Abstract

   Physical activities have beneficial effects on cardiovascular health,
   although the specific mechanisms are largely unknown. Cardiac resident
   macrophages (cMacs) and the distribution of their subsets are critical
   regulators for maintaining cardiovascular health and cardiac functions
   in both steady and inflammatory states. Therefore, we investigated the
   subsets of cMacs in mice after low-intensity exercise training to
   elucidate the exercise-induced dynamic changes of cMacs and the
   benefits of exercise for the heart. The mice were subjected to
   treadmill running exercise five days per week for five weeks using a
   low-intensity exercise training protocol. Low-intensity exercise
   training resulted in a suppression of body weight gain in mice and a
   significant increase in the ejection fraction, a parameter that
   represents the systolic function of the heart. Low-intensity exercise
   training induced the alterations in the transcriptome of the heart,
   which are associated with muscle contraction and mitochondrial
   function. Furthermore, low-intensity exercise training did not alter
   the number of lymphocyte antigen 6 complex, locus C1 (Ly6c)^− cMacs but
   instead remodeled the distributions of Ly6c^− cMac subsets. We observed
   an increase in the percentage of major histocompatibility complex class
   II (MHCII)^low cMacs and a decrease in the percentage of MHCII^high
   cMacs in the heart after low-intensity exercise training. Therefore,
   the benefits of exercise for cardiovascular fitness might be associated
   with the redistribution of cMac subsets and the enhancement of the
   ejection fraction.

   Keywords: Cardiac macrophage, Ejection fraction, Heart, Low-intensity
   exercise, Innate immunity

1. Introduction

   Cardiovascular diseases (CVDs) are the leading cause of death
   worldwide. In China, CVDs accounted for 45.50 % and 43.16 % of all
   deaths in rural and urban areas, respectively, in 2016 [[37]1]. In the
   United States, approximately 0.6 million people die from CVDs each year
   [[38]2]. Over the past 30 years, there has been a significant increase
   in the incidence of CVDs among young individuals aged 18–45 years
   [[39]3]. Core health behaviors such as smoking, physical activity,
   diet, body weight, as well as health factors including cholesterol,
   blood pressure, and glucose control, are closely associated with
   cardiovascular health [[40]4]. Physical activity and physical fitness
   have shown benefits for cardiovascular health. Epidemiological studies
   among CVD patients have indicated that the risk of death is 4.5 times
   higher in the least fit healthy adults compared to the most fit
   individuals. The intensity of exercise is a more important predictor of
   death than other risk factors such as smoking, high blood pressure,
   high cholesterol, and diabetes. Furthermore, participating in a formal
   exercise program can reduce the death rate of myocardial infarction
   patients by 20 %–25 % [[41]5]. However, the specific benefits of
   exercise for cardiovascular health and the underlying mechanisms are
   still not well understood.

   Physical activity may reduce the risk of CVDs by increasing levels of
   circulating high-density lipoprotein, improving insulin sensitivity,
   and promoting cardiac adaptations [[42]6]. Some clinical and animal
   studies have suggested that the benefits of exercise for CVDs, type 2
   diabetes, and cancers are associated with the anti-inflammatory effects
   of exercise. These effects include increased secretion of
   anti-inflammatory cytokines such as interleukin-6 (IL-6) and IL-10,
   elevated numbers of circulating IL-10-secreting regulatory T cells, and
   reductions in Toll-like receptor expression on monocytes and downstream
   pro-inflammatory responses [[43]7]. However, the roles of
   exercise-induced alterations in innate immunity are still largely
   unknown [[44]8].

   Cardiac resident macrophages (cMacs) play a critical role in the host's
   response to cardiac injury, tissue damage, and microbial insults. In a
   steady state, cMacs are actively involved in the uptake of
   dysfunctional mitochondria and other extracellular vesicles released by
   cardiomyocytes. Depletion of cMacs can lead to inflammasome activation,
   impaired autophagy, accumulation of abnormal mitochondria in
   cardiomyocytes, metabolic alterations, and ventricular dysfunction
   [[45]9]. Cardiac macrophages (cMacs) consist of heterogeneous cell
   populations including the resident subsets originate from yolk
   sac-derived erythromyeloid progenitors and circulating monocytes
   [[46]10,[47]11]. The resident cMac subsets have distinct roles within
   the cardiac microenvironment, and their relative ratio is crucial for
   cardiac inflammation and aging [[48]12]. However, the specific roles of
   cMac subsets in the benefits of exercise for cardiac function are not
   yet well understood.

