Tissue-specific small heat shock protein 20 activation is not associated with traditional autophagy markers in Ossabaw swine with cardiometabolic heart failure

Abstract

The small heat shock protein 20 (HSPB6) emerges as a potential upstream mediator of autophagy. Although autophagy is linked to several clinical disorders, how HSPB6 and autophagy are regulated in the setting of heart failure (HF) remains unknown. The goal of this study was to assess the activation of the HSPB6 and its association with other well-established autophagy markers in central and peripheral tissues from a preclinical Ossabaw swine model of cardiometabolic HF induced by Western diet and chronic cardiac pressure overload. We hypothesized HSPB6 would be activated in central and peripheral tissues, stimulating autophagy. We found that autophagy in the heart is interrupted at various stages of the process in a chamber-specific manner. Protein levels of HSPB6, Beclin 1, and p62 are increased in the right ventricle, whereas only HSPB6 was increased in the left ventricle. Unlike the heart, samples from the triceps brachii long head showed only an increase in the protein level of p62, highlighting interesting central versus peripheral differences in autophagy regulation. In the right coronary artery, total HSPB6 protein expression was decreased and associated with an increase in LC3B-II/LC3B-I ratio, demonstrating a different mechanism of autophagy dysregulation in the coronary vasculature. Thus, contrary to our hypothesis, activation of HSPB6 was differentially regulated in a tissue-specific manner and observed in parallel with variable states of autophagy markers assessed by protein levels of LC3B, p62, and Beclin 1. Our data provide insight into how the HSPB6/autophagy axis is regulated in a preclinical swine model with potential relevance to heart failure with preserved ejection fraction.

NEW & NOTEWORTHY Our study shows that the activation of HSPB6 is tissue specific and associated with variable states of downstream markers of autophagy in a unique preclinical swine model of cardiometabolic HF with potential relevance to HFpEF. These findings suggest that targeted approaches could be an important consideration regarding the development of drugs aimed at this intracellular recycling process.

INTRODUCTION

Autophagy plays a critical role in a number of metabolic, cardiovascular, musculoskeletal, and inflammatory disorders that are directly or indirectly linked to the development of cardiomyopathy (16, 30). Autophagy can have both physiological and pathological regulatory roles, making it difficult to determine its role in cardiovascular disease and assess its potential as a therapeutic target. For example, treatment with chloramphenicol succinate in swine increased Beclin 1 (a marker of autophagy initiation) levels when administered before a myocardial ischemia-reperfusion procedure, but did not alter its expression when delivered during reperfusion (31). This study demonstrates the complexity of autophagy signaling and inherent issues in predicting the dynamic nature of its regulation in a systematic manner.

The ubiquitously expressed small heat shock protein 20 (HSPB6) is emerging as an upstream mediator of autophagy that may have dual regulatory roles. Activation of HSPB6 has been observed in failing human heart samples (27). In mice, constitutively cardiac-specific overexpression of phosphorylation of HSPB6 at serine-16 (p-HSPB6-Ser16) caused cardiac fibrosis, remodeling, and left ventricular dysfunction that was mediated via an interleukin-6 (IL-6) mechanism (7). Conversely, other studies have shown that p-HSPB6-Ser16 improves myocardial contractility, promotes coronary smooth muscle vasorelaxation, and may regulate skeletal muscle blood flow to enhance glucose uptake during exercise (2, 6, 35, 36, 40, 41), highlighting its dichotomous nature.

Thus, the purpose of this short report was to assess the protein levels of the HSPB6/autophagy axis in the heart, skeletal muscle, and coronary arteries of a newly developed Ossabaw swine model of cardiometabolic heart failure (HF) (23). This large animal model, recently described as a valuable “multihit” model for investigating complex pathophysiological mechanisms relevant to heart failure with preserved ejection fraction (HFpEF) (34), combines Western diet and chronic cardiac pressure overload interventions. We hypothesized that HSPB6 would be activated in central (heart and coronary artery) and peripheral (skeletal muscle) tissues of pigs with cardiometabolic HF and associated with the increase of other autophagy markers. Given the possible dual regulatory roles of HSPB6, the rationale for an increase in HF in animals (i.e., a pathological role) was based on the chronic stressors of diet, cardiac afterload, and a strong inflammatory phenotype previously observed in this large animal model (23). Contrary to our hypothesis, activation and protein levels of HSPB6 were differentially regulated in a tissue-specific manner and observed in parallel with variable states of downstream autophagy activation assessed via examination of the microtubule-associated proteins 1A/1B light chain 3B (LC3B), sequestosome1/p62 (p62), and Beclin 1.

