled with antibodies against p75^NTR and ChAT. In the AMC group we observed cytoplasmic p75^NTR and ChAT immunoreactivity in nbM neurons that displayed well-defined axonal and dendritic processes (Figures 1A and 1D). DSD-neurons displayed a similar morphology but less intensely stained with less well-defined processes (Figures 1B and 1E) compared to the AMC cases. By contrast, neurons that displayed each protein had a globose appearance with virtually no visible processes in DSD+ cases (Figures 1C and 1F). Quantitative analysis revealed a significant decrease in the number of p75^NTR positive neurons in DSD+ compared to AMCs cases (Kruskal-Wallis followed by a Dunn’s test, p = 0.0001), but a similar reduction was not seen between DSD- and the AMC group. Although the averaged p75^NTR positive neuron number in DSD- was greater than DSD+, we found a non-significant difference between groups (Kruskal-Wallis follow by a Dunn’s test, p > 0.05). There was a significant reduction in the number of ChAT positive neurons in DSD+ compared to both DSD- and AMC cases (Kruskal-Wallis followed by a Dunn’s test: AMC>DSD+, p = 0.0002; DSD->DSD+, p = 0.04).

Figure 1.

Images and quantitation of cholinotrophic basal forebrain neuron pathology within the nucleus basalis of Meynert in DSD+, DSD- and AMC cases

(A–N) Photomicrographs show p75^NTR (brown, A-C), ChAT (purple, D-F) positive cells in AMC, DSD-, and DSD+ cases. Note the decrease in the number and intensity of both p75^NTR and ChAT labeled neurons in DSD- cases compared to an even greater reduction in DSD+ and globose shaped cells (black arrows) in DSD+ cases (C, F). Arrows in (A–F) mark cells shown at a higher magnification in the boxed areas located at the lower right corner of each panel. Dark brown AT8 (G, I) and TauC3 (H, J) bearing neurofibrillary tangles (NFTs) were observed only in the DSD- and DSD+ cases. Black arrows indicate globose shaped NFTs (G, H, and I) that are shown at a higher magnification in boxed areas adjacent to the lower magnification images. Note that not all neurons within the nbM (thin black arrows) contained tau pathology (G, H) in DSD- cases. Moreover, TauC3 staining revealed two NFT phenotypes that displayed either peripherally located or intense labeling that filled the entire structure (see boxed images adjacent to (H) and (J)). Scale bar in F = 50 μm and inset = 20 μm applies to panels A-E. Scale bar in J = 20 μm and inset = 20 μm applies to (G–J). Histograms show a significant reduction in both p75^NTR (K) and ChAT (L) positive cells in DSD+ compared to AMC. Although no significant were found in number of AT8 or TauC3 NFT positive cells (M), there was an increase in NTs in DSD+ compared to DSD- (N). ACM n = 5, DSD- n = 5, DSD+ n = 10. Data shown are presented as mean ± SEM. Statistical significance was determined using the Kruskal–Wallis’s test followed by Dunn’s test for comparisons across three clinical groups, and the Mann–Whitney test for comparisons between two groups (DSD- vs. DSD+). Significance levels (∗) were set at: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Quantification of tau pathology in CTBF neurons within the nbM in DSD+ and DSD-

Tau positive neuropil threads (NTs) and NFT counts within the nbM were performed using antibodies against AT8 and TauC3. Virtually no AT8 or TauC3 immunostaining was seen in the nbM in AMC cases (data not shown). AT8 positive NFTs and NTs were observed in 100% of DSD+ but in only 60% of DSD-cases. TauC3 labeled NFT were present in 54% of the total DSD+ and 40% of DSD-cases. Although AT8 positive NFTs were associated with a higher number of NTs in DSD+ (Figure 1I), a trend in this direction was not observed in DSD- (Figure 1G). TauC3 staining revealed NFTs that displayed either condensed cytoplasmic immunoreactivity at the periphery of the cell or a densely stained globose shaped resembling skein of yarn profiles in both DS cohorts (Figures 1H and 1J). Statistical analysis found no significant difference in the number of AT8 or TauC3 positive NFTs between DSD+ and DSD-cases. However, the number of AT8 labeled NTs were significantly greater in DSD+ compared to DSD- (Figures 1M and 1N; Mann-Whitney test, p = 0.03). Qualitative analysis of Nissl-stained sections revealed a greater number of neurons displaying cholinergic-like phenotypes in the nbM of DSD− compared to DSD+ cases. Many Nissl-positive neurons did not show AT8 immunoreactivity (Figures 1G and 1I).

Quantification of p75^NTR neurons with and without AT8 and MAP2 in the nbM of DSD+ and DSD-

The number of nbM p75^NTR positive cells, whether tau-positive or negative, that displayed MAP2 staining was evaluated in DSD- (Figures 2A–2D) and DSD+ (Figures 2E–2H) cases. We found a significant reduction in p75^NTR positive AT8 negative neurons in DSD+ compared to DSD- (Figure 2S; Mann-Whitney test, p = 0.0079), while no difference was seen in dual stained for p75^NTR and AT8. We also found a significant reduction in p75^NTR positive cells that colocalized with MAP2 in DSD+ compared to DSD- (Figure 2S; Mann-Whitney test, p = 0.0079), while no difference was found in p75^NTR positive MAP2 negative neurons between DS groups.

Figure 2.

Immunofluorescent images and quantification of p75^NTR neurons with and without AT8 and MAP2, as well as Thioflavin S single and AT8/Thioflavin S dual-labeled nbM cells in DSD+ and DSD-cases.

