Doxycycline decreases amyloidogenic light chain-induced autophagy in isolated primary cardiac myocytes

Abstract

Background:

Immunoglobulin light chain (AL) cardiac amyloidosis is characterized by extracellular deposition of amyloid fibrils in the heart and is potentially fatal. Untreated, it manifests clinically as heart failure with a precipitous decline and a median survival of <6 months. AL cardiac amyloidosis is associated with impaired extracellular matrix homeostasis in the heart with increased matrix metalloproteinase (MMP) levels. This commmunication provides novel insights into a potential role for doxycycline, a non-selective MMP inhibitor in AL cardiac amyloidosis.

Methods/results:

Adult rat ventricular myocytes stimulated with AL (obtained from cardiac amyloidosis patients) increased MMP-2 and MMP-9 activities (P < .05); the expression of autophagy marker microtubule associated protein 1 LC-3 isoform II (LC3-II) (P < .01), and the autophagy-related proteins ATG-4B (P < .05) and ATG-5 (P < .05) as compared to untreated cardiomyocytes. Doxycycline abrogated MMP activities (P < .0001) and decreased AL-induced autophagy via ATG-5 (P < .05).

Conclusions:

These in vitro studies demonstrated that doxycycline, in addition to inhibiting MMP, also modulated AL-induced autophagy in cardiomyocytes and provide potential insights for future therapeutic targets for AL-induced proteotoxicity. Novel therapies for cardiotoxicity and heart failure in AL cardiac amyloidosis remain an important unmet need.

1. Introduction

Cardiac amyloidosis is a rare systemic disease where extracellular proteinaceous material, known as amyloid, is deposited in the heart. Clinical manifestations are predominantly heart failure (HF). Amyloid fibrils derived from precursor proteins immunoglobulin-derived light chains (AL) and transthyretin proteins are the most common that exhibit a cardiac phenotype. Overproduction of immunoglobulin-derived AL causes AL amyloidosis, a plasma cell dyscrasia. Similarly, lack of AL clearance triggers amyloid deposition in many organs, including the heart, which occurs in up to 50% of all AL patients [1].

If the underlying plasma cell dyscrasia is untreated, the prognosis for AL cardiac amyloidosis is dismal with a median survival of <6 months [1–3]. Thus definitive therapy for AL is directed at eliminating the underlying malignant plasma cell clone. However AL amyloidosis patients with symptomatic HF and cardiac involvement are often excluded because of the high treatment-related morbidity and mortality [1;4]. Additionally, anti-plasma cell therapies aimed to eradicate cells that produce AL amyloid protein have no effect on the pre-formed amyloid which is deposited in the heart. Therefore, despite achieving hematologic response, many of these patients still succumb to organ failure, including HF. As such there is a significant need to understand the mechanisms of cardiotoxicity and for new pharmacological approaches to address the amyloid deposition that results in symptomatic HF.

Tissue destruction and organ dysfunction ensues as AL is deposited in the heart. At the cellular level, this induces oxidative stress, lysosomal dysfunction, and disrupts autophagic flux causing cardiomyocyte death [5]. AL deposition in the heart also alters the extracellular matrix milieu by activating collagen degradation which is under the control of the matrix metalloproteinases (MMPs). We previously showed that circulating MMP-2 and MMP-9 levels were elevated in patients with AL cardiac amyloidosis, and also correlated with markers of cardiac stress (increased brain natriuretic peptide (BNP) and troponin I levels) [2,3].

Doxycycline has purported beneficial effects in proteinopathies and reduced amyloid fibrillogenesis in preclinical studies [6]. A single centre human study showed that the addition of doxycycline to standard chemotherapy may reduce mortality in cardiac AL amyloidosis [7]. There are also several clinical trials (NCT03474458; NCT02207556; NCT03401372) presently investigating the effect of doxycycline treatment in AL amyloidosis patients. We therefore sought to investigate doxycycline’s mechanisms of action in AL cardiac amyloidosis.

2. Methods

2.1. Human light chain protein

Light chain protein was obtained from patients with AL cardiac amyloidosis and this was approved by the Boston University Medical Center (BUMC) Institutional Review Board. Urine purification was performed by the Boston University Amyloidosis Center as previously described [5].

2.2. Cardiomyocyte isolation and culture

Adult rat ventricular myocytes (ARVM) were isolated as previously described [8]. Briefly, ARVM (90–95% purity) were collected from the hearts of adult male Sprague-Dawley rats and plated non-confluently on laminin-coated (1 μg/cm²; Invitrogen, Carlsbad, CA) plastic culture dishes at a density of 30–50 cells/mm². Cells were maintained at 37 °C in Dulbecco’s Modified Eagle Medium (DMEM, Invitrogen) containing 2 mg/mL BSA, 2 mmol/L L-carnitine, 5 mmol/L creatinine, 5 mmol/L taurine (Sigma-Aldrich, St. Louis, MO), 100 IU/mL penicillin, and 10 g/mL streptomycin (Invitrogen).

