Osteopontin Expression in Cardiomyocytes Is Increased in Pediatric Patients With Sepsis or Pneumonia

Sepsis and pneumonia are major causes of death in the United States, and their pathophysiology includes infection with inflammation and immune dysfunction. Both sepsis and pneumonia cause cardiovascular dysfunction. The expression of Osteopontin (OPN) in cardiomyocytes of patients with sepsis or pneumonia, and its role the induced cardiac dysfunction have not been thoroughly investigated. OPN is a matricellular protein synthesized by multiple diseased tissues and cells including cardiomyocytes. Here, we studied the expression of OPN protein using immunofluorescence in human myocardial autopsy tissues from pediatric and mid age or elderly patients with sepsis and/or pneumonia. Fourteen human myocardial tissues from six pediatric patients and eight mid-age or elderly patients were studied. Immunofluorescence was used to investigate the expression of OPN in paraffin-embedded heart sections co-stained with the myocyte markers Actin Alpha 1 (ACTA1) and Myosin Light Chain 2 (MLC2). A quantitative analysis was performed to determine the number of ACTA1 and MLC2 positive cardiomyocytes that express OPN. The results showed that OPN expression was significantly increased in cardiomyocytes in the hearts from pediatric patients with sepsis and/or pneumonia (N = 3) relative to pediatric patients without sepsis/pneumonia (N = 3), or adult to elderly patients with sepsis/pneumonia (N = 5). Among the older septic hearts, higher levels of cardiomyocyte OPN expression was seen only in conjunction with severe coronary arterial occlusion. This is the first study to document increased OPN expression in cardiomyocytes of pediatric subjects with sepsis or pneumonia. Our findings highlight a potentially important role for OPN in sepsis- or pneumonia-mediated cardiac dysfunction in pediatric patients.

Pneumonia is an inflammatory disease of the lung caused by bacteria, viruses, or fungi, and it is responsible for high morbidity and mortality in adults and children (Mattila et al., 2014;le Roux and Zar, 2017). Some reports have shown that pneumococcal pneumonia can lead to cardiac complications including heart failure, arrhythmias, and myocardial infarction (Musher et al., 2007;Corrales-Medina et al., 2012). Besides these studies on increased levels of OPN in the plasma of patients with interstitial (Kadota et al., 2005) and community-acquired pneumonia (Chang et al., 2016), and in the plasma and lung of mice infected with Streptococcus pneumonia (van der Windt et al., 2011), the cardiomyocyte expression of OPN and its role in pneumonia-mediated cardiac dysfunction have not been investigated.
This study quantifies the expression of OPN protein using immunofluorescence in human myocardial autopsy tissues from pediatric and elderly patients with sepsis or pneumonia.

Human Myocardial Tissue Samples
A total of 14 human myocardial and two control kidney autopsy tissues were obtained from the Department of Pathology at the University of Miami Leonard M. Miller School of Medicine. The myocardial tissues belonged to six pediatric (<2 years old) patients and eight adult or elderly (48-77 years old) patients. Table 1 lists demographic data for autopsy cases used for this study. These kidney tissues (especially the diseased one) were used as positive controls for OPN expression as it is known that renal tubular cells express high levels of OPN under pathological conditions (Ding et al., 2018).
This study was carried out in accordance with the recommendations of the University of Miami Institutional Review Board (IRB protocol # 20180439). All autopsy specimens were de-identified and there was no need for consent forms.