(A–U) Immunofluorescent neurons labeled with p75^NTR (A, E, green), MAP2 (B, F, red), AT8 (C, G, Cyan), ThS (blue) (I, L) and merge images (D, H, K, N, R) in DSD- and DSD+ cases. In the DSD-cases there were greater p75^NTR cells AT8 negative neurons and p75^NTR MAP2 dual labeled neurons (A-D) compared to the DSD+ cases (E–H). Note that not all p75^NTR cells colocalize with AT8 or MAP2 in both DS groups (D, H). To determine the stage of a tangle tissue was stained for ThS, a marker of advanced pathology and the early stage AT8 phosphorylation antibody. Note that there are only a few ThS-labeled tangles that displayed AT8 in both DS groups (yellow arrows), compared to single ThS tangles, which were greater in DSD+ than in DSD- (white arrows) (I-L). Note that ThS positive NFTs that do not contain AT8 also do not colocalize with MAP2 (O-R). Scale bar in F = 25 μm applies to panels A-G, N = 25 μm applies to panels I-M and R = 10 μm and applies to O-Q. Graph showing a significant reduction in both p75^NTR AT8 immuno-negative, and p75^NTR MAP2 dual labeled neurons in DSD+ compared to DSD- (S). ThS-positive cells were greater (T), while the percentage of double-labeled cells with AT8 and ThS decreased (U) in DSD+ compared to individuals without dementia. DSD- n = 5, DSD+ n = 5. Data are presented as mean ± SEM. Statistical significance was determined using Mann–Whitney test for comparisons between DSD- and DSD+. Significance levels (∗) were set at: ∗p < 0.05, ∗∗p < 0.01.

Quantification of ThS single and AT8/ThS dual labeled nbM cells in DSD+ and DSD-

We performed ThS histochemistry to reveal the presence of β-pleated sheet structures in NFTs. In DSD-cases, almost 100% of AT8 positive NFTs colocalized with ThS (Figures 2I–2K), whereas not all ThS positive NFTs colocalized with AT8 in DSD+ cases (Figures 2L–2N). Statistical analysis revealed that the percentage of AT8/ThS dual-labeled cells was significantly lower in DSD+ compared to DSD- (Figure 2U, Mann-Whitney test, p = 0.007) and the number of ThS positive NFTs was greater in DSD+ than in DSD- (Figure 2T, Mann-Whitney, p = 0.01). However, ThS and AT8-did not colocalize with MAP2 in nbM cells in DSD+ (Figures 2O–2R). ThS staining did not reveal the presence of β-amyloid plaques in the neuropil containing the nbM, but these lesions occurred in the DS cortex.¹⁵

Frontal cortex (FC) cholinotrophic protein levels in DSD+, DSD- and AMC

In DS, the FC, which is innervated by CTBF neurons displays an atypical cortical lamination^55,56 as well as other cellular alterations^57,58,59,60 compared to neurotypical developing counterparts. Therefore, we performed Western blots to determine levels of cholinotrophic proteins in the FC in DS compared to AMC cases. Analysis revealed that FC proNGF (Figure 3A) and p75^NTR (Figure 3B) protein levels were significantly upregulated in DSD+ compared to AMC, but not in DSD-cases in either comparison (Figure 3, Kruskal-Wallis followed by Dunn’s test: p75^NTR, p = 0.02; proNGF, p = 0.03). Cortical ChAT (Figure 3C) levels decreased significantly in DSD+ compared to AMC, but no differences were found between the DS groups (Figure 3, Kruskal-Wallis followed by Dunn’s test, p = 0.02). No significant changes in cortical TrkA (Figure 3D) proteins were observed between DSD+ and DSD-groups, nor when compared to AMCs.

Figure 3.

FC cholinotrophic protein levels in AMC, DSD- and DSD+

(A–D) Representative immunoblots, and bar graphs show a significant upregulation of (A) proNGF and (B) p75^NTR, while (C) ChAT protein was downregulated between AMC and DSD+. (D) TrkA protein levels were stable across the groups analyzed. ACM n = 5, DSD- n = 5, DSD+ n = 13. Data are presented as mean ± SEM. Statistical significance was determined using the Kruskal–Wallis’s test followed by Dunn’s test for comparisons across three clinical groups. Significance levels (∗) were set at: ∗p < 0.05.

Association between cholinergic and tau markers within the nbM in DS

Spearman correlations were performed to determine relationships between cholinergic nbM neuron markers and tau pathology, as well as with cortical protein levels across the DS groups. We found that both the number of p75^NTR and ChAT positive cells negatively correlated with the number of AT8 positive NTs (p75^NTR; Spearman correlation r = −0.80, p = 0.00001, ChAT; Spearman correlation r = −0.79, p = 0.00001, see Table 2) and ThS-positive NFTs across groups (p75^NTR; Spearman correlation r = −0.67, p = 0.002, ChAT; Spearman correlation r = −0.68, p = 0.001, see Table 2). FC p75^NTR protein levels negatively correlated with p75^NTR cell number (Spearman correlation r = −0.79, p = 0.004, see Table 2), while a positive association was found with AT8 labeled NTs (Spearman correlation r = 0.92, p = 0.0009, see Table 2), ThS positive NFTs (Spearman correlation r = 0.92, p = 0.0007, see Table 2) and FC proNGF protein levels (Spearman correlation r = 0.65, p = 0.0007, see Table 2). Finally, an inverse correlation was found between FC ChAT levels and AT8 positive NFTs across the DS groups (Spearman correlation, r = −0.89, p = 0.0005, see Table 2).

Table 2.

