Cardiac Tissue Damage in a Female Animal Post-COVID Model: Relevance of Chemokine-Mediated Inflammation

This study was conducted in accordance with the Ethical Committee of Animal Experimentation of INIA-CSIC and by the Division of Animal Protection of the Comunidad de Madrid (PROEX 115.5-21). It followed the guidelines for the Care and Use of Laboratory Animals of the European Community (European Directive 2010/63/EU). Thus, the local Research Ethics Committee of Universidad Rey Juan Carlos (Madrid, Spain, code: ENM159/202311202122021, date of approval: 22 January 2022) approved the study design.

A sample of C57BL/6 female mice was used for transgene production, i.e., transgenic mice expressing the human angiotensin-converting enzyme 2 (ACE2) receptor under the human cytokeratin 18 (K18) gene promoter's control (hACE2) [22]. Subsequently, 6-week-old, 20–25 g, female hACE2 mice were anesthetized with 5–2% of isoflurane (+0.5–1 L/min) and intranasally inoculated with 50 μL of the Omicron variant of SARS-CoV-2 (lineage BA.1.17, 104 TCID50/mouse) as previously described [21]. All experimental procedures related to the animal model were conducted in the biosafety level 3 facilities at the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA-CSIC), Madrid (Spain).

An initial sample of twelve (n = 12) hACE2 female mice infected with the SARS-CoV-2 Omicron variant (infected group) and eleven (n = 11) uninfected hACE2 animals were used. All mice received food and water ad libitum and were monitored daily for clinical signs and body weight. Animals that showed advanced clinical features of SARS-CoV-2 infection and a significant deterioration in their health status (e.g., weight loss exceeding 20% of initial body weight or severe symptoms such as lethargy, inactivity, hunched posture, or markedly abnormal respiration) were anesthetized and sacrificed by cervical dislocation. All animals that survived and reached the end of the experiment (28 days post-infection) were sacrificed, and cardiac tissue was obtained for biological study (Figure 1). The period of 28 days post-infection was chosen because it correlates to 3 years in the human lifespan, which corresponds to a post-COVID condition [23].

Following mouse sacrifice, cardiac tissue was rapidly dissected and immediately frozen at −80 °C. Cardiac tissue was homogenized in ice-cold RIPA buffer supplemented with a protease inhibitor cocktail (Roche, Spain) and using a TissueLyser II system (Qiagen, Hilden, Germany) for protein extraction. All homogenates were centrifuged, and the supernatant was extracted. Total protein concentrations were determined using the NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Protein expression analysis by Western blot was performed using cardiac tissue from 7 surviving animals in the infected group and a similar sample in the uninfected group

For electrophoresis, 40 µg of protein from the heart was loaded onto a precast Mini-Protean^® TGXTM gel (Bio-Rad, Madrid, Spain) and transferred to a polyvinylidene fluoride (PVDF) membrane (Bio-Rad, Madrid, Spain). Membranes were blocked with 3% non-fat dry milk at room temperature for 1h and then incubated at 4° overnight with the primary following antibodies: Toll-like receptor-4 (TLR4) 1:500 (Novus Biosciences, Toronto, ON, Canada, Cat#NB100-56566); Myeloid differentiation primary response gene 88 (MyD88) 1:500 (Abcam, Cambridge, UK, Cat#2064); Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-κβ p65) 1:1000 (Abcam, Cat#16502); tumour necrosis factor-alpha (TNF-α) 1:1000 (Abcam, Cambridge, UK, Cat#34674); IL-6 1:500 (Abcam, Cat#290735); IL-1β 1:1000 (Abcam, Cambridge, UK, Cat#283818); IL-18 1:750 (Invitrogen, Waltham, MA, USA, Cat#PA5-79482); integrin alpha D (CD11d) 1:1000 (Abcam, Cambridge, UK, Cat#231534); iNOS 1:2000 (Invitrogen, Waltham, MA, USA, Cat#PA3-030A); PAI-1 1:500 (Abcam, Waltham, MA, USA, Cat#7205); Connexin 43 1:10,000 (Cell Signaling Technology, Waltham, MA USA, Cat#3512). After primary incubation, membranes were incubated for 1 h at room temperature with the specific secondary antibody (Goat anti-Mouse IgG (H+L) Secondary Antibody, HRP (Thermo Fisher Scientific, Inc., Waltham, MA USA, Cat#31430) or goat anti-rabbit IgG (H+L) secondary antibody, AP (Thermo Fisher Scientific, Inc., Waltham, MA USA, Cat#31340). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) 1:5000 (Abcam, Waltham, MA, USA, Cat#8245) was used as a loading control with secondary antibody (1:10,000).

