The Effect of Human HSP70 Administration on a Mouse Model of Alzheimer’s Disease Strongly Depends on Transgenicity and Age

Abstract

In humans, heat shock protein 70 is a key component of the machinery that protects neuronal cells from various stress conditions and whose production significantly declines during aging. Herein, we investigated the protective effect of sub-chronic intranasal administration of human Hsp70 on the state of neurons in the temporal cortex and areas of the hippocampus of old transgenic (Tg) 5XFAD mice (11–13 months), representing a late-onset model of hereditary Alzheimer’s disease. Quantitative analysis of the various neuronal pathologies between the two groups (Tg versus nTg) revealed maximal levels of abnormalities in the brains of aged Tg mice. Importantly, intranasal application of HSP70 had profound beneficial effects on neuron morphology in the temporal cortex and hippocampal regions when applied to the aged Tg mice but not in the case of age-matched, non-transgenic, littermate animals. Furthermore, the effect of HSP70 administration on neurons in the hippocampus and temporal cortex differed characteristically between the groups. Using RNA-Seq, we identified a lot of differentially expressed genes in the hippocampus of old Tg mice compared with those of nTg mice. Most importantly, we observed HSP70-induced upregulation of multiple genes participating in antigen processing and presentation especially the members of major histocompatibility complex (class I and II) in the brains of old 5XFAD Tg animals, suggesting that Hsp70 executes its beneficial role via activation of adaptive immunity. Overall, our data enable to conclude that Hsp70 treatment may be a safe and effective therapeutic application against Alzheimer-type neuropathologies manifested at the late stages of the disease.

Keywords

5XFAD Aging hippocampus neuronal pathology recombinant HSP70 temporal cortex transcriptome

INTRODUCTION

At the present time, investigation of heat shock proteins (HSPs), especially HSP70, is attracting the attention of many researchers from different branches of knowledge, including medicine. While normally HSP70 is an endogenous intracellular protein with many different functions in the cell, it may exit the cell and behave as chaperokine [1 –4]. Hsp70 protects cells from death induced by various stimuli and inhibits different cell death pathways. Hsp70 plays an important protective role in neurodegenerative disorders, including but not limited to Alzheimer’s disease (AD), Parkinson’s disease, Huntington’s disease, and prion diseases [4 –7]. Hsp70 is also involved in ameliorating many disturbances associated with the normal aging process [2 , 9]. The chaperone activities of Hsp70 include preventing aggregation of denatured proteins and active participation in scavenging of toxic protein aggregates to the sites of their clearance [1 , 11]. In the CNS, HSP70 apparently plays an important protective role, although the exact mechanisms underlying such protection are not exactly known [12]. It is evident that inducible HSP70 stimulates multiple systems that are responsible for cell survival. Indirect effects of HSP70, such as suppression of apoptosis, lysosome stabilization [13], stimulation of the immune response [13, 14], suppression of the early pre-oligomeric stages of Aβ self-assembly [15, 16] and tau protein hyperphosphorylation, inhibition of pro-inflammatory signaling [17], as well as an increase in the survival of endogenous neural progenitor cells, may account for the observed protection of the CNS in both the case of neurodegeneration (e.g., AD) and the normal aging process [2, 18].

It was demonstrated in different animal models that molecular chaperones play important roles in protection from protein damage during aging and that normal Hsps expression is required for longevity [1]. There are multiple observations indicating that higher basal levels of HSP70s expression are characteristic of longer-lived species [19]. Therefore, various stress response genes (often called “vitagenes”), particularly HSP70, are now major candidates in gene-longevity association studies. Vitagenes, which include HSP70, confer protection against oxidative stress and neuronal inflammation, which play a major role in many age-related pathologies, including AD [18, 20].

It was shown that chaperone activity and the rate of HSP synthesis decreases as people age [1, 21]. A decrease in the expression of HSP70 has been repeatedly documented in the aged brain [22]. While it is well known that major symptoms of sporadic and familial AD mainly affect the elderly, the particular characteristics of the human aged brain that facilitate the development of AD are not known. It is also clear that these particular consequences of the normal aging process facilitate the development of AD and represent important targets for pharmacological intervention to prevent or delay the progression of this widespread neuropathology [6 , 24].

Along these lines, we previously demonstrated in wild-type (wt) mice that chronic intranasal administration of exogenous human HSP70 enhanced the lifespan of different age groups and decreased the accumulation of age-related markers such as lipofuscin in old mice [9]. Similarly, chronic intranasal administration of HSP70 exerts dramatic neuroprotective effects in two complementary mouse models of Alzheimer-type neurodegeneration [3]. One of these models is the 5XFAD mouse model of amyloid deposition expressing five familial AD (FAD) mutations that are additive in driving human amyloid-β (Aβ)₄₂ overproduction [25, 26]. Furthermore, we previously performed quantitative analysis of neuronal pathologies in young 5XFAD mice using transgenic (Tg) versus non-transgenic (nTg) littermate controls by deep sequencing and demonstrated that HSP70 administration modulated the spectrum of DEGs in Tg animals. Thus, the treatment resulted in upregulation of genes responsible for amine transport, transmission of nerve impulses, and other important pathways usually impaired in AD patients. Importantly, in the previous investigations, we used young male transgenic 5XFAD mice (2.5 months old), representing an early-onset model of hereditary AD [7].

It was also previously demonstrated by another group studying 5XFAD mice that various manifestations of AD-like degeneration, such as intraneuronal Aβ₄₂ accumulation, Caspase-3 activation, neuronal loss, and cognitive disturbances, occur in a temporal order in the transgenic 5XFAD animals [25]. Consequently, 5XFAD mice represent a very appropriate model of AD-type neurodegeneration, which includes neuron loss and the early appearance of intraneuronal Aβ₄₂, making this model a robust tool to investigate the mechanisms underlying the well-documented neuroprotective role of Hsp70 and other chaperones.