Correlations between nbM cells count, tau and cortical protein levels across groups

	p75NTR+cells (nbM)		ChAT+cells (nbM)	AT8+NFT (nbM)	AT8+NT (nbM)	TauC3+NFT (nbM)	ThS+NFT (nbM)	proNGF (FC)	p75NTR (FC)	TrkA (FC)	ChAT (FC)
p75NTR+cells	r	1	0.89663	−0.32557	−0.80918	−0.23759	−0.67605	−0.61188	−0.79910	−0.20548	0.27398
(nbM)	p		<0.00001	0.12055	0.00001	0.29970	0.00207	0.04983	0.00454	0.54154	0.41192
ChAT+cells	r	0.89663	1	−0.31779	−0.79268	−0.12178	−0.68572	0.51481	0.69704	−0.04100	0.16401
(nbM)	p	0.00000		0.13020	0.00001	0.59897	0.00168	0.10832	0.02075	0.90820	0.62752
AT8+NFT	r	−0.32557	−0.31779	1	0.83617	0.48393	0.88075	0.07169	0.34889	0.32021	−0.89372
(nbM)	p	0.12055	0.13020		<0.00001	0.02623	0.000001	0.83585	0.29097	0.33365	0.00053
AT8+NT	r	−0.80918	−0.79268	0.83617	1	0.26680	0.93589	0.61290	0.92787	0.49373	−0.67249
(nbM)	p	0.00001	0.00001	<0.00001		0.23003	<0.000001	0.08717	0.00093	0.17970	0.05529
TauC3+NFT	r	−0.23759	−0.12178	0.48393	0.26680	1	0.67898	0.04910	0.52786	0.44193	−0.72427
(nbM)	p	0.29970	0.59897	0.02623	0.23003		0.00100	0.91488	0.18333	0.26786	0.05536
ThS+NFT	r	−0.67605	−0.68572	0.88075	0.93589	0.67898	1	0.62578	0.92616	0.45056	−0.77597
(nbM)	p	0.00207	0.00168	0.000001	<0.000001	0.00100		0.11250	0.00357	0.27143	0.03393
proNGF	r	−0.61188	−0.51481	0.07169	0.61290	0.04910	0.62578	1	0.65020	0.37273	0.20000
(FC)	p	0.04983	0.10832	0.83585	0.08717	0.91488	0.11250		0.00078	0.26075	0.55741
p75NTR	r	−0.79910	−0.69704	0.34889	0.92787	0.52786	0.92616	0.65020	1	0.54051	0.55455
(FC)	p	0.00454	0.02075	0.29097	0.00093	0.18333	0.00357	0.00078		0.00775	0.08181
TrkA	r	−0.20548	−0.04100	0.32021	0.49373	0.44193	0.45056	0.37273	0.54051	1	0.27273
(FC)	p	0.54154	0.90820	0.33365	0.17970	0.26786	0.27143	0.26075	0.00775		0.41813
ChAT	r	0.27398	0.16401	−0.89372	−0.67249	−0.72427	−0.77597	0.20000	0.55455	0.27273	1
(FC)	p	0.41192	0.62752	0.00053	0.05529	0.05536	0.03393	0.55741	0.08181	0.41813

The values in bold in the table represent statistically significant correlations. Statistical significance was determined using Spearman correlation. Significance levels were set at: p < 0.005.

r: correlation coefficient.

p: statistical significance of that correlation.

Discussion

Dysfunction of NGF signaling pathways contribute to CTBF neuronal degeneration and cognitive decline in AD and DS.^61,62,63,64 Although there is a large body of literature on the involvement this system in the selective vulnerability of the CTBF cortical connectome in AD,^{24,25,26,27,28,29,30,31,65} less is known about individuals with DS with and without dementia. In this study, we quantified cholinotrophic neuron numbers in the nbM and assessed levels of proNGF, TrkA, p75^NTR, and ChAT proteins in the FC in DSD+ and DSD- cases compared to age-matched disomic controls. Here, we report a significant reduction in p75^NTR-positive nbM neurons in DSD+, but not DSD- compared to age-matched non-trisomic cases, while ChAT-positive cells were significantly lower in DSD+ compared to both DSD- and non-trisomic controls. Although there was no significant difference between the number of AT8 or TauC3-positive NFT between DS groups, we found a significantly greater number of AT8 positive NTs and ThS-labeled NFTs in DSD+ cases. Biochemical analysis revealed that FC levels of p75^NTR and proNGF were increased only in DSD+ compared to age-matched disomic controls. FC TrkA levels did not differ from AMC in both DS groups. FC ChAT levels decreased in DSD+ compared to disomic controls. We also found levels of p75^NTR in the FC that positively correlated with AT8-positive NTs and ThS-positive NFTs in the nbM, while FC ChAT levels negatively correlated with AT8-positive NFTs in nbM.