   Here, the systolic function of heart and the cMacs subsets were
   investigate in mice which were applied a low-intensity exercise
   training protocol to understand the benefits of exercise in
   cardiovascular health. Low-intensity exercise training was found to
   elevate systolic function of heart and remodel cardiac resident
   macrophages. Thus, the benefits of exercise to cardiovascular health
   might be associated with the redistribution of cMacs in heart.

2. Materials and methods

2.1. Mice and treatment

   C57BL/6J male mice (8 weeks old) were housed in specific pathogen-free
   facilities at Xiamen University, following the guidelines of the Animal
   Care and Use Committee (Permit No. XMULAC20200150). The mice underwent
   a treadmill running exercise protocol for 5 days per week over a period
   of 5 weeks. In the first week, the mice were subjected to adaptive
   training with the following protocol: 6 m/min for 5 min, 9 m/min for
   5 min, 12 m/min for 5 min, and 15 m/min for 5 min. Subsequently, for
   the next 4 weeks, the mice were forced to run at 6 m/min for 5 min and
   12 m/min for 35 min. Control mice were housed in the cages. After 5
   weeks of treatment, tissues were harvested ([49]Fig. 1A).

Fig. 1.

   [50]Fig. 1
   [51]Open in a new tab

   Low-intensity exercise training represses the gain of body weight in
   mice. (A) Experimental procedure and training protocol. (B) Food intake
   of mice with or without training. (C) Body weights of mice with or
   without training. (D) Body weight gains of mice with or without
   training for 5 weeks. Data are shown as mean ± SD. n = 6.

2.2. Echocardiography

   Cardiac function in mice was evaluated using trans-thoracic
   high-resolution echocardiography (Vevo 2100, FUJIFILM VisualSonics)
   before and after 5 weeks of training. Echocardiography was performed
   with the mice under light isoflurane anesthesia (1.0–2.25 %; RWD),
   adjusted to maintain a heart rate of 450–550 beats per minute. Image
   acquisition was conducted using the MS-550D 22–55 MHz linear array
   transducer (FUJIFILM VisualSonics). The left ventricular ejection
   fraction (EF%) was calculated by tracing the systolic and diastolic
   endocardium in the parasternal short-axis view. Echocardiographic data
   were digitally stored and analyzed offline using the provided software.
   All echocardiographic studies were conducted under consistent
   conditions, and the investigators were blinded to the treatment of
   individual mice during the experiments and outcome assessments.

2.3. cMacs purification and flow cytometry analysis

   The whole heart was perfused with saline prior to harvest.
   Subsequently, the heart was mechanically disrupted and digested in HBSS
   buffer containing 0.5 μg/ml Liberase (Roche) for 30 min at 37 °C with
   gentle rocking. The cells were dissociated and passed through a 100 μm
   strainer for filtration.

   All cells were then stained with fluorescently labeled antibodies for
   30 min on ice in MACS buffer, which consisted of 1 × PBS with 1 % BSA
   and 2 mM EDTA. Prior to staining, FcR blocking reagent (Miltenyibiotec)
   was used to prevent nonspecific binding. Flow cytometry (BD LSR
   Fortessa) was employed to analyze the stained cells. To label dead
   cells, Fixable viability dye (FVD, Thermo Fisher) was used according to
   the manufacturer's instructions. Antibodies used for staining are
   listed in [52]Table 1. Non-staining and single-stained samples were
   prepared using splenocytes to set up compensation. Data analysis was
   performed using FlowJo software (v.10.0.7, Tree Star Inc). After data
   cleaning and concatenation, t-distributed stochastic neighbor embedding
   (t-SNE) analysis was applied to CD45^+ CD11b^+ F4/80^+ cells using
   FlowJo software to identify the distinct cell clusters and patterns
   between two groups.

Table 1.