It was a challenge to compare the effects of sub-chronic treatment with recombinant HSP70 in 5XFAD mice sampled at different ages. Thus, using old 5XFAD mice (11–13 months old), we performed histological analysis of the temporal cortex and hippocampal areas representing the main brain regions responsible for learning and memory. In addition, we carried out transcriptome analysis of the hippocampal regions of these 5XFAD mice that represent a late-onset model of AD. The obtained results were compared with our previous data in young animals from the same line [7]. The analysis presented herein demonstrated that HSP70 treatment resulted in a clear-cut beneficial effect in Tg mice of both age groups but differently affected the brain regions of young and aged Tg and nTg animals at both the histological and transcriptional levels. Taken together, our studies revealed possible targets for recombinant HSP70 intervention and provide evidence that sub-chronic intranasal administration of recombinant Hsp70 may have a significant therapeutic potential to cure or delay neurodegeneration in middle-aged and old people.

MATERIALS AND METHODS

Animals and treatment

Adult male 5XFAD mice (11-13 months old) were used in the experiments and treated with Hsp70 (4μg/mouse in 4μl, intranasal injections) daily for 1.5 months. The mice were divided into the following four groups: Tg+NaCl (n = 5), Tg+Hsp70 (n = 4), nTg+NaCl (n = 9), and nTg+Hsp70 (n = 4). The same dose and treatments were used in our previous studies [7, 9]. The maintenance protocol for the animals as well as the description of 5XFAD mice were previously provided [7]. Mice were genotyped by PCR analysis of DNA extracted from ear biopsies as previously described. All procedures involving animals were reviewed and approved by the Animal Care and Use Committee Branch of the Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry. The procedures corresponded to principles and procedures approved by the NIH USA Committee for Animal Care and Experimental Use (NIH Publications N 8023, revised 1978).

Brain tissue histology and isolation of human recombinant Hsp70

Brain tissue preparation and histology as well as preparation of human recombinant HSP70 were performed essentially as described in our previous study of 5XFAD mice [7]. Briefly, to investigate morphology of neurons from the temporal cortex and areas of the hippocampus, 15μm-thick sections were subjected to Nissl staining with cresyl violet acetate (Sigma) and inspected under a Nikon Eclipse E200 optical microscope. The form and size of neurons, as well as intensity of their staining, were estimated. Only neurons with clearly seen contours, nucleus, and nucleoli were taken into account. One thousand cells from each area of each animal were analyzed. A PDP-12 digitizer computer system (Germany) was used for a comparative analysis of cellular compositions of the temporal cortex and CA1 and CA3 areas of the hippocampus. The different kinds of the pathology of the neuronal morphology were associated with the severe dystrophycal processes: pyknosis, karyolysis, cytolysis, and vacuolization. The pyknomorphic neurons have shrunken nuclei with chromatosis, fragmentation, and conglutination of the blocks of the tigroid. The last stage of cell destruction has confined angularity or mace-like particles. Karyolysis in neurons was characterized by eroding of the nuclear and nucleolus border and a hypertrophy of the nucleolus and enlightenment of the nucleolus space and full or partial loss of the nuclear structure. Cytolysis makes apparent the octopi of the nucleus and the disappearance of the contour of the cells, the lysis of the tigroid. The vacuolization in the neurons appeared as the alveolar structure with the increased nucleus and nucleolus and confluent lacune in the neuronal cytoplasm. The healthy neurons, which were called “normal” neurons, were characterized by moderately stained cytoplasm with evenly distributed tigroid, their centrally positioned nuclei were light in color with easily distinguishable nucleoli. The frequency (%) of each neuronal abnormality versus the quantity of healthy neurons was estimated. The neuron density was determined per mm². Neurons were counted in 10 view fields (x20). The details are given in the Supplementary Material.

Preparation and sequencing of mRNA libraries

Total RNA from hippocampi of 5XFAD male mice was isolated and analyzed essentially as described in our previous study [7]. The experimental details of preparation and sequencing of mRNA libraries are given in the Supplementary Material. The reads are deposited in NCBI Sequence Read Archive, Bioproject ID PRJNA482728.

Table 1

The frequency of various neuronal pathologies in the brain regions of Hsp70 treated and untreated old 5XFAD mice

Brain areas	Groups	The proportion of normal and pathological neurons from the total number (%)					Neuronal density (n/mm²)
		Pyknosis	Karyolysis	Cytolysis	Vacuolization	Normal neurons
Temporal cortex	nTg NaCl (n = 9)	1.49±0.12	6.12±0.29	14.6±0.43	3.7±0.17	74.06±0.67	1112.7±9.1
	nTg Hsp70 (n = 4)	2.75±0.29^***	6.45±0.34	18.73±0.59^***	4.68±0.23^***	67.4±0.67^***	1007.3±11.1^***
	Tg NaCl (n = 5)	1.82±0.22	8.22±0.33^***	19.16±0.45^***	5.04±0.24^***	65.9±0.94^***	1111.2±11.1
	Tg Hsp70 (n = 4)	1.23±0.18	5.77±0.33⁰⁰⁰	17.9±0.86^***	2.9±0.2^*000	72.2±1.03⁰⁰⁰	1130.1±14.5
CA1	nTg NaCl(n = 9)	1.55±0.13	5.86±0.23	14.6±0.42	3.42±0.15	74.4±0.67	2776.2±12.4
	nTg Hsp70(n = 4)	1.85±0.25	5.53±0.41	16.93±0.63^***	3.38±0.18	72.35±0.99	2773.1±12.2
	Tg NaCl (n = 5)	1.3±0.15	6.28±0.19	17.71±0.47^***	4.5±0.24^***	70.36±0.37^***	2827.0±18.1^*
	Tg Hsp70 (n = 4)	1.17±0.17	4.9±0.26^*00	16.87±0.62^**	3.03±0.26 ⁰⁰⁰	74.03±0.75⁰⁰	2783.4±24.09
CA3	nTg NaCl (n = 9)	1.59±0.12	6.1±0.29	15.241±0.35	4.1±0.15	73.1±0.72	1978.8±10.0
	nTg Hsp70 (n = 4)	1.2±0.14	5.63±0.22	15.58±0.52	4.8±0.26^**	72.8±0.69	2083.1±10.7^***
	Tg NaCl (n = 5)	1.56±0.17	7.46±0.42^**	17.22±0.64^**	4.52±0.26	68.04±1.63^***	1975.3±16.3
	Tg Hsp70 (n = 4)	1.57±0.23	5.67±0.23⁰⁰	16.73±0.55	4.2±0.23	71.8±0.86⁰	2012.3±16.0

^*significance of the observed differences in relation to nTg group: ^*p < 0.05; ^**p < 0.01; ^***p < 0.001. ⁰significance of the observed differences in relation to Tg group: ⁰p < 0.05; ⁰⁰p < 0.01; ⁰⁰⁰p < 0.001. Data are given as mean±SEM. Comparison between the groups were performed separately for the temporal cortex and both areas of the hippocampus with two-tail Student’s t-test to compare replicate means with special emphasis on the effects of Hsp70 treatment in Tg and nTg groups.