MATERIALS AND METHODS

Experimental design.

Analysis of autophagy was performed in tissues collected from the same animals previously used in a recently published study by our laboratory (23). All animal protocols were in accordance with the Principles for the Utilization and Care of Vertebrate Animals Used in Testing Research and Training and approved by the University of Missouri Animal Care and Use Committee (Protocol No. 9779). Briefly, 2-mo-old (15–20 kg) intact female Ossabaw swines were divided into two groups as follows: a nonsham sedentary control group (CON; n = 5) and a Western diet-fed and aortic-banded HF group (WD-AB; n = 4). Animals in the CON group were fed a standard chow diet (5L80, Laboratory Diet; 3.03 kcal/g; carbohydrate, 71%; protein, 18.5%; fat, 10.5%; 500 g/day), whereas WD-AB animals consumed a Western diet [5B4L, Laboratory Diet; 4.14 kcal/g; carbohydrate, 40.8% (17.8% of total calories from high-fructose corn syrup); protein, 16.2%; fat, 43%/2% cholesterol wt/wt] (23, 25). The WD-AB group was subjected to 10 mo of dietary intervention (started at 2 mo of age) and 6 mo of chronic pressure overload (established at 6 mo of age) as previously described (9, 22, 23). At 1 yr of age, after a total of 10 mo of Western diet feeding to induce obesity and 6 mo of cardiac pressure overload, animals were anesthetized with a Telazol (5 mg/kg)-xylazine (2.25 mg/kg) mixture and maintained on propofol (6–10 mg·kg⁻¹·min⁻¹ with bolus as needed) for in vivo hemodynamic experiments before euthanasia and tissue harvest as previously described (23).

Protein extraction.

Left ventricle (LV), right ventricle (RV), right coronary artery (RCA), and triceps brachii long-head (TLH) tissues were carefully harvested, snap-frozen in liquid nitrogen, and stored at −80°C. Powdered tissue was homogenized in ice-cold RIPA Lysis and Extraction buffer (Thermo Fisher Scientific, Waltham, MA; Cat. No. 89901) containing freshly added protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA; Cat. No. 78441). Homogenates were incubated at 4°C under gentle agitation for 30 min and then centrifuged (15,000 g for 15 min at 4°C). Following centrifugation, the supernatants were collected and assessed for protein concentration using a BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA; Cat. No. 23227).

Western blot analysis.

Ten micrograms of total protein were loaded into SurePage, Bis-Tris 4–12% (GenScript, Piscataway, NJ; Cat. No. M00654). Proteins were transferred to a 0.45-µm PVDF membrane and blocked with 5% slim milk or BSA in Tris-buffered saline with 0.1% Tween 20 for 1 h at room temperature (RT) under gentle agitation. After washing step (three times for 10 min at RT), membranes were probed with primary antibodies overnight at 4°C under gentle agitation. The primary antibodies include anti-Hsp20 (HSPB6, Abcam, Cambridge, MA; Cat. No. ab184161; 1:5,000), phospho-Hsp20-Ser 16 (p-HSPB6-Ser16, Thermo Fisher Scientific, Waltham, MA; Cat. No. PA5-38278; 1:1,000), LC3B (MilliporeSigma, Saint Louis, MO; Cat. No. L7543; 1:1,000), p62 (MilliporeSigma, Saint Louis, MO; Cat. No. MABC32; 1:1,000), Beclin 1 (Abcam, Cambridge, MA; Cat. No. ab62557; 1:2,000), and calsequestrin (Thermo Fisher Scientific, Waltham, MA; Cat. No. MA3-913; 1:1,000). After washing step, the membranes were incubated with the appropriate secondary antibody at RT under gentle agitation. Membranes were then treated with chemiluminescence reagent Luminata Forte Western HRP Substrate (MilliporeSigma, Burlington, MA; Cat. No. WBLUF0100) and scanned with the KODAK 4000R Imager and Molecular Imaging Software. Calsequestrin (CSQ) and Coomassie blue staining images were used to report loading control quality. No statistical differences were observed in loading controls between groups; thus, absolute protein levels normalized relative to the CON group were used for analysis.