Protein bands were visualized using the Clarity™ Western ECL substrate (Bio-Rad^®, Madrid, Spain) and detected with a Chemidoc™ XRS+ system or ChemiDoc™ MP Imaging System. Band intensities were analyzed via densitometry using ImageLab™ software 3.0 (Bio-Rad^®, Madrid, Spain) and normalized to GAPDH.

Statistical analyses were conducted using GraphPad Prism software, version 8.0 (San Diego, CA, USA). The normality of quantitative data was assessed using the D'Agostino-Pearson test. Results are presented as mean ± standard error of the mean (S.E.M.). Biomarker levels were compared using either an unpaired Student's t-test or the Mann–Whitney U test, depending on the distribution of the data. In cases where outliers were detected, they were omitted for statistical comparison.

To explore the involvement of innate immune activation in cardiac inflammation, we analyzed the expression of the TLR4/MyD88/NF-κB signaling pathway in the heart. No significant differences in TLR4 (p = 0.340, Figure 2A), MyD88 (p = 0.410, Figure 2B) or NF-κB p65 (p = 0.780, Figure 2C) expression levels in cardiac tissue between infected and non-infected mice were observed.

To characterize the cardiac pro-inflammatory profile in infected female mice, the expression of IL-6, IL-18, IL-1β, TNF-α and CD11d was examined. A significant increase (p = 0.028) in the pro-inflammatory cytokine IL-6 was observed in infected female mice when compared with non-infected mice (Figure 3A). In contrast, there were no significant differences in the expression of cardiac IL-18 (p = 0.548, Figure 3B), IL-1β (p = 0.455, Figure 3C), and TNF-α (p = 0.125, Figure 4A) between infected and non-infected mice. Additionally, a significant increase (p = 0.016) in CD11d expression (Figure 4B) was also observed in infected female mice, suggesting a sustained activation of the chemokine-mediated inflammatory signaling in cardiac tissue, potentially reflecting leukocyte retention and prolonged immune cell presence.

Although both IL-6 and CD11d were individually found to be significantly upregulated in the cardiac tissue of infected mice at 28 days post-infection compared to non-infected controls, Pearson correlation analysis showed no significant linear relationship between their expression levels. The correlation coefficient (r = 0.274), coefficient of determination (R2 = 0.075) and p-value (p = 0.655) indicate a small and non-significant association (Figure 5).

To analyse the possible existence of damage to cardiac tissue, measurements of iNOS, PAI-1 and connexin 43 (Cx43) were included. iNOS reflects inflammatory activation and oxidative stress, PAI-1 is associated with fibrinolytic alterations and risk of fibrosis, while Cx43 is essential for electrical communication between cardiomyocytes. Changes in the expression of these markers have been linked to tissue damage in chronic inflammatory processes [24,25,26].

No significant differences in the expression of iNOS (p = 0.4084, Figure 6A) and PAI-1 (p = 0.5345, Figure 6B) or cardiacCx43 (p = 0.2879, Figure 6C) were observed between infected and non-infected mice.

The TLR4/MyD88/NF-κB axis has been widely implicated in myocardial inflammation across various conditions, including viral infections. However, in our post-COVID model, no significant differences were observed in the expression of TLR4, MyD88, or NF-κB p65 in cardiac tissue, 28 days after SARS-CoV-2 infection, suggesting that this pathway may not be involved in the sustained cardiac inflammation observed post-infection. Moreover, the absence of TLR4 pathway activation would suggest that the inflammatory response is not driven by pathogen-associated molecular pattern (PAMP) recognition or acute cytokine storm, but rather by localized chemokine-mediated signaling. This aligns with findings from human studies showing persistent myocardial injury and microthrombi in COVID-19 survivors, even in the absence of active viral replication [30,31].