Bioinformatics analysis

Raw sequence data processing (QC, trimming, alignment, read quantification) was performed with PPLine tool [27]. To attenuate the effect of different RINs we quantified reads in the 750 bp 3’-tail regions of transcripts as described [28]. Additionally, we split all the genes into 10 bins depending on the average transcript length, and, within each bin, we normalized the read counts using TMM (trimmed mean of M-values) method from edgeR [29]. The same procedure was performed to eliminate the dependency of FC on the ‘absolute’ gene expression level. Differential gene expression analysis was performed using edgeR package [29]. Gene Ontology, KEGG, Reactome enrichment analyses were performed using topGO and clusterProfiler packages. In the present work, we also included and re-analyzed our previous RNA-Seq data derived for young 5XFAD Tg mice [7]. For details, see the Supplementary Material.

RESULTS

Similarities and differences in the neuronal morphology during the aging process of Tg and nTg mice

In our previous studies, we demonstrated that human HSP70 administration protects spatial memory in two independent mouse models of AD and ameliorates neuronal pathologies occurring in the course of normal aging [3, 7]. While we are well aware that many promising findings in mouse models of AD are difficult to replicate in humans, we think that the comparative analysis of sub-chronic injections of HSP70 offer some tantalizing possibilities in different age-dependent models of neurodegeneration. With this in mind, we investigated the morphological state and density of neurons in the temporal cortex and hippocampal CA1 and CA3 areas of old 5XFAD mice with the goal of comparing the results with our previous data obtained from 5XFAD mice at early stages of pathology progression [7]. Coronal brain sections of 5XFAD Tg mice and nTg littermate controls were used for analysis of neurons. We investigated the state of neurons in brain regions that are responsible for learning and memory in old Tg and nTg 5XFAD mice, depending on HSP70 treatment. The monitored neuronal pathologies included pyknosis, karyolysis, cytolysis, and vacuolization (Table 1). The illustrations of the mentioned pathologies and normal healthy neurons have been previously published [3, 7]. Additionally, neuron density and the proportions of healthy (normal) neurons and cells with different kinds of pathology were estimated. Supplementary Table 1 includes the results of our previous studies of young 5XFAD mice [7], and Supplementary Table 2 summarizes the results accumulated in the process of analysis of neuron morphology and density in the old Tg and nTg mice with subchronic treatment HSP70 or saline. The p values of pairwise differences obtained in the nonparametric Mann-Whitney U Test for experimental groups, differed by age, genotype, or medicinal effect are presented in Supplementary Table 3.

To illustrate the effects of aging and HSP70 administration on the morphology of neurons, we compared the parameters indicated above in pairs of groups representing maximal interest in terms of dependence on HSP70 treatment and transgenicity. Such an approach enables the presentation of all the major data in a graphical form (Fig. 1) and monitoring of different brain regions in treated and naïve Tg and nTg littermates from both age groups.

Fig.1

Graphical illustration of the observed differences in neuron density and morphology in the compared groups of mice, depending on transgenicity and Hsp70 treatment. Heatmap shows the relative changes in neuronal characteristics in Tg and nTg mice. Cell color is green if the parameter is improved (positive changes) and red if deterioration of the parameter is detected (negative changes).

In the heatmap (Fig. 1), the compared pairs of groups are depicted in seven columns (1–7). In the columns the positive or negative changes of different morphological characteristics in compared pairs are indicated by different colors. It is evident that with age in Tg and nTg groups (Fig. 1.1: old nTg versus young nTg and Fig. 1.2: old Tg versus young Tg), the extent of various neuronal pathologies is increased in the hippocampus, especially in the CA3 region. Specifically, old Tg mice exhibit a decrease in the proportion of normal neurons in all regions studied (Fig. 2A), accompanied by enhanced cytolysis and vacuolization in the cortex and both studied hippocampal areas (Fig. 1.2). Surprisingly, old nTg mice are characterized by improvement of the cortex which includes the decrease in pyknosis and vacuolization (Fig. 1.1). Another interesting feature common for old Tg and nTg mice is an increase in the neuronal density in the temporal cortex in comparison with that in young mice of the same genotype, which may be explained by migration of neuronal progenitor cells from certain brain niches of adult mice to the cortex. However, these presumptive migrating progenitor cells apparently preserve the population of neurons with normal structure in the cortex of only the old nTg mice, while age-matched Tg animals are characterized by an increase in the proportion of neuronal cells with various pathologies especially in the cortex (Fig. 1.2). The comparison of old Tg mice with the age-matched nTg group (Fig. 1.3), as expected, revealed structural deterioration of all studied regions. It is of note that despite such deterioration, old Tg mice do not differ from age-matched nTg animals in terms of neuronal density in all brain regions investigated (Fig. 1.3 and Fig. 2B).

Fig.2

The effect of intranasal administration of Hsp70 on the proportion of normal neurons from total number cells counted in % (a) and neuron density (b) in the temporal cortex and two areas of the hippocampus in old 5XFAD mice. Square brackets above the columns indicate which group is compared with.

Hsp70 treatment exerts beneficial effect on neurons of old Tg mice

Keeping in mind different trends observed in the morphology and density of neurons in Tg and nTg mice that occur in the course of aging process, we decided to compare the effect of sub-chronic HSP70 intranasal injections in the groups of mice of different age. Mann-Whitney U Test revealed significant effects of HSP70 on the majority of the traits examined as its own main effect or in interaction with the genotype. Hence, we carried out FDR test based on p-values of all compared groups for all morphological parameters and independently for each studied brain region (see Supplementary Table 3).