2.3. Drug treatments

ARVM were treated with AL for 24 h. In some experiments ARVM were also treated with 50 μg/mL doxycycline (DOXY, Sigma-Aldrich) for 30 min. DOXY was diluted in water to a concentration of 1mg/mL and kept in the dark at −20°C. DOXY was freshly diluted from the stock for every treatment. A dose response was performed using DOXY 5, 10, 50 and 100 μg/mL based on prior DOXY doses [6]. Controls were treated with the vehicle substance.

2.4. Cell viability

After treatment with DOXY, cardiomyocytes were assessed for viability using the MTT assay. ARVM were treated with varying doses of DOXY. MTT solution (50μL/mL) was added and protected from light. The media was removed and DMSO was added to solubilize the cells. Sorensen’s buffer was added to DMSO, and 100 μL was transferred to a 96-well plate. Absorbance was measured (A570) using a plate reader, and cell viability was determined as a percentage of the reading of the control group.

2.5. Zymography

MMP activities were determined by in-gel zymography as previously described [8].

2.6. Immunoblot

Isolated ARVM were subjected to western blotting for protein expression. Following treatments, ARVM were lysed, scraped, and collected in cold lysis buffer (Cell Signaling Technology, Danvers, MA). Samples were prepared and subjected to SDS-PAGE (12% tris-glycine gels, Lonza, Rockland, ME) and transferred to PVDF membranes using a wet electroblotting system (Bio-Rad). Membranes were then probed with the following antibodies: anti-LC3 (1:2500); anti-ATG4B (1:2500) and anti-ATG5 (1:2500) (Novus Biologicals, Littleton, CO). Goat anti-rabbit or goat anti-mouse horseradish peroxidase conjugated (1:5000) were used as secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). Blots were detected with ECL™ Western Blotting Detection Reagent (Buckinghamshire, England), and chemiluminescence was quantified by densitometry using ImageJ measuring software (National Institutes of Health). Protein expression was normalized for equal protein loading using Anti-GAPDH monoclonal antibody as loading control (Thermo Scientific. Rockford, IL).

2.7. Statistical analysis

Normality of distributions was verified by Kolmogorov-Smirnoff normality test. Differences between groups were analysed by one-way ANOVA followed by a Newman-Keuls Multiple Comparison Test. P ≤ .05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism (GraphPad Software, Inc., La Jolla, CA).

3. Results

3.1. MMP-2 and MMP-9 activities were increased in ARVM stimulated with amyloidogenic light chain

ARVM were incubated for 24 h with amyloidogenic light chain isolated from patients with cardiac amyloidosis (AL). Both MMP-2 and MMP-9 activities were significantly increased in cardiomyocytes stimulated with AL light chain compared to untreated (vehicle) cardiomyocytes by 1.6- and 1.3-fold, respectively (P < .05 vs. vehicle for both) (Fig. 1).

Fig. 1.

Doxycycline, a non-specific inhibitor of metalloproteinases (MMPs), decreased amyloidogenic light chain-induced autophagy in isolated rat ventricular cardiomyocytes. Graphic presentation and representative zymogram of MMP-2 and MMP-9 activities (n = 6 experiments). Data shown as mean±SEM. AL, amyloidogenic light chain; DOXY, doxycycline.

3.2. Doxycycline effects on MMPs

We next examined the role of doxycycline (DOXY) on MMP inhibition in vitro, but first perfomed a dose-response for DOXY. DOXY reduced MMP-2 activity in ARVM in a dose-dependent manner. To exclude cell mortality as a reason for reduced MMP-2 activity, we performed the MTT viability assay which showed no significant cell mortality up to the DOXY dose of 50 μg/mL, whereas at 100 μg/mL doxycycline reduced cell viability (data not shown). Therefore, the reduced MMP-2 activity seen up to 50 μg/mL was not due to reduced cell numbers. DOXY (50 μg/mL) decreased both MMP-2 and MMP-9 activities in ARVM stimulated with AL (63 and 60% reduction, respectively vs. AL alone; P b .0001) (Fig. 1).

3.3. Inhibition of MMPs by doxycycline was accompanied by a change in autophagic markers in AL-treated cardiomyocytes

As previously shown [5], the autophagy marker, microtubule associated protein 1 LC3 isoform II (LC3-II) protein expression was markedly increased in AL-treated cardiomyocytes (2.6-fold increase vs. vehicle; P < .05). This may indicate either increased induction of autophagy or defective clearance. Additionally autophagy-related 4B cysteine peptidase (ATG4B, aka autophagin-1) and ATG5 protein expression were also increased in cardiomyocytes after treatment with AL by 2.7- and 1.7-fold, respectively (P < .05 vs. vehicle for both).

Treatment with DOXY resulted in a 38% reduction in LC3-II expression vs. AL alone in cardiomyocytes (P < .01) and importantly although DOXY abrogated ATG5 protein levels by 25% (P < .05 vs. AL alone), there was no significant affect on ATG4B protein expression vs. AL alone (Fig. 2).

Fig. 2.