Staining and Image Acquisition
Immunofluorescence was used to investigate the OPN expression in paraffin-embedded sections. Samples were incubated for 45 min at 70 degrees, then dewaxed with two 5-min xylene washes and hydrated by 3-min graded ethanol washes of 100% (twice), 95%, 80%, and 70% followed by two 4-min water immersions. Antigen retrieval was then done by steaming the slides for 75 min in 1X Citrate Antigen Retrieval Buffer (Ab93678). Samples were rinsed with PBS then permeabilized with 0.2% Triton X-100 (Sigma-Aldrich) in PBS for 30 min, followed by blocking with 10% Donkey serum in TBST for 45 min. The samples were then incubated overnight at 4 • C with three primary antibodies, Anti-Alpha Skeletal Muscle Actin, Alpha-Sr 1 (ACTA-1) (Abcam Ab28052; dilution 1:200), anti-hOPN (R&D AF1433; dilution 1:50), and anti-Myosin Light Chain (MLC2) (Abcam Ab92721; dilution 1:200) in blocking solution. ACTA-1 and MLC2 were used as cardiomyocyte markers. The following day the slides were washed with PBS then incubated for 1 h at room temperature with conjugated secondary antibodies Donkey anti-Mouse IgG, Alexa Fluor 568 (ThermoFisher Scientific A-10037; 1:400), Donkey anti-Goat IGG, Alexa Fluor 647 (ThermoFisher Scientific A-21447, 1:400), and Donkey anti-Rabbit IgG, Alexa Fluor 488 Samples are listed in the order of highest to lowest OPN expression in the cardiomyocytes in 14 heart autopsy samples and in the renal tubules in two kidney autopsy samples used as controls for OPN staining.
(ThermoFisher Scientific A-21206; 1:400) in blocking solution. DAPI was used to stain nuclei before mounting with ProLong Gold Antifade (Invitrogen P36934) and coverslips. After 2 days, the slides were scanned at 20 × magnification using the Olympus VS120-L100 Virtual Slide Microscope (Tokyo, Japan). At least five slides per specimen were stained to confirm results. Representative confocal images (shown in Figures 1-9) were captured at 40x magnification and 0.6x digital zoom using Z-stacking on a Zeiss LSM710 confocal microscope.

Image Analyses and Quantification of OPN Positive Myocytes
All immunostaining scanned images were analyzed using Olympus OlyVIA 2.9 software. A quantitative analysis was performed independently by two investigators to manually count the number of ACTA1 and MLC2 positive cardiomyocytes that express OPN in representative fields. Five images per sample were counted and averaged.

RESULTS
To quantify the OPN protein expression, immunofluorescence was performed in paraffin-embedded sections of myocardial biopsies obtained from patients with different conditions (described in Table 1). All myocardial tissues in addition to two control kidney tissues were stained for OPN. As expected, the pediatric diseased kidney had a very high expression of OPN in the renal tubules ( Figure 1A) and the adult healthy kidney had a moderate expression of OPN in the renal tubules ( Figure 1B). All myocardial tissues were additionally co-stained with the myocyte markers ACTA1 and MLC2 to unequivocally identify the OPN positive cardiomyocytes.  From the 14 myocardial tissues analyzed, the three pediatric hearts with sepsis and/or pneumonia, showed the highest expression of OPN (Figures 2, 3, 5). Except for the heart from the old patient with sepsis and several severe coronary occlusions FIGURE 4 | OPN expression is increased in the cardiomyocytes from the elderly patient with septic shock and coronary artery disease. In the elderly patient with septic shock and >75% occlusion in the left anterior descending (LAD) artery and 100% occlusion in the right circumflex artery (RCA), OPN (white in A) is highly expressed in MLC (green in B), and ACTA1 (red in C) positive cardiomyocytes (arrows). (D) is the merged panel. Scale bar = 50 µm.  (Figure 4), all the other hearts from the pediatric patients and the mid-age and old patients had barely had any OPN expression in the cardiomyocytes (Figures 6-9). The microscopic images   (Figures 1-9) are presented in the order of high to low expression of OPN. The quantification of the number of OPN positive cardiomyocytes in each of the fourteen myocardial tissues and   Frontiers in Physiology | www.frontiersin.org in the three groups (pediatric with or without sepsis/pneumonia and elderly with sepsis/pneumonia) is shown in Figure 10.