HSP70 treatment results in significant beneficial effects for most of the pathologies assessed and importantly increases the proportion of healthy (normal) neurons in all investigated brain regions in the old Tg mice (Fig. 1.4 and 1.6; Fig. 2B). While as expected, the comparison of untreated Tg and nTg old mice reveals deterioration of neurons in studied brain areas in the Tg group (Fig. 1.3), we did not find a significant effect of HSP70 treatment on overall neuronal density in any of the brain structures in the old Tg mice. This implies that the neuroprotective effect of HSP70 treatment is probably not due to activation of neurogenesis (Fig. 1.4 and 1.6; Fig. 2B). To better evaluate the effect of HSP70 treatment on the brain of old Tg mice, we compared the neuronal state of the treated Tg mice with that of the age-matched untreated nTg mice (Fig. 1.4). The comparison indicates that HSP70 treatment improves most vital parameters of old Tg mice and makes the proportion of health neurons and neuronal density in this group of animals equal to the correspondent parameters of old nTg mice (Fig. 1.4). Therefore, HSP70 treatment apparently slows down the process of pathological changes of neurons in Tg mice (compare Fig. 1.3 with 1.4; Fig. 2A; Supplementary Tables 2 and 3).

The effect of HSP70 treatment on neurons of old nTg mice is not as straightforward and does not seem to improve neuronal parameters. Characteristically, HSP70 treatment administered to old nTg mice does not show any beneficial effects (CA3 region being an exception) but leads to deterioration in terms of most criteria in the temporal cortex (Fig. 1.5), while in the hippocampal areas, most of the parameters are not affected. Interestingly, HSP70 treatment of old nTg mice leads to an increase in neuronal density of the CA3 area and simultaneously annihilates the improvement in the morpho-functional state of cortical neurons observed during the course of aging in nTg mice (compare Fig. 1.5 and 1.1). Importantly, the comparison of neuronal characteristics in old nTg+HSP70 group with those of young nontreated nTg (Fig. 1.7) enables to conclude that most of the detected negative changes due to HSP70 treatment in the cortex make old nTg animals more similar to the young nTg mice (compare Fig. 1.7 and 1.5), i.e., treatment with HSP70 delays manifestations of aging process in the cortex in nTg mice and in all studied brain structures in Tg mice (compare Fig. 1.7 and 1.4).

Thus, the above data illustrate the severity of neuronal dysfunction caused by transgenicity in older animals that did not receive HSP70 and clear-cut positive effect of HSP70 therapy on the survival and function of neurons even at the late stage of neurodegeneration process.

Transcriptomic changes in hippocampus depend on age and transgenicity of 5XFAD mice

In total, 13,800–14,050 genes have passed expression level threshold (in terms of CPM, read counts per million) in various comparisons (Tg versus nTg, HSP versus NaCl).

According to the results of RNA-Seq transcriptomic profiling, the major effect of transgenicity is the activation of immune response and overexpression of more than one hundred immune-related genes in both young and old animals, irrespectively of HSP treatment. This is illustrated with GO-centric gene expression profiles (Fig. 3) and with Venn diagram (Fig. 4). These genes participate in B-cell (and at lesser extent – T-cell) receptor pathways, phagocytosis, lysosome-related processes, antigen processing and presentation, cell chemotaxis, and many other immune-related activities. The lists of immune-related overexpressed genes significantly overlap between old and young mice and include the same genes (131 common genes that passed FDR threshold), but the degree of their increase is much higher in older mice.

Fig.3

Transcriptomic effect of transgenicity: expression level changes (Tg versus nTg) of genes participating various KEGG pathways and Gene Ontology (GO) terms that are most affected by transgenicity. In each cell, the sorted binary logarithms of expression level fold changes (LogFC) are shown. LogFC range is from –2 (i.e., 4-fold downregulation; blue) to +2 (i.e., 4-fold overexpression; red). Cell borders indicate the gene set enrichment (GSEA) p-value for a term/pathway. ‘min. p (up/down)’–minimal GSEA p-value for up/down-regulated genes (across all the analyses). ‘N genes’–the number of genes in a pathway/term.

Fig.4

Venn diagram comparing differentially expressed genes (DEG) between Tg and nTg mice in three subgroups: young (yellow, total 249 DEGs), old NaCl (orange, total 2849 DEGs), old HSP (blue, total 2206 DEGs). We included genes with p < 0.01 and expression level changes greater than 1.5-fold (both up- or downregulated in Tg mice). For each gene subset in the diagram, Gene Ontology (GO) enrichment analysis has been performed. Enriched GO terms are presented with colored text. Font size is proportional to the number of genes in a term (see the ‘font size legend’). Font color indicates the enrichment test FDR (blue–dark green–light green color scale; see the ‘font color legend’).

Besides the set of 131 genes that are differentially expressed (between Tg and nTg) in all three subgroups, both old HSP70 treated and old mice treated with saline (NaCl) demonstrated expression level changes of 1413 genes, that are not altered in young animals. This gene set is slightly enriched with participants of mRNA processing and regulation of apoptotic process (Fig. 4). Interestingly, 1283 genes are differentially expressed (between Tg and nTg) only in old mice that did not receive HSP70. The expression of these genes is not significantly changed (between Tg and nTg) in either old HSP70-treated or young animals (Fig. 4). This gene set is enriched with regulators of neurogenesis and neuron differentiation, gliogenesis, and axon regeneration. Such transcriptomic pattern reflects more severe neurodegeneration process taking place in old Tg animals that did not receive HSP therapy and illustrates apparent beneficial rejuvenating effect of HSP70 treatment.

This is also shown in Fig. 3. In old mice that did not receive HSP70 (but neither young nor HSP-treated old animals), transgenicity is followed by predominant downregulation of genes involved in the formation and function of glutamatergic, GABAergic, cholinergic, dopaminergic and serotonergic synapses; long-term depression and potentiation; calcium signaling. This observed effect may be the result of the reduced density of healthy neurons in old Tg NaCl mice.