The NGF TrkA receptor is encoded by the Trk protooncogene and signals through autophosphorylation of tyrosine kinases^66,67,68 and the retrograde transport of the NGF/TrkA complex activates downstream cell survival mechanisms in CTBF neurons by activating Akt, a member of the serine-threonine protein kinase family.⁶⁹ Although TrkA has been investigated in AD,^28,29,70 there are few detailed reports examining Trk receptors positive neurons or cortical protein levels in DS. In this regard, a single studied evaluated TrkA positive cholinotrophic neuron numbers in adults with DS with dementia (mean age 52.3 ± 3 years) compared to healthy aged controls (mean age 64.5 ± 6 years) using an antibody that labeled the cytoplasm of multipolar and fusiform neurons that exhibited extensive dendritic processes within the nbM.⁷⁰ There was a 47% reduction in TrkA neurons in DS cases with dementia⁷⁰ and this cell loss was validated in Nissl-stained sections, suggesting a frank loss of these nbM neurons in DS, which contrasts to a phenotypic down regulation of both p75^NTR and TrkA but not ChAT neurons within the nbM during the progression of AD.^71,72 Perhaps different mechanisms underline neuronal degeneration related to trisomy vs. age-related events between these neurological disorders. In this study, TrkA positive neurons appeared shrunken with a globose shape, and exhibited reduced dendritic processes and broken varicose fibers. There was a gradient of cellular TrkA protein labeling ranging from light to dark in both dystrophic and non-dystrophic neurons in DS, suggesting a defect in the production of this cell survival receptor even prior to frank cellular degeneration in DS with dementia.⁷⁰ This study also reported a reduction in TrkA cortical levels, which ranged from 40% to 60% of normal levels in the occipital and temporal cortical regions to a nearly total loss in the parietal cortex in DS.⁷⁰ However, in the current study, we found that FC TrkA levels were similar in both DS groups and in age-matched non-trisomic cases, suggesting that TrkA in FC is less affected. Functionally, these findings suggest differential cortical defects in the anterograde transport of TrkA receptors and the retrograde transport of bound NGF to neurons within the nucleus basalis resulting in atrophy and/or degeneration of this highly vulnerable continuum of cells in DS.⁴⁹ A reduction in the number of nbM neurons containing both the protein and genes for TrkA has been reported in prodromal and frank AD.^50,51 By contrast, hippocampal TrkA protein levels are reduced in prodromal but return to near normal levels in AD.²⁹ In a similar vein, there is an upregulation of ChAT activity in the FC and hippocampus during the progression of AD^52,53 suggesting brain resilience or plasticity during the onset of AD. This remains to be determined in asymptomatic cases of DS.

In the present study, we also observed a significant reduction in p75^NTR positive neurons in the nbM of individuals with DS with dementia compared to age-matched disomic controls, but no significant difference was seen in comparison to individuals with DS without dementia. Although it is well accepted that p75^NTR interacts with the TrkA receptor to activate cell survival pathways,^{49,67,73,74,75,76,77} it also acts together with its co-receptors sortilin, found in the intracellular membrane⁷⁸ and neurotrophin receptor homolog-2 (NRH2), that enhances the complexing of the pan-neurotrophin p75^NTR receptor, with proNGF shifting the balance from survival to cell death. Although there are no studies of these co-receptors in DS, both p75^NTR and its co-receptors were preserved in the hippocampus in people who died with an antemortem clinical diagnosis of mild cognitive impairment and AD compared to aged controls.²⁹ Like TrkA, there is a dearth of information defining the effect that trisomy has upon p75^NTR containing neurons within the nbM in DS. A review of the literature revealed a single report that mentioned p75^NTR positive neurons within the nucleus basalis in DS.⁷⁰ In the present study, we found dense cytoplasmic p75^NTR multipolar immunoreactive neurons in nbM with well-defined axons and dendritic processes in age-matched non-trisomic controls. In DSD- cases, these neurons were less immunoreactive and displayed a modified multipolar appearance that included poorly defined axons and dendritic processes. By contrast, p75^NTR labeled neurons displayed a globose appearance and lacked cellular processes in DSD+ cases. Although we found a significant decrease in the number of p75^NTR immunoreactive neurons in DSD+ compared to age-matched disomic controls, there was only a non-significant trend of more p75^NTR positive neurons in DSD- compared to DSD+ cases suggesting cognitive resilience to the effect of trisomy. However, despite a significant reduction in p75^NTR labeled nbM neurons, cortical p75^NTR levels were increased in DS cases with dementia. This disconnect maybe due to the de novo appearance of p75^NTR positive cortical neurons in DS⁴⁶ and AD,⁷⁹ which may represent a re-expression of a neuronal phenotype seen in the developing human cortical subplate⁸⁰ suggesting an example of a compensatory response to trisomy. It is also possible that p75^NTR accumulates in the axons of cholinergic neurons, perhaps in axonal fragments or neurites associated with amyloid plaques.⁸¹ Another factor that may play a role in the stable levels of p75^NTR or the lack of changes in TrkA levels found in the FC in DS is the observation that reactive astrocytes and microglia, express both NGF receptors.^82,83,84,85 In fact, triplication of HSA21 contributes to the overexpression of multiple genes associated with pro- and anti-inflammatory processes,^86,87,88 that contribute to an exacerbated inflammatory response in the face of brain injury resulting in an increase in the production of p75^NTR or TrkA in glial cells.^89,90,91 However, whether a dysfunctional cholinotrophic system activates p75^NTR in glial cells remains to be determined in DS.^89,90,91 These findings suggest that reducing cortical p75^NTR levels could aid in CTBF survival in DS.

A recent series of studies suggests that defective metabolic processing of proNGF into mNGF contribute to defective CTBF neuron function in DS and AD.^{92,93,94,95,96} ProNGF, binds with higher affinity to the cell death receptor p75^{NTR,73,97,98,99,100,101,102} , which is dependent upon its interaction with NGF metabolic proteins.¹⁰³ Defects in cortical NGF metabolism include lower levels of tissue plasminogen activator (tPA) and plasminogen, which are involved in the maturation of proNGF, heightened neuroserpin, which is involved in cell growth resulting in higher levels of proNGF, increased activity of matrix metalloproteinase-9 (MMP-9), the main NGF-degrading protease, culminating in greater degradation of mNGF in DS and AD.^45,104 Here, we report that FC levels of proNGF are increased in DSD+ but showed a trend to increase in DSD-cases. Moreover, it was reported that levels of proNGF, MMP-1, MMP-3, and MMP-9 activity are increased at asymptomatic stages of DS from baseline to 1-year and predicted an increase cognitive impairment as determined in a 2 years follow-up evaluation suggesting that cortical NGF metabolic pathways are compromised in the prodromal stages of DS.¹⁰⁵ Interestingly, infusion of Aβ oligomers induces an increase in proNGF levels as well as the synthesis and activation of MMP-9, resulting in a rise in degradation of mNGF in rodents.^106,107 Perhaps over the life span of those with DS, amyloid toxicity contributes to defects in cortical NGF dysmetabolism, an increase in proNGF binding with the p75^NTR, shifting the biologic actions of the NGF system from cell survival to cell death enhancing the selective vulnerability of cholinotrophic neurons within the nbM in DS. The present study did not detect amyloid pathology within the neuropil or neurons within the nbM using Thioflavin S histochemistry, suggesting that cholinergic neuronal cell vulnerability is more closely linked to tau aggregation and NGF dysregulation than to amyloid deposition in this region in DS. However, cortical amyloid plaque deposition may play a role in CTBF cell degeneration via the retrograde transport of a toxic form of amyloid to these neurons.^96,108 These findings support the concept that multi-pharmaceutical approaches are needed to treat NGF pathway dysfunction in DS as well as in AD.