Statistical analysis.

All data are presented as means ± SD with significance reported at the P < 0.10 and P < 0.05 levels (4, 43). Cohen’s d sample effect size was calculated for significant differences reported at these levels to determine the magnitude and direction of change in protein autophagy markers (3, 33). Power analyses were conducted to determine the appropriate number of pigs to detect differences between groups as recommended by Kim and Seo (11) using the Sealed Envelope power calculator (https://www.sealedenvelope.com/power/continuous-superiority/). For input, we used arterial and ventricular protein data from intact female control and aortic-banded Yucatan (24) and Ossabaw swine (23) based on a range of mean differences of 20–30% and a pooled standard deviation for each group ranging from 0.10 to 0.17, with two-sided significance level set to P < 0.05 and power at 80%. This power analysis indicated n = 4 or 6 per group would be sufficient to detect group differences. Data analyses were performed using GraphPad Prism version 8, with unpaired two-sided Student’s t test used to compare the groups. For RV total HSPB6 protein, the P value was adjusted for unequal variance (as assessed by Levene’s test) using Welch’s t test.

RESULTS

Activation of HSPB6.

There was an increase in the p-HSPB6-Ser16 and total protein level of HSPB6, but not activation (p-HSPB6/HSPB6 ratio) of the WD-AB group compared with the CON group in the LV (Fig. 1, A and B). In the RV, p-HSPB6-Ser16, total HSPB6, and p-HSPB6/HSPB6 ratio were all increased in the WD-AB animals group (Fig. 1, C and D). In the TLH muscle (Fig. 1, E and F), there were no group differences in p-HSPB6-Ser16, total HSPB6, or p-HSPB6/HSPB6 ratio. A significant decrease in total HSPB6 protein level was observed in the RCA of the WD-AB animals group (Fig. 1, G and H). Interestingly, p-HSPB6-Ser16 protein bands were not detected in RCA samples. Cohen’s d effect size analyses revealed a large (>0.8), positive effect on HSPB6 protein level in the LV and RV (d = 1.6–2.2) and a large, negative effect in the RCA (d = −1.4). No differences were observed in CSQ protein levels (loading control) (Fig. 1, I and J). These data indicate tissue-specific differences in HSPB6 regulation between striated muscle samples (i.e., the heart and TLH) and the coronary vasculature.

Fig. 1.

Protein level of HSPB6, phosphorylation of HSPB6 at Ser16, and p-HSPB6/HSPB6 ratio are increased in the LV and RV of Western diet-fed, aortic-banded Ossabaw swine. Representative Western blots of phosphorylated HSPB6 at Serine 16 (p-HSPB6-Ser16) and total HSPB6 from the LV (A), RV (C), TLH (E), and RCA (G). Quantification of p-HSPB6-Ser16, total HSPB6, and p-HSPB6/HSPB6 ratio in the LV (B), RV (D), TLH (F), and RCA (H). Representative CSQ Western blot and its quantification are shown in (I) and (J), respectively. All data are means ± SD, n = 4 or 5, d = effect size. CSQ, calsequestrin; LV, left ventricle; RCA, right coronary artery; RV, right ventricle; TLH, triceps brachii long head.

Protein levels of the autophagy markers Beclin 1, LC3B, and p62.

Beclin 1 protein level did not change in the LV of the WD-AB group compared with the CON group (Fig. 2, A and B). Beclin 1 was increased in the RV of the WD-AB group compared with the CON group (Fig. 2, C and D). In the TLH (Fig. 2, E and F) and RCA (Fig. 2, G and H), no differences in Beclin 1 protein level were observed between groups. Cohen’s d effect size revealed a large, positive effect on Beclin 1 protein level in the RV (d = 1.4). Equal loading was demonstrated by Coomassie blue staining (Fig. 2, I and J).