Unlike the transgenicity status, sub-chronic HSP70 administration itself does not result in massive and drastic alterations of gene expression profiles. It is important to note that HSP70 administration does not attenuate the Tg-induced activation of immune response in brain tissues, but even significantly increases the expression of several genes involved in antigen processing and presentation, especially the members of major histocompatibility complex (MHC) class I and II (Fig. 3). This is also illustrated with the heatmap (Fig. 5). It is known that antigen processing and presentation provides a link between innate and adaptive immunity. Besides the members of MHC complexes, we observed slight upregulation of two genes encoding catalytic subunits of immunoproteasomes, Psmb8 and Psmb9. Therefore, heat shock protein therapy may result in the reinforcement of antigen processing and presentation via MHC of both classes. This effect is evident in both young and old Tg mice, but is more pronounced in the old animals. Characteristically, this effect of HSP70 treatment is lacking in non-transgenic mice (Fig. 5).

Fig.5

Heatmap showing the relative expression levels of genes involved in antigen processing and presentation (top 25 differentially expressed genes). Cell color is blue if the gene is expressed at a level below the average, and orange indicates above average. CPM column indicates average read counts per million (‘absolute’ transcript abundance). LogFC columns show binary logarithms of expression level fold changes in HSP-treated mice compared to the placebo group (orange – upregulation, blue – downregulation). As seen from the figure, Hsp70 administration is typically followed with upregulation of genes encoding members of the major histocompatibility complex predominantly in transgenic animals. These genes are also overexpressed in old Tg mice compared to old nTg animals.

The significant transcriptomic changes in HSP70-treated Tg mice were also detected for the genes involved in amine synthesis and transport, including transmission of nerve impulse. However, this transcriptomic response is mosaic and non-consistent, i.e., not all HSP70-treated Tg mice exhibited such changes. Thus, the significant upregulation of Th, Avp, Agt, Ddc genes, that are included in ‘amine transport’ GO term, is observed in 3 of 5 young HSP70-treated Tg mice and 2 of 4 old nTg HSP70-treated mice. A similar observation was made regarding several genes involved in neurotransmitter secretion and neuropeptide signaling pathways, including Cartpt, Pomc, Penk, Ecel1, Glra2, Glra3. However, this seems to be downstream effect of HSP70 therapy, whereas the major mechanism of Hsp70 protective action is probably realized by enhancing adaptive immune response observed in Tg mice.

We are well aware that hippocampus is composed of various cell types in different proportions and the data obtained from the transcriptome analysis is a mixture of cellular responses but since both hippocampal areas (CA1 and CA3) responded similarly (Fig. 1) to HSP70 treatment the transcriptome analysis probably resembles the response of the hippocampus as a whole.

DISCUSSION

The main goal of our investigation was to compare the effects of sub-chronic treatment of 5XFAD mice of different ages with human recombinant HSP70 and, in parallel, monitor the age-related morphology of neurons and transcriptional changes in brain regions in young and old Tg and nTg mice. The results were compared with the previously accumulated data obtained in young animals of the same strain (2.5-month-old 5XFAD mice), representing an early-onset model of hereditary AD [7].

As expected, in old Tg and nTg mice, the extent of various neuronal pathologies is increased, especially in the CA3 region of the hippocampus, in comparison to young animals (Fig. 1.1 and 1.2). However, old Tg mice, in comparison with the age-matched nTg group (Fig. 1.4), exhibit more severe structural deterioration of all studied regions. Surprisingly, old nTg mice exhibit some improvement in the neuronal parameters of the cortex in comparison with young nTg mice (Fig. 1.1). This improvement probably represents a compensatory reaction to the deterioration in both areas of the hippocampus observed in old nTg animals (Fig. 1.1). Notably, progression of memory deficits with aging in Tg 5XFAD mice in comparison to nTg animals of age control was measured using fear conditioning, Y/T-maze alternation, Morris water maze, and other approaches [30 –32].

Another interesting feature common for old Tg and nTg mice is the observed increase in the neuronal density in the temporal cortex in comparison with young mice of the same genotype (Fig. 1.1 and 1.2). Density can be different due to the changed volume of parenchyma, changes in white matter, altered vascular bed, or if the dynamics of loss of neurons is slower than that of the volume, particularly it may be in the case of AD. However, we detected an equal increase in the neuronal density only in the temporal cortex but not in hippocampus in both Tg and nTg old mice. Therefore, we speculate that presumptive activation of neurogenesis probably takes place at a certain stage of the aging process in both compared groups, while the output of this probably compensatory phenomenon is quite different in Tg and nTg mice. Activation of neurogenesis during the aging process represents the most plausible explanation for the observed phenomenon of increase of neuron density in the cortex. This observation strengthens the concept of brain reserve, which states that the brain is able to manage or compensate for age-related changes or pathologies occurring in one structure (in the hippocampus in our case). While it is widely accepted that the subventricular zone of the lateral ventricles and subgranular zone of dentate gyrus of hippocampus are the classical areas of neurogenesis in the adult brain, there are a lot of data suggesting that neurogenesis may occur in other brain areas, including the cortex, especially in the case of pathology [33 –35]. It is known that in certain cases, for example, focal ischemia, many of neuronal progenitors migrate in chains toward the ischemic striatum or to the hippocampal pyramidal cell layer in the case of an injured hippocampus [36, 37].

Therefore, the source of the new neurons in the cortex may be due either to redifferentiation of special astroglial cells into new neurons or to migration of neuronal precursor cells from canonical brain niche of adult neurogenesis in subventricular zone. Along these lines, another group studying the same model (5XFAD) demonstrated that at the end of the 5-6-month time period, some of both Tg and nTg mice exhibited certain increases in the quantity of large pyramidal neurons in cortical layer 5 [26]. However, these presumptive migrating progenitor cells apparently preserve the population of neurons with a normal structure in the cortex only in old nTg mice, while age matched Tg animals are characterized by an increase in neuronal cells with various pathologies in all studied brain regions (Fig. 1.2). This finding is not surprising since 5XFAD mice exhibit not only neuronal loss but dendrite degeneration and synapse loss as well [25], leading to dramatic deterioration of neuron morphology in the old Tg animals (Fig. 1.2).