   Antibodies used in this study.
   Antibodies Source Identifier
   PE anti-mouse CD45 antibody (Clone 30-F11) Biolegend Cat# 103105
   PerCP/Cy5.5 anti-mouse CD11b antibody (Clone M1/70) Biolegend Cat#
   101228
   APC anti-mouse F4/80 antibody (Clone BM8) Biolegend Cat# 123116
   Brilliant Violet 510™ anti-mouse I-A/I-E antibody (Clone M5/114.15.2)
   Biolegend Cat# 107635
   FITC anti-mouse Ly-6C antibody (Clone HK1.4) Biolegend Cat# 128005
   CD68 Rabbit pAb ABclonal Cat# A13286
   Cardiac troponin T (TNNT2) Rabbit mAb ABclonal Cat# A4914
   [53]Open in a new tab

2.4. Bioinformatic analysis

   RNA-seq data were acquired from the Gene Expression Omnibus (GEO)
   database (GSE:229475). Statistical analysis was performed in R.
   Transcripts per kilobase Million (TPMs) were calculated for raw count
   data by dividing the number of counts by exon length in kilobases
   (RPK), dividing the total number of counts by 1 million (CM), and then
   dividing the RPK by CM. The edgeR package was used to identify the
   significantly differentially expressed genes (DEGs) with fold change
   (Fc) > 1.5 and p-value<0.05. Gene set enrichment analysis (GSEA) was
   performed using the MSigDB Gene annotation gene set (M2 CP: Canonical
   pathways, available at
   [54]https://www.gsea-msigdb.org/gsea/msigdb/index.jsp) with GSEABase.
   Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment
   analysis were performed using the clusterProfiler package.

2.5. Immunohistochemical (IHC) staining

   Hearts were fixed in 4 % paraformaldehyde, followed by embedding in
   paraffin. Paraffin-embedded tissues were then sectioned to produce 5 μm
   thick sections. These sections were subjected to deparaffinization and
   rehydration processes. To block endogenous peroxidase activity, the
   sections were treated with 3 % H[2]O[2] for 30 min. Non-specific
   binding sites were blocked by incubating the sections with 1 % BSA in
   PBS for 30 min. Subsequently, the sections were incubated with primary
   antibodies against CD68 (1:50 dilution) and TNNT2 (1:50 dilution)
   overnight at 4 °C. Peroxidase-conjugated secondary antibodies were
   applied for detection. After the secondary antibody incubation, the
   slides were counterstained with hematoxylin for 5 min, followed by
   dehydration and mounting. The stained sections were examined and images
   captured using a bright field microscope (TissueFAXS P1,
   Tissuegnostics). The absolute number of CD68^+ cells per 20 × field was
   quantified by NIH ImageJ [[55]13].

2.6. Statistics

   Statistical analysis was conducted using Prism 9 software. For
   comparing two groups, an unpaired two-tailed Student's t-test was
   performed. The normality of the data was assessed using the
   Shapiro-Wilk test, and the variances were compared using the F test. In
   [56]Fig. 1C, a two-way ANOVA with Šídák's multiple comparisons test was
   applied. The data were presented as mean ± standard deviation (SD). p
   values ≤ 0.05 were considered statistically significant. p values are
   denoted in figures as: not significant (NS), p > 0.05; *, p ≤ 0.05; **,
   p < 0.01; ***, p < 0.001.

3. Results

3.1. Low-intensity exercise training represses the gain of body weight in
mice

   To investigate the impact of exercise on cardiovascular health, a group
   of mice underwent a low-intensity training regimen ([57]Fig. 1A). There
   was no significant difference in food intake between the control group
   and the exercise group ([58]Fig. 1B). The body weights of control mice
   exhibited an increasing trend with age, whereas the exercised mice had
   significantly lower body weights at the indicated time points ([59]Fig.
   1C). Furthermore, the exercise regimen led to a noticeable reduction in
   the increasing trend of body weights observed in the control mice, and
   the body weight gains in the exercise group were significantly reduced
   ([60]Fig. 1D). These findings suggest that low-intensity exercise
   training can effectively control the body weight of mice without
   affecting their food intake.