Fig. 2.

Beclin 1 protein level is increased in RV of Western diet-fed, aortic-banded Ossabaw swine. Representative Western blots of Beclin 1 from the LV (A), RV (C), TLH (E), and RCA (G). Quantification of Beclin 1 in the LV (B), RV (D), TLH (F), and RCA (H). Representative membrane stained with Coomassie blue and its quantification are shown in (I) and (J), respectively. All data are means ± SD, n = 4 or 5, d = effect size. LV, left ventricle; RCA, right coronary artery; RV, right ventricle; TLH, triceps brachii long head.

In the LV, LC3B-I protein level (the endogenous nonlipidated form of the LC3B protein) was increased in the WD-AB group compared with CON group (Fig. 3, A and B). However, this was not associated with changes to LC3B-II or the LC3B-II/LC3B-I conversion ratio, which are both markers of autophagosome formation. There was no change in LC3B-I, LC3B-II, or the LC3B-II/LC3B-I conversion in the RV (Fig. 3, C and D) or TLH (Fig. 3, E and F) between groups. In the RCA, no difference in LC3B-I or LC3B-II protein level was observed (Fig. 3, G and H). However, the LC3B-II/LC3B-I conversion ratio was increased in the WD-AB group compared with CON group (Fig. 3, G and H), indicating possible activation of the autophagy conjugation system. Cohen’s d effect size revealed a large, positive effect on LC3B protein levels in the LV and RCA (d = 1.4–2.6). Calsequestrin protein level (loading control) was not different between groups (Fig. 3, I and J).

Fig. 3.

Microtubule-associated proteins 1A/1B light chain 3B (LC3B) conversion is increased in the RCA of Western diet-fed, aortic-banded Ossabaw swine. Representative Western blots of LC3B-I and LC3B-II from the LV (A), RV (C), TLH (E), and RCA (G). Quantification of LC3B-I, LC3B-II, and LC3B-II/LC3B-I ratio in the LV (B), RV (D), TLH (F), and RCA (H). Representative CSQ Western blot and its quantification are shown in (I) and (J), respectively. All data are means ± SD, n = 4 or 5, d = effect size. CSQ, calsequestrin; LV, left ventricle; RCA, right coronary artery; RV, right ventricle; TLH, triceps brachii long head.

An increase in p62 protein level in RV (Fig. 4, C and D) and TLH (Fig. 4, E and F) was found in the WD-AB group compared with the CON group, reflecting tissue- and chamber-specific accumulation of dysfunctional proteins and organelles. No changes in p62 expression were seen in the LV (Fig. 4, A and B) or RCA (Fig. 4, G and H). Cohen’s d effect size revealed a large, positive effect on p62 protein level in the RV and RCA (d = 1.5–1.7). Equivalent loading was again demonstrated by Coomassie blue staining (Fig. 4, I and J).

Fig. 4.

Sequestosome1/p62 (p62) protein level is elevated in the RV and TLH of Western diet-fed, aortic-banded Ossabaw swine. Representative Western blots of p62 from the LV (A), RV (C), TLH (E), and RCA (G). Quantification of p62 in the LV (B), RV (D), TLH (F), and RCA (H). Representative membrane stained with Coomassie blue and its quantification are shown in (I) and (J), respectively. All data are means ± SD, n = 4 or 5, d = effect size. LV, left ventricle; RCA, right coronary artery; RV, right ventricle; TLH, triceps brachii long head.