Importantly, HSP70 treatment increased the proportion of normal neurons in all investigated brain regions in old Tg mice (Figs. 1.6 and 2A) and made them equal to the age-matched untreated nTg group in terms of most vital parameters (Fig. 1.3). Therefore, HSP70 treatment, which has beneficial effects on the temporal cortex and the hippocampus of old Tg mice, apparently delays aging, may extend the lifespan, and improve life quality, as we showed previously using wild-type mice [9]. Based on the accumulated data, 5XFAD transgenic mice may be considered as a model of accelerated aging. It is of note that our speculative suggestion to consider AD as a manifestation of an accelerated aging process has not been universally accepted yet. However, it is clear that aging and AD are intrinsically interwoven with each other and applied HSP70 treatment could be insightful in both aging and AD.

The accumulated data on the morphology of neurons in different age groups of 5XFAD mice after HSP70 treatment are summarized in Fig. 6. It is evident that with age, in both groups (Tg versus nTg) the response to HSP70 significantly decreases, while in Tg mice of both ages, the treatment results in significant beneficial effects in all studied brain structures (Figs. 1.6 and 6). In contrast, application of HSP70 in the case of old nTg mice, results in deterioration of neuronal state mainly in the cortex (Fig. 1.5). However, comparison of neuronal morphology in old nTg and old nTg+HSP70 with young nTg mice, indicates that most changes occurring in the cortex of old nTg+HSP70 mice, make them more similar to young nTg animals. Therefore, HSP70 treatment applied in old nTg and old Tg mice delays manifestation of aging characteristics. It is clear that HSP70 treatment applied apparently manifests highly beneficial effects in the transgenic animals especially at the advanced stage of neuropathology development (Figs. 1.6 and 6). In both Tg and nTg mice, HSP70 treatment significantly delayed aging process which corroborates our previous studies demonstrating that long-term administration of recombinant HSP70 significantly enhanced the lifespan of different age groups of wt mice [9].

Fig.6

The results of Hsp70 treatment on morphology of neurons in the investigated brain regions (the temporal cortex and hippocampal CA1 and CA3 areas) in 5XFAD mice of different genotypes and ages. Six morphological characteristics that were monitored in the study (cytolysis, pyknosis, karyolysis, vacuolization, neuronal density, and the proportion of normal intact neurons in the brain), were taken as 100%.

According to the results of RNA-Seq transcriptomic profiling, the presence of the transgenic cassette results in the activation of adaptive and innate immune response, inflammatory response, synthesis of chemokines and leukocyte chemotaxis (Figs. 3 and 4). Basically, this agrees with the previous findings [38, 39].

A significant part of upregulated genes is specifically expressed in leukocytes. For example, cystatin F (Cst7), a protease inhibitor that is present in endosomes and lysosomes and selectively expressed in immune cells [40], is the 1st top DEG. It is 3-fold greater expressed in old Tg mice compared to young Tg, and is nearly absent in non-transgenic animals. The same observation was noticed for many chemokines and cytokines, cathepsins B, E, S, Z, and other genes. This transcriptomic effect seems to be the result of infiltration of leukocytes in brain tissues, and this process in much more severe in old mice. Infiltration of immune cells has been previously shown for 5XFAD model [41].

In contrast, HSP70 treatment does not result in such massive alterations of gene expression profiles (Fig. 4). For young mice, there is a tendency for HSP70 to upregulate genes participating in neuronal transmission and neurotransmitter biosynthesis. Noteworthy, these changes may reach large values in their amplitude (up to 5 ... 10-fold or more), and they occur simultaneously for several genes in the same mice. However, such changes are manifested not in all animals of the HSP70 subgroup (7 of 13). Heterogeneity of such “expression bursts” between samples and their similarity between multiple genes suggest their dynamic nature and may represent some of positive downstream effects of HSP70 treatment and neuronal system condition of individual mice.

More attention should be paid to the impact of HSP70 administration on the transcriptomes of old animals. As illustrated above (Fig. 4), transcriptomes of non-treated old Tg mice (compared to nTg) demonstrate enrichment with differentially expressed genes (predominantly, down-regulated) that participate in neuro- and gliogenesis. These genes are not altered either in young, or in old mice treated with HSP70. The amplitude of these changes is not high (10–50%), but they are massive: more than a hundred neuronal-related genes is affected. This transcriptomic effect may result from the significantly decreased proportion of healthy neurons in all studied brain areas of non-treated old Tg animals compared to the old non-treated nTg mice (Fig. 1.4). Such observation underlines the positive role of HSP70 on neuronal system condition of Tg animals and perfectly agrees with our histological data.

The most striking and consistent transcriptomic effect of HSP70 on Tg mice is the upregulation of genes participating in antigen processing and presentation (Figs. 3 and 5). It is more pronounced in old mice. Mainly, we are talking about H2-D1 and H2-K1, the highly expressed (CPM = 70–80) members of MHC-Ia complex. These expression gains are also observed for member of other MHC complexes (Ia, IIa, Ib, IIb) but they are expressed at a lower level (CPM = 3–15).

Presentation of antigen provides a link between innate and adaptive immunity. In this key process, components of the innate immune system, such as macrophages or dendritic cells, stimulate the adaptive immune response by presenting the captured antigen to the T-lymphocytes. Innate immunity and inflammation are well-known factors of AD onset and progression. Importantly it was recently demonstrated that innate immunity response aggravates, and adaptive–attenuates Alzheimer-like symptoms and ameliorates neuronal state and function in 5XFAD mice. It was also shown that the treatment of adaptive immune-deficient mice (Rag-5XFAD) or microglial cells with pre-immune IgG dramatically enhances Aβ clearance [42].

It is well known that the effect of HSP70 strongly depends on its localization: inside cells, HSP70 plays a role of molecular chaperone to help in protein folding or the resolution of protein aggregates. Being released from the cell, HSP70 acts as messenger communicating the cells’ interior protein composition to the immune system for initiation of immune responses against intracellular proteins. It can act as cytokine and peptide adjuvant, thereby promoting both the innate and adaptive immune responses [12, 14].