The current study investigated the contribution that tau phosphorylation plays in the development of NFT and NT pathology within CTBF neurons using antibodies against AT8, that mark early-stage and TauC3, a late-stage maker of tau pathogenesis in both DS groups. The cholinotrophic region of the nbM was virtually free of AT8 or TauC3 immunostaining in non-trisomic individuals. By contrast, AT8 positive NFTs were observed in 100% of the DSD+ but only in 60% of DSD-cases. TauC3 labeled neurons were present in 54% of the DSD+ compared to 40% of the DSD-cases. No significant difference was seen in the number of AT8 and TauC3 positive NFTs within the nbM between DS cases, while AT8 NTs were more prevalent in the DSD+ than the DSD- cases like that reported in DSD+ frontal cortex,¹⁵ suggesting an early axonal transport deficit in individuals with DSD+ mimicking that seen in AD.⁴³ Furthermore, we used ThS histochemistry, which labels mature and ghost tangles¹⁰⁹ to determine of the stage of NFTs within the nbM in DS. ThS-bearing NFTs were more abundant in DSD+ than DSD-, but not for either AT8 or TauC3 immunostaining suggesting a more advanced phase of NFT evolution in DSD+ cases. It is interesting to note that AT8 immunostaining has been shown not reveal the true stage of neuronal tau pathology within the nbM in late-onset AD¹¹⁰ indicating the need to confirm NFT staging using additional markers of tau pathology.⁴³ Moreover, the absence of AT8 or MAP2 positivity in ThS-positive tangles, together with a negative correlation between ThS labeled NFTs and number of p75^NTR/ChAT cells in the nbM, also suggests a more advanced tangle phenotype. It is important to note that the p75⁺/AT8⁻/MAP2⁺ phenotype likely represents a healthy state of a CTBF neuron in the nbM, whereas the p75⁺/AT8⁺/MAP2⁺ phenotype reflects a tau-pathological state. The fact that only p75⁺/AT8⁻/MAP2⁺ cells disappear in DSD+ cases does not necessarily imply that they are more vulnerable. To address this question, we used Thioflavin S histochemistry. The DSD+ group displayed a higher number of ghost tangles compared to DSD- cases, suggesting a more advanced stage of tangle evolution. The loss of the p75⁺/AT8⁻/MAP2⁺ phenotype may reflect changes in NGF trophic support prior to the onset of overt tau hyperphosphorylation, eventually leading to cell death and the formation of ghost tangles. Moreover, we found a strong negative correlation observed between the number of p75^NTR -positive and Thioflavin-positive cells, but not AT8-positive NFTs suggesting that when evaluating total tau pathology in the nbM, it is important to consider not only the surviving p75⁺/AT8⁺/MAP2⁻ neurons but also ghost tangles.

In DS, the formation of intracellular NFTs is associated with a frank loss and not a phenotypic downregulation of p75^NTR in nbM cells as seen in prodromal AD⁴³ suggesting a cellular resilience to tau pathology early in the development of CTBF dysfunction during the onset of AD.^44,111,112 What predisposes cholinotrophic tangle pathology to be more advanced in individual DSD+ cases remains unknown. A candidate may be the dual-specificity tyrosine-phosphorylated and regulated kinase 1A (Dyrk1A) gene that is triplicated within the DS critical region.¹¹³ Increased levels of Dyrk1A contribute to neurofibrillary degeneration, in part, through the hyperphosphorylation of tau by activating glycogen synthase kinase-3β (GSK-3β) promoting tau self-aggregation¹¹⁴ or the differential splicing of the tau gene resulting in a greater 3R/4R tau ratio and a greater number of 3R tau containing cortical neurons in DS cases with AD-like dementia compared to sporadic AD.¹¹⁵ Although studies have shown that cortical levels of DykrA1 protein are increased in DS compared to non-trisomic controls,^114,115 our group found no differences in DyrkA1 expression in single FC tau bearing projection neurons¹⁵ or cortical protein levels⁴⁶ between DS with and without dementia. It has been speculated that HSA21 triplication of the Dyrk1A gene, underlies the several-fold increase in the number of Dyrk1A and tau-positive NFTs seen in the nucleus basalis in DS with dementia.¹¹⁶ Dyrk1A overexpression activates NGF mediated PC12 cell differentiation and regulates cell growth,¹¹⁷ suggesting that mutation of this gene enhances cholinotrophic cell dysfunction in DS. Perhaps drugs that stabilize or inhibit Dyrk1A (e.g., epigallocatechin gallate)^118,119 in combination with NGF therapies are potential targets to delay the onset of CTBF neuron dysfunction in DS with translation to AD.