Recent studies have shown that, although the SARS-CoV-2 spike protein can activate TLR4 in human cardiac cells [32] during the acute phase or in the presence of active viral components, this does not correspond to the post-infectious state of our model. Structural investigations have confirmed that the S1 domain of the viral spike interacts with the TLR4/MD2 complex, promoting inflammatory signaling [33], but this interaction seems to depend on the active presence of the virus or circulating viral fragments, which may explain the lack of activation observed in our animal model at 28 days post-infection. It is important to note that other studies in animal models have shown that activation of the TLR4/MyD88/NF-κB axis can induce significant cardiac inflammation in conditions such as acute myocardial infarction or post-traumatic stress [34,35], but these models differ from the post-infectious context without active viral replication that characterizes the SARS-CoV-2 infection model. These findings reinforce the idea that post-COVID cardiac inflammation may be mediated by mechanisms distinct from classical PAMP recognition, such as those activating TLR4. Instead, it may involve inflammation induced by persistent tissue damage, endothelial dysfunction, or endogenous danger signals (DAMPs), opening new avenues for the study of therapies targeting these mechanisms.

Regarding cytokines, the overexpression of IL-6 and CD11d in our study reinforces the hypothesis of localized and sustained inflammation in a post-COVID phase. In relation to IL-6, the COVID-19 pandemic revealed various long-term cardiovascular complications linked to increased inflammatory responses, particularly through the IL-6 activity [36]. High levels of IL-6 during and after an acute SARS-CoV-2 infection are linked with poor outcomes, such as acute respiratory distress syndrome, myocarditis, endothelial dysfunction, and thrombotic events [31,37,38,39]. It is important to highlight that human studies have assessed IL levels in the plasma but not in specific organs. Our study would confirm an overexpression of IL-6 specifically within the cardiac tissue, supporting the hypothesis that SARS-CoV-2 directly affects the heart, potentially leading to endothelial injury, inflammatory cytokine storm syndrome, hypercoagulability, or thrombosis. Therefore, the observed overexpression of IL-6 and CD11d levels (see below) in the cardiac tissue provides a potential substrate that explains the hypothesis of endothelial dysfunction proposed for explaining post-COVID cardiac problems [40].

In the present research, at 28 days post-infection, CD11d expression was markedly elevated in cardiac tissue, whereas levels of TNF-α, IL-1β, and IL-18 remained within the normal range. This dissociation suggests that CD11d upregulation may occur independently of classical innate immune activation or may reflect a persistent, non-canonical inflammatory mechanism. CD11d, an integrin involved in leukocyte adhesion and retention, is typically upregulated in chronic inflammatory environments and may reflect sustained immune cell presence in cardiac tissue identified [41]. Although direct evidence of CD11d involvement in post-COVID-19 cardiac tissue is currently lacking, previous studies have demonstrated the role of related integrins, such as CD11b, in macrophage retention and vascular remodeling in COVID-affected hearts. Werlein et al. conducted a comprehensive multimodal analysis of cardiac autopsy samples from COVID-19 patients and identified a distinct angiocentric, macrophage-driven inflammatory process [42]. Their findings revealed a significant infiltration of CD11b⁺/TIE2⁺ macrophages in perivascular regions, which were associated with intussusceptive angiogenesis—a non-sprouting form of vascular remodeling-suggesting a mechanism of sustained immune cell presence and tissue-specific inflammation [42]. These macrophages were notably elevated in patients hospitalized more than 10 days after infection onset, and their presence correlated with increased expression of angiogenic and hypoxia-related genes such as VEGFA, FGF2, and HIF1. Given the functional similarities within the β2 integrin family, the observed CD11d overexpression in our model may reflect a comparable mechanism of sustained immune cell presence and tissue-specific inflammation. This highlights the need to further investigate sex-specific and non-specific pathways involved in post-COVID cardiovascular sequelae, which may inform more targeted therapeutic strategies.