Taking into account trace quantity and fast proteolysis of intranasally injected HSP70 in the brains of treated mice [43], we suggest that in this case the effect of exogenous HSP70 is not realized by its well-known antiapoptotic and chaperonic qualities, i.e., pathways of intracellular HSP70. Probably in the case of Tg mice, especially the old ones, the introduced HSP70 behaves as a danger signal that activates genes participating in peripheral adaptive immunity that are necessary for clearing of pathogenic Aβ and restraining AD pathology. Thus, the observed overexpression of MHC-I and MHC-II in Tg mice indicates an activation or reprogramming of microglia caused by exogenous HSP70, enhancing Aβ clearance by microglia and macrophages as has been originally suggested [44].

Therefore, one may speculate that the demonstrated beneficial effects of HSP70 sub-chronic administration on neuron structure and density in the old Tg mice may represent a new therapeutic target to ameliorate AD-like pathologies even in the late stages of the disease.

Footnotes

ACKNOWLEDGMENTS

Work is supported by grants of Russian Science Foundation (RSCF) 18-15-00392 and 17-74-30030. Grant of Ministry of Education and Science of the Russian Federation (agreement No. 14.Z50.31.0014).

Authors want to thank Prof. Gregory Enikolopov fron Stony Brook University for careful reading of the MS and many useful suggestions.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Calderwood

, Murshid

, Prince

(2009) The shock of aging: Molecular chaperones and the heat shock response in longevity and aging—a mini-review. Gerontology 55, 550–558.

Morimoto

, Cuervo

(2014) Proteostasis and the aging proteome in health and disease. J Gerontol A Biol Sci Med Sci 69(Suppl 1), S33–S38.

Bobkova

, Garbuz

, Nesterova

, Medvinskaya

, Samokhin

, Alexandrova

, Yashin

, Karpov

, Kukharsky

, Ninkina

, Smirnov

, Evgen’ev

(2014) Therapeutic effect of exogenous hsp70 in mouse models of Alzheimer’s disease. J Alzheimers Dis 38, 425–435.

Kim

, Han

, Lee

, Yenari

(2018) The 70-kDa heat shock protein (Hsp70) as a therapeutic target for stroke. Expert Opin Ther Targets 22, 191–199.

Guzhova

, Margulis

(2006) Hsp70 chaperone as a survival factor in cell pathology. Int Rev Cyt 254, 101–149.

Leak

(2014) Heat shock proteins in neurodegenerative disorders and aging. J Cell Commun Signal 8, 293–310.

Evgen’ev

, Krasnov

, Nesterova

, Garbuz

, Karpov

, Morozov

, Snezhkina

, Samokhin

, Sergeev

, Kulikov

, Bobkova

(2017) Molecular mechanisms underlying neuroprotective effect of intranasal administration of human Hsp70 in mouse model of Alzheimer’s disease. J Alzheimers Dis 59, 1415–1426.

Murshid

, Eguchi

, Calderwood

(2013) Stress proteins in aging and life span. Int J Hypertherm 29, 442–447.

Bobkova

, Evgen’ev

, Garbuz

, Kulikov

, Morozov

, Samokhin

, Velmeshev

, Medvinskaya

, Nesterova

, Pollock

, Nudler

(2015) Exogenous Hsp70 delays senescence and improves cognitive function in aging mice. Proc Natl Acad Sci U S A 112, 16006–16011.

10.

Kim

, Hipp

, Bracher

, Hayer-Hartl

, Hartl

(2013) Molecular chaperone functions in protein folding and proteostasis. Annu Rev Biochem 82, 323–355.

11.

Duncan

, Cheetham

, Chapple

, van der Spuy

(2015) The role of HSP70 and its co-chaperones in protein misfolding, aggregation and disease. Subcell Biochem 78, 243–273.

12.

Calderwood

, Gong

, Murshid

(2016) Extracellular HSPs: The complicated roles of extracellular HSPs in immunity. Front Immunol 7, 159.

13.

Kirkegaard

, Gray

, Priestman

, Wallom

, Atkins

, Olsen

, Klein

, Drndarski

, Petersen

, Ingemann

, Smith

, Morris

, Bornæs

, Jørgensen

, Williams

, Hinsby

, Arenz

, Begley

, Jäättelä

, Platt

(2016) Heat shock protein-based therapy as a potential candidate for treating the sphingolipidoses. Sci Transl Med 8, 355ra118.

14.

Malyshev

(2013) HSP70 in the immune responses. In Immunity, Tumors and Aging: The Role of HSP70, vol. 6, Springer Briefs in Biochemistry and Molecular Biology. Springer, Dordrecht.

15.

De Mena

, Chhangani

, Fernandez-Funez

, Rincon-Limas

(2017) secHsp70 as a tool to approach amyloid-beta42 and other extracellular amyloids. Fly 11, 179–184.

16.

Rivera

, Capone

, Cauvi

, Arispe

, De Maio

(2018) Modulation of Alzheimer’s amyloid beta peptide oligomerization and toxicity by extracellular Hsp70. Cell Stress Chaperones 23, 269–279.

17.

Rozhkova

, Yurinskaya

, Zatsepina

, Garbuz

, Karpov

, Surkov

, Murashev

, Ostrov

, Margulis

, Evgen’ev

, Vinokurov

(2010) Exogenous mammalian extracellular HSP70 reduces endotoxin manifestations at the cellular and organism levels. Ann N Y Acad Sci 1197, 94–107.

18.

Hoshino

, Suzuki

, Matsushima

, Yamakawa

, Suzuki

, Mizushima

(2013) Suppression of Alzheimer’s disease-related phenotypes by geranylgeranylacetone in mice. PloS One 8, e76306.

19.

Salway

, Gallagher

, Page

, Stuart

(2011) Higher levels of heat shock proteins in longer-lived mammals and birds. Mech Ageing Dev 132, 287–297.

20.

Martin

, Boucher

, Fontaine

, Delarasse

(2017) Distinct inflammatory phenotypes of microglia and monocyte-derived macrophages in Alzheimer’s disease models: Effects of aging and amyloid pathology. Aging Cell 16, 27–38.