Figure 4 summarizes the differences in the pathophysiology of the cholinotrophic cortical projection system between individuals with DS with and without dementia. Overall, we found a significant loss of p75^NTR and ChAT positive neurons, accompanied by elevated cortical proNGF and p75^NTR in DSD+ cases. ChAT protein levels were reduced in both DS groups, while TrkA did not change within DS groups and controls. Since both proNGF and the p75^NTR are related to cellular dysfunction, this increase suggests a shift toward degeneration away from survival and maintenance of CTBF neurons in DS. Although the number of AT8 or TauC3-positive NFTs within the nbM was not different between DS phenotypes, there was a greater number of AT8 positive NTs and ThS-labeled NFTs in DSD+ cases suggesting increased early axonal damage and more advanced NFT pathology in these individuals. Finally, the lack of a change in FC levels of p75^NTR, and an increase in proNGF levels, less tau bearing NFTs and differences in gene expression in single tau bearing NFTs in the FC in individuals with trisomy that did not develop dementia,¹⁵ suggests biological mechanisms associated with cognitive resilience despite classic AD pathology in people with DS without memory impairments similar to those aged individuals that lived into the sixth^23,120 seventh^{20,121,122,123} and eighth¹²⁴ decades of life with AD pathology. Together these findings suggest that drugs that target TrkA, p75^NTR/proNGF, NGF metabolites, tau phosphorylation and Dyrk1A are potential approaches to slow the degeneration of the cholinotrophic basal cortical memory connectome in DS with translation to AD.

Figure 4.

Summary of changes in the cholinotrophic basal forebrain connectome in DS

Diagrammatic sagittal view of the human brain (A) and a modified stacked bar graph (B) illustrating differences in the pathobiology of the cholinotrophic projection system between non-trisomy age-matched control (AMC), DS without dementia (DSD-) and DS with dementia (DSD+) individuals with DS. Frontal cortex protein levels for ChAT (pink), proNGF (green), p75^NTR (purple) and TrkA (light blue). Nucleus basalis NFTs of ThS (blue), AT8 (orange) and TauC3 (black) and counts of p75^NTR and ChAT (red) neurons. Created with BioRender.com.

Limitations of the study

The main limitation of this investigation is the lack of large numbers of DS individuals with and without dementia and age-matched non-trisomic controls. To help overcome this caveat, we and others developed the Down Syndrome Biobank Consortium with sites throughout the United States, Europe, and Asia⁵⁴ to generate well characterized tissue for distribution to DS researchers around the world. Another limitation was the lack of a TrkA antibody that labeled nbM neurons, despite testing several TrkA antibodies. Unfortunately, the TrkA antibody we used in our earlier investigation of TrkA neurons within the nbM in DS⁷⁰ was not available. It is also important to mention that none of the p75^NTR antibody tested for this study showed a well-defined 75kDa band in our western blots as presented by the manufacturers in cell culture. However, a Santa Cruz antibody produced a strong band, at approximately 100 kDa that corresponds to a p75^NTR receptor that binds to dimeric NGF.^125,126 Others reported a 100 kDa band using the same p75^NTR antibody that was diminished after deglycosylations, which revealed a 75 kDa band.¹²⁷ Although we demonstrated the specificity of this antibody using the corresponding blocking peptide (Santa Cruz Biotechnology, Cat# sc-271708-P) to preabsorb the p75^NTR antibody prior to its application in western blotting and immunostaining of human brain tissue containing the CTBF neurons within the nbM, it possible that the 100 kDa band seen in the protein blots represents a complex between p75^NTR and proNGF. Therefore, the specificity of this p75 antigen for use for western blotting should be viewed conservatively. Despite these limitations, this is the first-of-its-kind investigation of differences in the CTBF system between individuals with DS with and without dementia.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Professor Elliott J. Mufson (elliott.mufson@barrowneuro.org).

Materials availability

This study did not generate unique reagents.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

We thank the participants in Down Syndrome Biobank Consortium, the UCI, and IDIBAPS brain banks. This work was supported by the National Institutes of Health [P01AG14449, RF1AG061566, RF1AG081286, P30AG066511, NS087121, P30AG066519, U19AG068054], BrightFocus Foundation [CA2018010], Barrow Neurological Foundation, and the Arizona Alzheimer's Consortium. The sponsors had no role in studying design, collection, analysis, and interpretation of data, in the writing of the report, and in the decision to submit the article for publication.

Author contributions

Conceptualization: E.J.M.; methodology: M.M.-R. and M.N.; data curation: E.J.M., S.E.P., M.M.-R., and E.H.; investigation: E.J.M., S.E.P., and M.M.-R.; resources: S.E.P. and E.H.; visualization: M.M.R; funding acquisition: E.J.M., S.E.P., and E.H.; supervision: E.J.M. and S.E.P.; writing – original draft: E.J.M. and M.M.-R.; writing – review and editing: E.J.M., S.E.P., and M.M.-R.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
p75NTR	Santa Cruz Biotechnology	Cat# sc-271708; RRID:AB_10714958
ChAT	Proteintech	Cat# 20747-1-AP; RRID:AB_10898169
TrkA	Fitzgerald Industries International	Cat# 20R-TR013; RRID:AB_1289098
proNGF	Alomone Labs	Cat# ANT-005; RRID:AB_2040021
p75NTR peptide pre-absorption	Santa Cruz Biotechnology	Cat#sc271708-P
ChAT	Sigma-Aldrich	Cat# AB144P; RRID:AB_2079751
p75NTR	Millipore	Cat# 07-476; RRID:AB_310649
AT8	Thermo Fisher Scientific	Cat# MN1020; RRID:AB_223647
TauC3	Innovative Research	Cat# AHB0061; RRID:AB_1502094
Biological samples
Healthy adult frontal cortex brain tissue	Banner Sun Health Research Institute (BSHRI)	https://www.bannerhealth.com
Down syndrome adult frontal cortex brain tissue	University of California Irvine Alzheimer Disease Researcher Center (UCI ADRC)	https://medschool.cuanschutz.edu/neurosurgery/research-and-innovation/services/down-syndrome-biobank
Down syndrome and healthy adult frontal cortex brain tissue	Biobanc Hospital Clinic Barcelona-IDIBAPS	https://medschool.cuanschutz.edu/neurosurgery/research-and-innovation/services/down-syndrome-biobank
Deposited data
Raw and analyzed data	This paper
Software and algorithms
GraphPad Prism (version 10)	GraphPad	RRID:SCR_002798
Kodak 1D 3.6	Kodak	https://kodak-1d.software.informer.com/