Despite the well-established role of IL-1β, IL-18, and TNF-α in acute inflammatory responses, our study did not observe significant changes in their expression in cardiac tissue 28 days post SARS-CoV-2 infection. This finding suggests that these cytokines may not be central mediators of sustained cardiac inflammation in the post-acute COVID-19 phase. IL-1β is a key pro-inflammatory cytokine involved in early immune response and has been implicated in acute myocarditis and systemic inflammation during COVID-19 [43]. However, its absence in our model aligns with recent findings showing that IL-1β levels tend to normalize in convalescent stages, even in patients with persistent symptoms [44]. This may reflect a resolution of the acute inflammatory phase or a shift toward alternative immune pathways. Interleukin-18 (IL-18), another inflammasome-related cytokine, has also been associated with myocardial inflammation in both COVID-19 infection and vaccine-related myocarditis [45]. Interestingly, while IL-18 has been shown to correlate with cardiac edema in convalescent patients [46], our model did not replicate this elevation, possibly due to differences in timing, viral variants, or tissue-specific immune regulation. It is also possible that IL-18 expression is more prominent in systemic circulation or pulmonary tissue than in the heart tissue during the post-acute phase. TNF-α, a central mediator of cytokine storm and acute cardiac injury, has been widely studied in COVID-19 pathophysiology. Elevated TNF-α levels have been linked to acute myocardial damage and activation of the NF-κB pathway [47]. However, in our model, TNF-α expression remained unchanged, suggesting that the TNF-NF-κB axis may not be persistently activated in cardiac tissue after viral clearance. This supports the hypothesis that post-COVID cardiac sequelae may be driven by non-classical inflammatory mechanisms, such as chemokine-mediated leukocyte retention or endothelial dysfunction, rather than sustained TNF-α signaling. Overall, the lack of significant changes in IL-1β, IL-18, and TNF-α expression in cardiac tissue highlights the complexity of post-COVID immune responses and underscores the importance of investigating alternative pathways beyond classical cytokine storm mediators.

Our findings show that, despite the upregulation of both IL-6 and CD11d in the cardiac tissue of infected mice, there was no significant linear correlation between the markers. This result is consistent with clinical evidence in post-COVID patients, where elevated IL-6 has been associated with myocardial damage and myocarditis, but without a clear relationship with other specific immune markers such as CD11d [46,48,49]. Furthermore, the lack of correlation observed in our model suggests that IL-6 and CD11d may be regulated by independent pathways, which is consistent with recent reviews highlighting the complexity of post-COVID immune mechanisms, where multiple cells and mediators participate in a non-linear manner in cardiac injury [37,50]. These findings reinforce the need for further studies exploring the interaction between cytokines and adhesion molecules in post-COVID cardiovascular pathophysiology, as well as their value as potential prognostic biomarkers.

Myocardial overexpression of PAI-1 is recognised as a profibrotic marker that promotes ventricular remodelling and fibrosis, contributing to cardiac dysfunction [51,52,53]. Similarly, increased iNOS is associated with oxidative stress and contractile impairment, exacerbating myocardial damage [54] and alterations in the expression or distribution of connexin 43 compromise electrical conduction, increase susceptibility to arrhythmias, and accelerate progression to heart failure [55]. In our study, the expression of PAI-1 and iNOS in cardiac tissue from female mice infected with SARS-CoV-2 showed modest increases (≈13%) compared to noninfected mice. This finding suggests that, in this PASC condition, there is no marked activation of inflammatory (iNOS) or fibrinolytic (PAI-1) pathways, which could indicate an absence of severe myocardial damage. These results are consistent with the hypothesis that the cardiac inflammatory response induced by SARS-CoV-2 is more pronounced in the acute phase and tends to normalize in later stages [56,57], reinforcing the need for studies evaluating the functional and structural evolution of the myocardium.

Thus, some limitations of this study could be pointed out. The relatively low number of animals may also reduce statistical power and generalizability of the findings. A second limitation of this study is that cytokine analysis was performed using Western blot rather than ELISA, which offers higher sensitivity and absolute quantification. However, the sample preparation protocol (homogenization in lysis buffer) precluded ELISA analysis. Future research should incorporate ELISA or multiplex immunoassays to complement protein expression data. Another limitation could be the absence of histological and immunofluorescence analyses of cardiac tissue, which would have provided valuable information on structural and cellular alterations following PASC.

Finally, it should be noted that sex has a greater influence on the initiation and progression of the COVID-19 disease. In fact, the differences in severity and mortality could be explained by the hormonal and genetic regulation [58]. Our study was conducted in females of the K18-hACE2 transgenic mouse model. This model is widely used to study SARS-CoV-2 infection, as it expresses the human ACE2 receptor under the control of the cytokeratin-18 promoter, allowing for efficient viral entry and replication [59]. However, this overexpression could not replicate the physiological regulation of ACE2, including modulation by sex hormones such as estradiol, which may partially mask their influence on susceptibility and disease progression. Future studies should incorporate sex comparisons and hormonal modulation to better understand these differences.