DISCUSSION

The results of this study indicate autophagic signaling is discontinuous in the WD-AB group when examining static protein levels of the autophagy markers HSPB6, LC3B, p62, and Beclin 1, with the added confounder that dysfunction is defined by tissue-specific variability. The finding that markers of autophagy measured in this study are generally not altered in a consistent manner between tissues demonstrates the difficulty of predicting autophagy status in a systemic manner, particularly in a disease state such as HF where the inherent physiological variability is likely increased. Our original hypothesis that activation of HSPB6 would be increased in this swine model of cardiometabolic HF was based on similar findings in human HF and more recent work in mice that linked its activation to increased fibrosis through an inflammatory process mediated in part by IL-6 (7, 27). In the same animals used in this study, we recently showed that the enhancement of molecular pathways known to contribute to systemic inflammation included activation of the tumor necrosis factor (TNF) signaling network in WD-AB animals, of which the activation of IL-6 was predicted with high confidence by ingenuity pathway analysis (23). Given the similar state of chronic inflammation commonly observed in parallel with HF present in this model, we certainly did not predict such discontinuity in the autophagy signaling process between tissues. HFpEF is a highly heterogeneous condition; thus, it is possible that our findings reflect this aspect of the disease, despite the idea that chronic metabolic stress and increased afterload as primary drivers of pathology could rationally be predicted to drive similar outcomes in multiple organ systems.

Autophagy status in the heart.

These ideas are corroborated by the results of this study that showed despite increased protein expression of p-HSPB6-Ser16 and HSPB6 in both the LV and RV, evidence supporting coherent linear activation of autophagy signaling was not observed in either cardiac chambers. In response to Western diet, chronic cardiac pressure overload, and an inflammatory process, HSPB6 protein levels were changed in the heart consistent with the rationale that HF increases its activation as indicated by the p-HSPB6/HSPB6 ratio, which could lead to increased autophagic activity (18, 27, 37). HSPB6 has also been previously shown to interact with Beclin 1 (18, 27), an indicator of increased autophagosome formation when upregulated (15). Although Beclin 1 protein was increased in the RV, consistent with the discontinuous nature of our findings, no changes in the lipidation of LC3B-I to LC3B-II were observed indicating a low rate of autophagosome formation. Furthermore, the increase in p62 protein level in the RV of the WD-AB group in the absence of increased autophagosome formation provides additional support that formation of the autophagosome and its subsequent degradation is impaired (1). Considered together, these data highlight dysregulation of the HSPB6/autophagy axis in the RV, characterized by an increase in HSPB6 and Beclin 1 that is not associated with downstream autophagosome formation or increased autophagosome cargo degradation as illustrated by no changes in the lipidation of LC3B-I to LC3B-II and increased p62 protein level, respectively.

In contrast, increased HSPB6 protein in the LV was not associated with any change in downstream autophagy markers indicating potentially important chamber-specific differences in regulation. In these same animals, we previously reported impaired mitochondrial function in the LV of the WD-AB group (23). The presence of dysfunctional mitochondria suggests reduced degradation and turnover of this organelle via autophagy and/or mitophagy (38), consistent with the current findings. In total, our findings indicate that autophagy signaling in the heart appears discontinuous at various stages of the process and occurs in a chamber-specific manner.

Autophagy status in the TLH skeletal muscle.

Results in the TLH also showed an increase in p62 protein level in the WD-AB group compared with the CON group, again indicative of impaired autophagic activity (20). Unlike the heart, this change was not associated with an increase in HSPB6 or other alterations to Beclin 1 or LC3B status. Although the HSPB6 protein is highly expressed in skeletal muscle from porcine (12, 39) and humans (8, 10), its direct relationship with autophagy in skeletal muscle is poorly understood. Previous reports have suggested that HSPB6 regulates insulin-induced glucose uptake in skeletal muscle of obese rats (40, 41), which is interesting because autophagy is energy sensitive (28), and overnutrition/obesity decreases autophagic activity in skeletal muscle (17, 26). Patients with HFpEF are often obese with concomitant metabolic syndrome (21, 42), and the WD-AB animals examined in this study display this HF phenotype (23). Discontinuous signaling between HSPB6 and p62 levels in the current study suggest that our large animal model of cardiometabolic HF may be useful to investigate autophagy in skeletal muscle, providing crucial information relevant to peripheral disease considered to be an important part of the HFpEF syndrome.

Autophagy status in the RCA.