3.2. Low-intensity exercise training elevates ejection fraction of the heart

   EF% is a measurement that indicates the percentage of blood being
   pumped out of the heart with each contraction. Improvements in EF% have
   been consistently associated with enhanced quality of life, reduced
   rehospitalization rates, and lower mortality in patients with heart
   failure [[61]14]. Firstly, the heart weight to tibia length ratios did
   not show any significant differences between the control mice and the
   exercised mice ([62]Fig. 2A), suggesting that cardiac hypertrophy did
   not occur as a result of the low-intensity exercise training [[63]15].
   Additionally, echocardiograms were performed before and after 5 weeks
   of training, and a slight but significant increase in EF% was observed
   after the training period ([64]Fig. 2B and C). These findings suggest
   that low-intensity exercise training may be beneficial for
   cardiovascular health by improving contractility. The low-intensity
   training protocol employed in this study appears to confer benefits to
   the heart.

Fig. 2.

   [65]Fig. 2
   [66]Open in a new tab

   Low-intensity exercise training elevates ejection fraction of the
   heart. (A) Heart weight/tibia length of mice with or without training
   for 5 weeks. (B) The typical echocardiographic images of indicated mice
   in the parasternal short-axis view. (C) Ejection fraction (EF%) of
   indicated mice. Data are shown as mean ± SD. n = 6.

3.3. Low-intensity exercise training alters the transcriptional profile
associated with cardiac muscle contraction and mitochondrial function

   In order to gain further insights into the impact of low-intensity
   exercise on the function of the heart, we performed RNA sequencing
   analysis using publicly available dataset, which provided a
   comprehensive view of the global transcriptional response of mice
   hearts to 12 weeks of low-intensity exercise. Through GO term
   enrichment analysis of differentially expressed genes, we found that
   the functions associated with low-intensity exercise were primarily
   related to energy metabolism, muscle function, and mitochondrial
   processes ([67]Fig. 3A and B). Additionally, the enrichment analysis of
   KEGG pathways and GSEA suggested that exercise may influence pathways
   associated with oxidative phosphorylation and cardiac muscle
   contraction ([68]Fig. 3C and D). Specifically, we observed an
   upregulation of genes involved in Na^+/K^+-ATPase (Atp1a1 and Atp1a2)
   and actin-tropomyosin and actin-myosin system (Tpm1, Tpm2, Tpm3, Tnnt2,
   Tnni3, Myl2, Myh6, Myh7 and Myl3) which contribute to cardiac muscle
   contraction ([69]Fig. 3E). Exercise-induced increased expression of
   tetraspanins, including TSPAN4 and TSPAN9, suggested that mitocytosis
   was enhanced, which is critical for eliminating damaged mitochondria to
   maintain mitochondrial pool quality and cellular homeostasis [[70]16]
   ([71]Fig. 3F). Taken together, our results indicate that low-intensity
   exercise training may enhance cardiac muscle contraction through a
   mitochondria-dependent pathway, which has previously been reported to
   be associated with cMacs [[72]9].

Fig. 3.

   [73]Fig. 3
   [74]Open in a new tab

   Low-intensity exercise training alters the transcriptional profile
   associated with cardiac muscle contraction and mitochondrial function.
   (A and B) Gene Ontology (GO) term enrichment analysis. (C) Kyoto
   Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis.
   (D) Gene Set Enrichment Analysis (GSEA) analysis. Exercise-induced
   significantly differentially expressed genes (DEGs) were used for the
   analysis. Heatmap indicated the exercise-induced DEGs which are
   associated with cardiac muscle contraction and (E) mitocytosis (F).

3.4. Low-intensity exercise does not change the number of cMacs in heart

   cMacs are known to play crucial roles in maintaining cardiac functions
   both under steady-state conditions and during inflammation
   [[75]9,[76]10]. However, the impact of low-intensity exercise training
   on cMacs and their subsets remains largely unknown. CD11b and F4/80
   were used to label cardiac macrophages and some monocytes. CD11b^+
   F4/80^+ cells accounted for 54.8 % of CD45^+ lymphocytes in the cardiac
   microenvironment. This ratio and CD11b^+ F4/80^+ cell counts did not
   significantly change following low-intensity exercise training
   ([77]Fig. 4A–C). Concurrently, there was no significant change in the
   count of CD68^+ cells in the heart sections, as identified by IHC,
   following exercise ([78]Fig. 4D). Given that Ly6c^+ cells are known to
   represent monocytes and CCR2-dependent progenitors for tissue
   macrophage populations [[79]12], the percentages of the Ly6c− cMacs
   subset within the overall cMacs population were analyzed. The subset
   accounted for 77.5 ± 11.4 % in control mice and 74.1 ± 12.6 % in
   exercised mice ([80]Fig. 4E and F). Furthermore, we evaluated the
   expression levels of genes associated with macrophage markers,
   including Ptprc (CD45), Cd68, Adgre1 (F4/80), and Itgam (CD11b), in the
   hearts of mice following low-intensity exercise training. The
   expression levels of these genes did not show significant changes after
   exercise training ([81]Fig. 4G). This suggests that the total abundance
   of macrophages in the heart remained relatively stable during the
   course of low-intensity exercise training.