Amyloidogenic light chain induced autophagy in isolated rat ventricular cardiomyocytes - doxycycline decreased only LC3-II and ATG-5 expression. Graphic presentation and representative immunoblot demonstrating protein levels of LC3-II, ATG4B and ATG5. GAPDH was used as loading control. (n=5 experiments) Data shown as mean±SEM. AL, amyloidogenic light chain; DOXY, doxycycline.

4. Discussion

In order to explore therapies for AL cardiac amyloidosis with HF, we investigated at the cellular level our prior clinical findings which demonstrated a relationship between AL cardiac amyloidosis and MMP activation [2;3]. In this brief communication, ARVM stimulated with AL light chains increased both MMP-2 and MMP-9 activities, as expected. Novel findings here include the increased expression of autophagic proteins ATG4B and ATG5 in AL-treated cardiomyocytes, supporting prior findings of impaired autophagy in cardiomyocytes treated with amyloidogenic light chains [5]. Doxycycline abrogated AL-induced MMP activities and decreased AL-induced autophagy via ATG5 but not ATG4B.

Cardiac deposition of AL amyloidogenic protein results in a precipitous decline in clinical status once symptoms of HF occur. There are no therapies for HF due to AL cardiac amyloidosis and indeed many guideline-directed therapies for HF are contraindicated [1]. Novel therapies should alter not only the amyloid fibrils and proteotoxic, prefibrillar misfolded proteins but also the implicated signaling pathways. In a single-centre clinical study, doxycycline demonstrated anti-amyloid effects when given together with anti-plasma cell chemotherapy [7].

In our study, MMP-2 and MMP-9 activities were significantly reduced with the addition of doxycycline to AL-stimulated cardiomyocytes compared to cardiomyocytes treated with AL alone. Others have also shown that doxycycline disrupts amyloid fibrils, and disaggregates amyloid deposits with a concomitant decrease in MMP-9 in transthyretin amyloidosis but not in AL amyloidosis [9,10]. MMP-9 levels are elevated in the serum, heart and kidneys in AL amyloidosis [3,6] and it is hypothesized that this increase may contribute to the pathogenesis of amyloid-induced organ damage [11]. However, it cannot be excluded that increased activation of MMPs in AL cardiac amyloidosis might reflect a compensatory response to AL light chain aggregation, and in this scenario reducing MMP activity may not be beneficial.

Our findings also demonstrated that doxycycline treatment decreased LC3-II expression in cardiomyocytes stimulated with AL. Autophagy is a dynamic process that allows cells to adapt to changing conditions, to mobilize energy reserve and optimize quality control of proteins and organelles and varied stimuli induce differential responses and becomes maladaptive in disease depending on the type and duration of the stress conditions [12]. Insights from murine studies suggest that in other forms of amyloid deposition, such as brain ischemia, increases neuronal autophagy and amyloid-β accumulation likely contribute to dementia. These authors suggest that brain autophagy might be a possible therapeutic target in patients with strokes [13]. Similarly, in AL cardiac amyloidosis, autophagy is considered pathological and its modulation by doxycycline is novel as there are, at present, very few therapeutic options available. We propose that this reduction is mediated, at least in part by ATG5, since ATG5 expression was decreased after doxycycline treatment in AL-stimulated cardiomyocytes. However, the relationship between the pleiotropic actions of doxycycline, such as MMPs inhibition and the mitigation of autophagy, were not investigated in this study and requires further investigation. Others have demonstrated in breast cancer cells that doxycycline decreased LC3-II expression and inhibited both cellular proliferation and the profibrotic process known as epithelial-mesenchymal transition [14]. There is also evidence of ATG5 regulation independently of ATG4B. Recent studies demonstrate that ATG5 modulates the immune system and crosstalks with apoptosis [15]. In fact, there is evidence that ATG5 plays a role in the regulation of diseases such as Crohn’s disease [16], systemic lupus erythematosus [17] and central nervous system autoimmune diseases, such as multiple sclerosis [18]. Thus future in-depth mechanistic insight into the ATG5 pathway may inform potential therapies in AL cardiac amyloidosis.

Antimicrobial doses of doxycycline are usually between 100 and 200 mg every 24 h and achieve plasma concentrations of 1.5 to 9.3 mg/L in healthy volunteers [19,20], however concentrations as high as 20 mg/L are found in elderly patients [21]. In addition to the dose-response performed in this study, the 50 mg/L (or 50μg/mL) dose was also selected based on prior in vitro studies where 15 and 250 mg/L showed an inhibitory dose-response effect on amyloid fibril formation and disruption [6]. Doxycycline has a wide safety margin and signs of increased toxicity have not been reported [21,22], but further safety and tolerability studies are warranted in larger animal models.

In summary, novel AL therapies for HF in AL cardiac amyloidosis are an important clinical need since patients usually succumb to clinical HF despite treatment [4] of the underlying plasma cell dyscrasia. Improved understanding of the cardiotoxic mechanisms of light chains in AL cardiac amyloidosis is needed. Here, we demonstrate that AL directly induces MMP activities and impairs autophagic flux in cardiomyocytes. Importantly, the non-specific inhibitor of MMPs, doxycycline restores normal autophagic levels in cardiomyocytes in a process that is mediated, at least in part by an ATG5-dependent mechanism.