In contrast to findings in the striated muscle samples examined from the heart and TLH, total HSPB6 protein expression in the RCA was decreased in the WD-AB group and associated with an increase in LC3B-II/LC3B-I ratio. These data demonstrate a different mechanism of discontinuous autophagy signaling in the coronary vasculature, whereby potential decreases in upstream activation are observed with indicators of increased autophagic activity. Although HSPB6 is known to induce vasorelaxation in the carotid (29) and coronary arteries (6, 36) in pigs, there are currently no studies to the best of our knowledge that have connected HSPB6 to autophagy in the RCA. Although autophagy is reported to regulate several vascular pathophysiological mechanisms (5), its overall impact on vascular health is controversial. Compounds inducing autophagy have been shown to prevent aging-induced arterial derangement (13, 14) and hypertension-induced arterial stiffening (19). Conversely, decreasing autophagy by administering 3-methyladenine or spautin-1 (i.e., prevention of autophagosome formation by targeting the VPS34 complex) induced stabilization of contractile phenotype of vascular smooth muscle (32). Previously, we demonstrated significant coronary microvascular impairment in these animals (23), and the dysregulation of autophagy in the RCA opens up a new avenue of exploration regarding its potential involvement in overall coronary vascular health. The disconnect between upstream HSPB6 protein and downstream LC3B-II/LC3B-I ratio in the RCA illustrates yet another tissue-specific example of discontinuous autophagic regulation that may have pathological relevance for HFpEF.

Limitations of the study and future directions.

Although this study measured some of the more well-known markers of autophagy in an effort to provide a baseline status in a novel large animal setting of cardiometabolic HF, we cannot rule out the possibility that regulation of the system may be occurring through a number of intrinsic autophagic mechanisms. Indeed, we recognize the assessment of autophagic flux (e.g., blockage of lysosomal fusion or inhibition of lysosomal acid-dependent hydrolases) is an important next step to fully elucidate the dynamic nature of the autophagy system in HF (20). Complementary approaches including the evaluation of VPS34/ATG14L/Beclin 1 complex activity, transmission electronic microscopy for visualization of autophagic vacuoles, and lysosomal content and activity will add important information regarding the continuum of autophagic signaling. Furthermore, a time-dependent evaluation of autophagy should be carefully addressed given the current findings provide insight into only one time point of disease. Although we cannot fundamentally resolve the overall importance of the changes in autophagy signaling observed herein to the general development of cardiovascular dysfunction in HF, we believe these findings provide an initial foundation by which to base future studies. Finally, despite the large variability intrinsic to the HF syndrome, a strength of this animal model is a notable pathologic phenotype supported by a strong statistical foundation. However, we acknowledge the potential for type II errors in the current study design could make it more difficult to detect subtle phenotypes that may be of physiological importance.

Conclusion.

Our data highlight discontinuous signaling between the activation of HSPB6 and other downstream autophagy markers, highlighting discrepancies with previous work linking the two mechanisms. The inconsistent manner of tissue-specific autophagy regulation observed in this study provides a rationale for the continued pursuit of understanding the relevance of these differences. As this is the first study to examine HSPB6 activation in a large animal model with relevance to HFpEF, we hypothesize that complete activation of this important cellular recycling process may be required to realize any therapeutic potential that manipulation of the autophagy system may hold for the treatment of HF.

GRANTS

This study was supported by a grant from the University of Missouri Research Board (to C. A. Emter and R. S. Rector); National Institutes of Health (NIH) Grant R01 HL112998 (to C. A. Emter); American Heart Association (AHA) Postdoctoral Fellowship Award 18POST33960472 (to K. A. S. Silva); Veterans Affairs Merit Grant I01BX003271 (to R. S. Rector); NIH Grants K01 HL125503 (to J. Padilla) and K01 AG041208 and R01 HL136292 (to T. L. Domeier); and AHA Award Postdoctoral Fellowship Award 16POST27760052 (to T. D. Olver).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.V.L., T.D.O., R.S.R., and C.A.E. conceived and designed research; K.A.S.S. performed experiments; K.A.S.S., E.V.L., R.S.R., and C.A.E. analyzed data; K.A.S.S., E.V.L., R.S.R., and C.A.E. interpreted results of experiments; K.A.S.S. prepared figures; K.A.S.S. drafted manuscript; K.A.S.S., E.V.L., T.D.O., T.L.D., J.P., R.S.R., and C.A.E. edited and revised manuscript; K.A.S.S., E.V.L., T.D.O., T.L.D., J.P., R.S.R., and C.A.E. approved final version of manuscript.