Fig. 4.

   [82]Fig. 4
   [83]Open in a new tab

   Low-intensity exercise training does not change the number of cMacs in
   heart. (A) Representative flow cytometry plots indicating the CD45^+
   cells and CD11b^+ F4/80^+ cells in the heart. Percentages (B) and cell
   counts (C) of CD11b^+ F4/80^+ cells among CD45^+ cells in the heart of
   mice with or without exercise. (D) Heart sections were stained for CD68
   and TNNT2 using immunohistochemistry. The absolute number of CD68^+
   cells per 20 × field was quantified. (E) Representative flow cytometry
   plots indicating the cMacs with Ly6c and MHCII staining. (F)
   Percentages of Ly6c^− cMacs among CD11b^+ F4/80^+ CD45^+ cells in the
   heart of mice with or without exercise. Data are shown as mean ± SD.
   n = 5. (G) The expression levels of indicated genes in the heart.
   Transcripts per kilobase million (TPM) were calculated from RNA-seq
   data. Data are shown as mean ± SD. n = 3.

3.5. Low-intensity exercise training remodels the subsets of cMacs in heart

   To gain further insights into the subsets of cMacs and their potential
   alterations induced by low-intensity exercise training, we performed
   t-SNE analysis based on the expression levels of CD11b, F4/80, MHCII,
   and Ly6c using flow cytometry data. Consistent with the previous
   quantification results ([84]Fig. 4E), the majority of cMacs belonged to
   the Ly6c^low population ([85]Fig. 5A). We observed a distinct cluster
   in the mice subjected to low-intensity exercise training, indicating a
   potential shift in the cMac population ([86]Fig. 5A). This cluster
   exhibited relatively low expression levels of MHCII and Ly6c ([87]Fig.
   5A). Additionally, Ly6c^− cMacs were further categorized into MHCII
   ^high and MHCII ^low subsets ([88]Fig. 5B). Notably, low-intensity
   exercise training led to a significant decrease in the MHCII ^high
   population of Ly6c^− cMacs and a significant increase in the MHCII ^low
   population, which has previously been reported as a CCR2^− resident
   cMacs subset [[89]10,[90]12,[91]17] ([92]Fig. 5C and D). Moreover,
   low-intensity exercise training significantly increased the expression
   levels of myeloid-epithelial-reproductive tyrosine kinase (MerTK) in
   the heart, which was highly expressed in MHCII^low resident cMacs
   subset [[93]18], but not CCR2 ([94]Fig. 5E). This suggests that
   low-intensity exercise training remodels the subsets of cMacs in the
   heart.

Fig. 5.

   [95]Fig. 5
   [96]Open in a new tab

   Low-intensity exercise training remodels the subsets of cMacs in heart.
   (A) t-SNE plots displaying the subsets of cMacs in the heart of mice
   with or without exercise. t-SNE analysis was performed using flow
   cytometry data. The expression levels of Ly6c and MHCII in the subsets
   of cMacs are shown. An arrow highlights a distinct cluster between the
   control and exercise groups. (B) Representative flow cytometry plots
   indicating the subsets of Ly6c^− cMacs with Ly6c and MHCII staining.
   The percentages of MHCII^high(C) and MHCII^low cMacs (D) in Ly6c^−
   cMacs in the heart of mice with or without exercise. Data are shown as
   mean ± SD. n = 5. (E) The expression levels of indicated genes in the
   heart are shown. Transcripts per kilobase million (TPM) was calculated
   from RNA-seq data. Data are shown as mean ± SD. n = 3.