Experimental model and study participant details

Frozen FC dissected free of white matter and tissue blocks immersion-fixed in 4% neutral buffered paraformaldehyde or 10% formalin containing the CTBF neuronal group within the nbM (Ch4) were obtained from 18 individuals with DS with dementia (DSD+) and 8 without dementia (DSD−). Tissue was obtained from the University of California Irvine Alzheimer Disease Researcher Center (UCI ADRC; 13 DSD+ and 5 DSD−) and Biobank Hospital Clinic Barcelona-IDIBAPS (IDIBAPS: 5 DSD+ and 3 DSD−), members of the DSBC (link: https://medschool.cuanschutz.edu/neurosurgery/researchandinnovation/services/downsyndromebiobank). Eight age-matched controls (AMC) were obtained from the Banner Sun Health Research Institute (BSHRI; n=5) and the Biobank Hospital Clinic Barcelona-IDIBAPS (HCB-IDIBAPS: n=3). Frozen tissue was available for western blot analysis from 5 AMC, 5 DSD-, and 13 DSD+ cases, while fixed tissue for histological studies was available from 8 AMC, 5 DSD-, and 10 DSD+ cases.

Table 1 shows the demographic and neuropathological characteristics of the DS cases examined. The diagnosis of DS was confirmed through the presence of an extra copy of HSA21, identified by fluorescence in situ hybridization and/or chromosome karyotyping. One DSD- case provided by UCI was confirmed as a partial trisomy without triplication of APP gene. Dementia status of UCI ADRC cases was determined based on the International Classification of Diseases and Related Health Problems-Tenth Revision (ICD-10) and the Dementia Questionnaire for Mentally Retarded Persons (DMR-IV-TR) criteria.¹²⁸ All DS individuals in the UCI ADRC cohort were followed until death. Evaluations included comprehensive physical and neurological examinations, participant history, and input from a reliable caregiver. Standardized cognitive and behavioral assessments, both direct and indirect, were also conducted. A dementia diagnosis was made when deficits in at least two cognitive domains were observed, along with a notable decline from the individual's baseline level of cognition. Potential confounders, such as depression, sensory impairments, and hypothyroidism, were excluded. Premorbid intelligence quotient (IQ) was also assessed for all DS cases in the UCI ADRC cohort. Dementia diagnosis for the IDIBAPS cases was determined by a neurologist trained in gerontology and discussion with a caregiver. Virtually all cases were genotyped for ApoE genotype and underwent a postmortem Braak staging evaluation as previously reported.¹⁵ Age-matched non-trisomic controls were obtained from the BSHRI and IDIBAPS. Human Research Committees of UCI ADRC, BSHRI and IDIBAPS approved this study.

Method details

Immunohistochemistry

Two 8-μm-thick paraffin-embedded sections containing the anterior subfields within the nbM were pretreated with citric acid (pH = 6) for 10 minutes for antigen retrieval. Afterward, sections were incubated with primary antibodies against goat anti-ChAT (Sigma-Aldrich Cat# AB144P, RRID:AB_2079751, [1:50]) rabbit anti-p75^NTR (Millipore Cat# 07-476, RRID:AB_310649, [1:100]), mouse anti-AT8 (Thermo Fisher Scientific Cat# MN1020, RRID:AB_223647, [1:500]) and TauC3 (Innovative Research Cat# AHB0061, RRID:AB_1502094, [1:100]) overnight at RT in a tris-buffered saline (TBS)/0.25% Triton X-100/1% goat serum solution. After several washes in TBS, tissues were incubated with a goat anti-rabbit/anti-mouse and horse anti-goat biotinylated secondary antibody, followed by incubation using the Vectastain ABC kit (1 hour) and developed in acetate-imidazole buffer containing 0.05% 3,3-diaminobenzidine tetrahydrochloride (Sigma, MO). The tissue was dehydrated through a graded series of alcohols, and cover-slipped with DPX. Immunohistochemical controls were performed to rule out any cross-reactivity or non-specific staining, including omission of the primary and secondary antibodies and pre-absorption of the p75^NTR antibody. AT8 and TauC3 sections were counterstained with 1% Mayer's hematoxylin, as previously described.¹²⁹

Immunofluorescence and Thioflavin-S histochemistry

8-μm-thick paraffin-embedded sections from the same cases (n=5 per group) used for single p75^NTR and ChAT immunostaining, were triple labeled for p75^NTR, AT8 and MAP2 immunofluorescence and Thioflavin-S (ThS) histochemistry. Sections were washed 3× for 10 minutes each in PBS, TBS, and TBS/Triton (0.25%), blocked in TBS/Triton/3% donkey serum for 1 hour at RT, and incubated overnight in TBS/Triton/1% donkey serum containing rabbit anti-p75^NTR (Millipore Cat# 07-476, RRID:AB_310649, [1:50]), guinea pig anti-MAP2 (Synaptic Systems Cat# 188 004, RRID:AB_2138181, [1:50]) and mouse anti-AT8 (Thermo Fisher Scientific Cat# MN1020, RRID:AB_223647, [1:100]), then washed in TBS/1% donkey serum, and incubated in same serum containing the appropriate secondary antibody for 1 hour at RT. Afterwards, sections were washed and defatted in equal parts chloroform and 100% ethanol (1 hour), rehydrated through graded alcohols (10 seconds), placed in distilled H₂O (5 minutes), placed in a 0.02% ThS solution in the dark (20 minutes), differentiated in 80% ethanol (30 minutes), and incubated with an autofluorescence eliminator reagent to eliminate lipofuscin autofluorescence (Millipore, Temecula, CA), and cover-slipped with Aqua-Mount (Lerner, Cheshire, WA).