DISCUSSION
Our study shows that OPN expression was increased in cardiomyocytes of pediatric patients with sepsis or pneumonia but not in those of elderly patients without evidence of comorbid Coronary Artery Disease. According to the Center for Disease Control and Prevention, in the United States 270,000 people die from sepsis each year (Rhee et al., 2017). For newborns sepsis was the seventh leading cause of infant death in 2016 (Kochanek et al., 2017), followed by influenza and pneumonia (Kochanek et al., 2017). In sepsis, respiratory tract infections, particularly pneumonia, are the most common and associated with the highest mortality (Mayr et al., 2014). Both sepsis and pneumonia have been reported to induce heart dysfunction (Merx and Weber, 2007;Singer, 2007, 2013;Corrales-Medina et al., 2011;Brown et al., 2014).
Myocardial dysfunction during sepsis and pneumonia has been variously associated with increased circulating cytokines and prostanoids, endothelin upregulation, impaired oxygen utilization, and increased levels of C-reactive protein (Kumar et al., 1996;Merx and Weber, 2007;Musher et al., 2007;Hunter and Doddi, 2010;Singanayagam et al., 2012). Increased OPN levels in plasma have also been reported (Kadota et al., 2005;Vaschetto et al., 2008), but the corresponding levels in cardiac tissues have not been investigated. OPN is overexpressed in the heart coincident with myocardial infarction (Trueblood et al., 2001), hypertrophy (Graf et al., 1997;Li et al., 2017), and myocarditis (Szalay et al., 2009) in a cell type-and stimulantdependent manner (Singh et al., 2010). Mechanistically, interleukin-1β combined with interferon-γ (IFN-γ) increase OPN expression in cardiac microvascular endothelial cells (Singh et al., 1995), while angiotensin II alone or combined with TNF-α or interleukin-1β increase OPN gene expression in cardiac fibroblast (Xie et al., 2004), and glucocorticoid hormone can increase the expression of OPN in cardiomyocytes and cardiac microvascular endothelial cells (Singh et al., 1995). Graf et al. (1997) showed that endothelin-1 and norepinephrine induce OPN expression in cultured rat neonatal cardiac myocytes. Interestingly high levels of circulating endothelin-1 have also been linked with sepsis (Bocking et al., 1979;Shindo et al., 1998;Rudiger and Singer, 2007), providing a possible explanation for our results in the pediatric specimens. In our studies the highest cardiomyocyte OPN levels were found in patient (H4), a baby born prematurely who died due to necrotizing enterocolitis. The subject had edematous heart muscle due sepsis and multi-organ failure, and patient (G), a child with fungal pneumonia. In both subjects the increased cardiomyocyte OPN expression may be linked with coincident inflammation. Our results are consistent with a central role for infection/inflammation in the activation of OPN expression in the pediatric heart. Notably, the children with full term ompholocele, holoprosencephaly, or premature birth, but no infection, showed only low-level cardiomyocyte OPN.
Interestingly, in the hearts of the older patients with sepsis, few cardiomyocytes expressed OPN. Among these, only the heart with severe coronary arterial occlusions showed high levels of cardiomyocyte OPN, consistent with the known link between Coronary Artery Disease, myocardial infarction, and OPN (Trueblood et al., 2001). It seems likely that age is a determining factor in the possible association between sepsis and cardiac OPN. Consistent with this, Luce et al. (2007) reported that cytokine responses differ widely between neonates and adults. Paliwal et al. (2012) reported higher OPN expression in macrophages of aged mice with skeletal muscle injury. Conversely, Schultz et al. (2002Schultz et al. ( , 2004 reported exaggerated inflammatory responses in term and preterm infants compared to adults with attenuation of the proinflammatory reaction during infection that can lead to severe organ dysfunction in infants. Developing neonatal cardiomyocyte also differ from adults and there may be different responses to sepsis and inflammation (Luce et al., 2007).
Patients with pneumococcal pneumonia are at increased risk for concurrent acute cardiac events (Musher et al., 2007). Increased levels of inflammatory cytokines, C-reactive protein, and fibrinogen, along with increased myocardial demand for oxygen and lowered blood oxygen levels, are among the factors that have been reported to lead to cardiac dysfunction in patients with bacterial pneumonia. Based on our findings, increased OPN level may contribute to these pneumonia-caused cardiac events. However, further studies with larger sample sizes are needed to confirm relationships between cardiomyocyte OPN, inflammation, and age.

CONCLUSION
In this study, we report markedly increased levels of OPN in cardiomyocytes of pediatric patients with sepsis or sepsis and pneumonia. Both sepsis and pneumonia involve activation of the immune system, inflammatory responses followed by cardiac dysfunction. This is the first study to report such increased cardiomyocyte OPN expression in pediatric sepsis or pneumonia. Our findings highlight a potentially important role for OPN in sepsisor pneumonia-mediated cardiac dysfunction in pediatric patients.

ETHICS STATEMENT
This study was carried out in accordance with the recommendations of the University of Miami Institutional Review Board (IRB protocol # 20180439). The study was exempt because the samples used were all autopsy samples.

AUTHOR CONTRIBUTIONS
CI, KP, and KJ-W performed the experiments. CI and KP analyzed the results. SG provided the pathology samples, clinical information, and data interpretation. CI and LS wrote the manuscript.

FUNDING
This work was supported by the following grants to LS: National Institute of Health (R56HL132209 and 1R01HL140468) and the Miami Heart Research Institute.