4. Discussion

   Here, low-intensity exercise training was shown to elevate the systolic
   function of the heart, alter the transcriptional profile associated
   with cardiac muscle contraction and mitochondrial function, and
   increase the number of MHCII ^low resident cMacs. Low-intensity
   exercise has been shown to accelerate mitochondrial ATP production,
   improve pulmonary oxygen kinetics [[97]19], stimulate bioenergetics,
   and increase fat oxidation in mitochondria [[98]20] in humans. These
   improvements in mitochondrial function contribute to improve cardiac
   function in diabetic cardiomyopathy [[99]21] and inhibit the
   development of cardiovascular disease [[100]22]. Increased expressions
   of Na^+/K^+-ATPase genes, including Atp1a1 and Atp1a2, was observed in
   the heart after exercise. Na^+/K^+-ATPase contributes to ATP hydrolysis
   and the supply of energy in the heart and neurons [[101]23,[102]24].
   The results support the notion that the benefits of exercise on the
   heart are associated with mitochondrial homeostasis.

   The MHCII ^low resident cMacs have been shown to express higher levels
   of Mertk, a macrophage phagocytic receptor that plays a role in
   efferocytosis and myocardial repair in both homeostasis and myocardial
   infarction [[103]18,[104]25]. In the healthy myocardium, viable
   cardiomyocyte-derived fragments containing damaged mitochondria serve
   as cellular waste [[105]26]. The elimination of these damaged
   mitochondria by cMacs ensures mitochondrial fitness and meets the
   high-energy demands of the heart [[106]9,[107]27,[108]28]. Depletion of
   cardiac macrophages or deficiency in Mertk leads to impaired
   elimination of mitochondria from the myocardial tissue, resulting in
   the accumulation of abnormal mitochondria in cardiomyocytes and
   metabolic alterations. These factors contribute to the repression of
   systolic function and the induction of left ventricular dysfunction
   [[109]9].

   Moreover, cMacs can stimulate cardiomyocytes to dedifferentiate and
   re-enter the cell cycle [[110]29]. They can also promote cardiac
   proliferation, particularly under low oxygen conditions [[111]30].
   Exercise has been reported to induce new cardiomyocyte generation via
   increased miR-222 expression in the heart [[112]31]. The decline in the
   population of self-renewing MHCII^low cMacs is considered an aging
   marker of the heart, and long-term exercise training is believed to
   benefit cardiovascular health by counteracting the aging process
   [[113]12,[114]32]. Thus, the exercise-induced redistribution of cMacs
   may exert multiple roles in maintaining cardiovascular fitness.

   Although it is challenging to directly determine the functional role of
   the MHCII^low cMacs subset in the benefits of low-intensity exercise
   training on the heart and mitochondrial homeostasis through gain- and
   loss-of- function assay, the results are in line with published reports
   suggesting a potential association. The upregulation of MHCII^low cMacs
   induced by low-intensity exercise training may be associated with the
   benefits for mitochondrial homeostasis and the systolic function of the
   heart.

   In conclusion, this study provides evidence that low-intensity exercise
   training can alter the distribution of cMacs and improve systolic
   function, potentially leading to benefits for cardiovascular health.

Ethics statement

   Animal experiments were conducted following the guidelines of the
   Animal Care and Use Committee at Xiamen University (Permit No.
   XMULAC20200150).

Funding

   The work was supported by grants from the National Natural Science
   Foundation of China (grant no. 82000304) to Liu B; the National Key
   Research and Development Program of China (2019YFE0113900), the Xiamen
   Medical and Health Key Project (3502Z20204005) and the Fujian
   Provincial Health Technology Project (2021QNB021) to C. Dai.

Data availability statement

   Data are available on request.

CRediT authorship contribution statement

   Gang Wang: Conceptualization, Formal analysis, Investigation,
   Methodology, Software, Writing - original draft. Lin Wang:
   Conceptualization, Formal analysis, Investigation, Methodology,
   Validation. Xuchao Wang: Formal analysis, Investigation. Heng Ye:
   Formal analysis, Investigation. Wei Ni: Formal analysis, Investigation.
   Wei Shao: Data curation, Funding acquisition, Resources, Writing -
   review & editing. Cuilian Dai: Data curation, Funding acquisition,
   Resources, Writing - review & editing. Binbin Liu: Conceptualization,
   Formal analysis, Funding acquisition, Project administration,
   Supervision, Visualization, Writing - original draft, Writing - review
   & editing.

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

References