Western blotting

FC levels of ChAT, p75^NTR, TrkA and proNGF were measured in 5 AMCs, 5 DSD- and 13 DSD+ samples. Briefly, frozen samples free of white matter were homogenized (150 mg/mL) in phosphate buffer containing protease inhibitors (Sigma, St. Louis, MO) and denatured in SDS loading buffer to a final concentration of 5 mg/ml. Proteins (50 μg/sample) were separated by SDS-PAGE (Lonza, Rockland, ME) and electrophoretically transferred to polyvinylidene fluoride membranes (Millipore, Burlington, MA). Membranes were blocked in tris-buffered saline/0.05% Tween-20/5% milk (1 h) at room temperature (RT). Antibodies were added to blocking buffer and membranes incubated overnight (4°C) in mouse anti-p75^NTR ([1:150], Santa Cruz Biotechnology Cat# sc-271708, RRID:AB_10714958), rabbit anti-ChAT ([1:500] (Proteintech Cat# 20747-1-AP, RRID:AB_10898169), rabbit anti-TrkA ([1:150], Fitzgerald Industries International Cat# 20R-TR013, RRID:AB_1289098) or rabbit anti-proNGF ([1:250], Alomone Labs Cat# ANT-005, RRID:AB_2040021), washed, incubated with horseradish peroxidase-conjugated with either a goat anti-mouse IgG (1:200) or goat anti-rabbit IgG (1:200) secondary antibody at RT, visualized by chemiluminescence (Kodak Image Station 440CF; Perkin-Elmer, Wellesley, MA), and quantified with Kodak 1. Four different proteins, each with a separate kDa, as well as tubulin were developed on the same blot (Figure S1). All experiments were run in triplicate using samples from the same tissue aliquots. Protein bands were quantified and normalized to β-tubulin ([1:5000], Millipore Cat# MAB3408, RRID: AB_94650). Regarding the p75^NTR antibody, we tested several commercially available antibodies from different host species that targeted various epitopes of this protein (Abcam Cat# ab38335, RRID: AB_2152644 and Abcam Cat# ab227509, RRID: AB_2892530, Sigma-Aldrich Cat# N3908, RRID: AB_260763 and Millipore Cat# 07-476, RRID: AB_310649, and Cell Signaling Cat# 8238S). Each antibody failed to detect the expected 75 kDa band, as previously reported for these antibodies.^127,130 Based upon numerous preliminary experiments, we chose an antibody that produced a strong band, at approximately 100 kDa (Santa Cruz Biotechnology Cat# sc-271708, RRID: AB_10714958) that likely corresponds to a p75^NTR receptor that binds to dimeric NGF.^125,126 A previous western blot study using this same p75 antibody reported a 100 kDa band that was attributed to a highly glycosylated form of p75,¹²⁷ suggesting antibody specificity. An alternative interpretation is that the 100 kDa band indicates that the antibody recognized a complex with proNGF. We demonstrated antibody specificity of the Santa Cruz Biotechnology p75^NTR antibody (Cat#sc271708-P) by peptide pre-absorption (1:15) prior to immunoblotting that blocked p75^NTR labeling (Figures S2A and S2B). Photos showing cholinotrophic neurons in the nbM immunostained with the p75^NTR Santa Cruz Biotechnology antibody (Figure S2C). The arrow in C marks the cell shown at higher magnification in the boxed area. (Figure S2C). Example showing raw images of complete immunoblots for ChAT (Figure S2D), TrkA (Figure S2E), and proNGF (Figure S2F). Immunoblots were adjusted for brightness in power point.

Quantification and statistical analysis

Quantitation of single and multi-labeled CTBF neurons within the nbM in down syndrome

Counts for single p75^NTR, ChAT, AT8, and TauC3 immunoreactive structures were performed at 40x magnification on consecutive sections containing anterior and lateral subfields of the nbM.^{131,132,133,134} Three non-overlapping regions of interest were analyzed per slide in two adjacent sections. Fiduciary landmarks were identified to avert analyzing overlapping cells within the nbM. All images and cell counts were performed using a Nikon Eclipse 80i microscope coupled with NIS-Elements Imaging software (Nikon, NY). We also counted p75^NTR positive neurons that were either AT8 positive or AT8 negative, along with MAP2 positive or MAP2 negative. Counts of AT8/ThS, dual-label cells and percentages were calculated against the total number of AT8-positive NFTs visualized with the aid of a Revolve Fluorescent Microscope (Echo Laboratories, CA) with excitation filters (wavelengths 405, 489, 555 and 681 nm for DAPI and ThS (emission blue), Cy2 (pseudocolor green; p75^NTR), Cy3 (pseudocolor cyan; AT8), and Cy5 (pseudocolor red; MAP2), respectively, as described previously.¹³⁵

Statistical analysis

Data were evaluated across clinical groups using the Mann–Whitney U test, Fisher’s exact test, and Kruskal–Wallis test, followed by Dunn’s post hoc test for multiple comparisons. Spearman’s rank correlation was used to assess associations, with statistical significance set at p < 0.005. For all other analyses, significance was defined as p < 0.05 (two-tailed). All statistical analyses and graphical representations were performed using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA).

Additional resources

This work is not part of or involves a clinical trial.

Published: July 1, 2025