21.

Martinez de Toda

, De la Fuente

(2015) The role of Hsp70 in oxi-inflamm-aging and its use as a potential biomarker of lifespan. Biogerontolog 16, 709–721.

22.

Hou

, Wei

, Luo

, Liu

(2010) Modulating expression of brain heat shock proteins by estrogen in ovariectomized mice model of aging. Exp Geront 45, 323–330.

23.

Calderwood

, Murshid

(2017) Molecular chaperone accumulation in cancer and decrease in Alzheimer’s disease: The potential roles of HSF1. Front Neurosci 11, 192.

24.

Xia

, Jiang

, McDermott

, Han

(2018) Aging and Alzheimer’s disease: Comparison and associations from molecular to system level. Aging Cell 17, e12802.

25.

Webster

, Bachstetter

, Nelson

, Schmitt

, Van Eldik

(2014) Using mice to model Alzheimer’s dementia: An overview of the clinical disease and the preclinical behavioral changes in 10 mouse models. Front Genet 5, 88.

26.

Eimer

, Vassar

(2013) Neuron loss in the 5XFAD mouse model of Alzheimer’s disease correlates with intraneuronal Abeta42 accumulation and Caspase-3 activation. Mol Neurodegener 8, 2.

27.

Krasnov

, Dmitriev

, Kudryavtseva

, Shargunov

, Karpov

, Uroshlev

, Melnikova

, Blinov

, Poverennaya

, Archakov

, Lisitsa

, Ponomarenko

(2015) PPLine: An automated pipeline for SNP, SAP, and splice variant detection in the context of proteogenomics. J Proteome Res 14, 3729–3737.

28.

Sigurgeirsson

, Emanuelsson

, Lundeberg

(2014) Sequencing degraded RNA addressed by 3’ tag counting. PloS One 9, e91851.

29.

Robinson

, McCarthy

, Smyth

(2010) edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140.

30.

Oakley

, Cole

, Logan

, Maus

, Shao

, Craft

, Guillozet-Bongaarts

, Ohno

, Disterhoft

, van Eldik

, Berry

, Vassar

(2006) Intraneuronal beta-amyloid aggregates, neurodegeneration, and neuron loss in transgenic mice with five familial Alzheimer’s disease mutations: Potential factors in amyloid plaque formation. J Neurosci 26, 10129–10140.

31.

Shukla

, Zheng

, Mishra

, Amin

, Steiner

, Grant

, Kesavapany

, Pant

(2013) A truncated peptide from p35, a Cdk5 activator, prevents Alzheimer’s disease phenotypes in model mice. FASEB J 27, 174–186.

32.

Landel

, Annweiler

, Millet

, Morello

, Féron

(2016) Vitamin D, cognition and Alzheimer’s disease: The therapeutic benefit is in the D-tails. J Alzheimers Dis 53, 419–444.

33.

Bernier

, Bedard

, Vinet

, Levesque

, Parent

(2002) Newly generated neurons in the amygdala and adjoining cortex of adult primates. Proc Natl Acad Sci U S A 99, 11464–11469.

34.

Kokoeva

, Yin

, Flier

(2005) Neurogenesis in the hypothalamus of adult mice: Potential role in energy balance. Science 310, 679–683.

35.

Ohira

, Takeuchi

, Shoji

, Miyakawa

(2013) Fluoxetine-induced cortical adult neurogenesis. Neuropsychopharmacology 38, 909–920.

36.

Ramaswamy

, Goings

, Soderstrom

, Szele

, Kozlowski

(2005) Cellular proliferation and migration following a controlled cortical impact in the mouse. Brain Res 1053, 38–53.

37.

Nakatomi

, Kuriu

, Okabe

, Yamamoto

, Hatano

, Kawahara

, Tamura

, Kirino

, Nakafuku

(2002) Regeneration of hippocampal pyramidal neurons after ischemic brain injury by recruitment of endogenous neural progenitors. Cell 110, 429–441.

38.

Landel

, Baranger

, Virard

, Loriod

, Khrestchatisky

, Rivera

, Benech

, Feron

(2014) Temporal gene profiling of the 5XFAD transgenic mouse model highlights the importance of microglial activation in Alzheimer’s disease. Mol Neurodegener 9, 33.

39.

Lee

CYD

, Daggett

, Gu

, Jiang

, Langfelder

, Li

, Wang

, Zhao

, Park

, Cooper

, Ferando

, Mody

, Coppola

, Xu

, Yang

(2018) Elevated TREM2 gene dosage reprograms microglia responsivity and ameliorates pathological phenotypes in Alzheimer’s disease models. Neuron 97, 1032–1048e1035.

40.

Kos

, Nanut

, Prunk

, Sabotic

, Dautovic

, Jewett

(2018) Cystatin F as a regulator of immune cell cytotoxicity. Cancer Immunol Immunother 67, 1931–1938.

41.

Thériault

, ElAli

, Rivest

(2015) The dynamics of monocytes and microglia in Alzheimer’s disease. Alzheimers Res Ther 7, 41.

42.

Marsh

, Abud

, Lakatos

, Karimzadeh

, Yeung

, Davtyan

, Fote

, Lau

, Weinger

, Lane

, Inlay

, Poon

, Blurton-Jones

(2016) The adaptive immune system restrains Alzheimer’s disease pathogenesis by modulating microglial function. Proc Natl Acad Sci U S A 113, E1316–1325.

43.

Yurinskaya

, Zatsepina

, Vinokurov

, Bobkova

, Garbuz

, Morozov

, Kulikova

, Mitkevich

, Makarov

, Funikov

, Evgen’ev

(2015) The fate of exogenous human HSP70 introduced into animal cells by different means. Curr Drug Deliv 12, 524–532.

44.

Kakimura

, Kitamura

, Takata

, Umeki

, Suzuki

, Shibagaki

, Taniguchi

, Nomura

, Gebicke-Haerter

, Smith

, Perry

, Shimohama

(2002) Microglial activation and amyloid-beta clearance induced by exogenous heat-shock proteins. FASEB J 16, 601–603.