SARS-CoV-2 infection and the subsequent COVID-19 disease manifestation involve extensive organ damage, including the lungs, heart, kidney, liver, and brain. The multi-organ complications of COVID-19 result from compromised vascular integrity, which leads to fluid retention, excessive inflammatory cell infiltration, endotheliitis, and the activation of coagulation pathways. Viral damage and the multi-organ hyperinflammatory state, often referred to as the “cytokine storm,” are thought to cause tissue damage through repetitive insult and induce a continuously remodeling fibrotic state. Fibrosis has emerged as a serious and long-lasting comorbidity of COVID-19. However, the lingering effects of the disease, often referred to as post-COVID-19 syndrome, remain elusive. An effective therapeutic has yet to sufficiently address both the underlying causes of fibrosis and the observed aberrant viral-induced tissue damage. This review will explore the intersection of the inflammatory and fibrotic state in COVID-19 and potential therapeutics to address these convergent pathologies.
A major co-morbidity of COVID-19 is acute respiratory distress syndrome (ARDS). An estimated 50% of hospitalized COVID-19 patients with COVID-19-induced pneumonia develop ARDS, which has a 52.4% mortality rate . ARDS is associated with diffuse alveolar damage, fibrin-rich hyaline membranes, increased epithelial and endothelial cell permeability, fluid leakage into the pulmonary interstitium, gross disruption of gas exchange, hypoxia and respiratory failure, and subsequent pulmonary fibrosis . Evidence of pulmonary fibrosis is evident in moderate to severe COVID-19 patients through lung volume loss, architectural distortion, and fibrotic bands in the late or remission stages of COVID-19  and in autopsies . Pulmonary fibrosis is also found in SARS patients 4 and 5 weeks after recovery and in autopsies [5,6,7,8]. A 3 year follow-up revealed that ten out of forty-six (21.74%) patients had restrictive ventilation dysfunction, but fortunately, most recovered by 15 years .
Endothelial cells have emerged as a central player in the pathogenesis of ARDS and multi-organ failure in patients with COVID-19. The SARS-CoV-2 virus damages the vasculature both directly and indirectly. Pathological analysis of the pulmonary vasculature of COVID-19 patients shows high levels of microthrombi and vessel growth through intussusceptive angiogenesis and, to a lesser extent, sprouting angiogenesis, which is positively correlated with hospitalization duration . COVID-19 lung autopsies show diffuse alveolar damage with widespread micro- and macro-thrombi consistent with that of ARDS . Inflammatory-related genes associated with intussusceptive angiogenesis have been reported to be upregulated in both ARDS and COVID-19 patients . By direct infection, SARS-CoV-2 can cause endotheliitis and subsequent endothelial cell death . SARS-CoV-2 can indirectly cause an increase in vasculature permeability through activation of the kallikrein–bradykinin pathway by SARS-CoV-2/ACE2 binding and subsequent ACE2 inhibition .
SARS-CoV-2 infection also results in pulmonary endothelial cell recruitment of neutrophils, producing reactive oxygen species (ROS) . ROS increases inter-endothelial gaps by increasing endothelial cell contractility while simultaneously loosening endothelial cell junctions. Fluid retention then arises from IL-1β and TNF activation of glucuronidases which degrades glycocalyx and upregulates hyaluronic acid synthase 2 . Endothelial cells simulated by IL-1β and TNF increase P-selectin, von Willebrand factor, and fibrinogen expression which bind and aggregate platelets . The protein C/protein S pathway, a disintegrin and metalloproteinase with a thrombospondin type 1 motif member 13 (ADAMTS13), tissue-type plasminogen activator (t-PA), and other coagulation factors are activated and contribute to the dissolution of the fibrin-mesh. This triggers fibrinolysis through the conversion of plasminogen to plasmin . The accumulation of fibrin breakdown products, such as fibrin-derived D-dimers, leads to the ischemic complications found in COVID-19 patients with approximately 76% of hospitalized adults presenting with elevated D-dimer levels . Overall, widespread clotting has proven to have critical consequences as 70% of fatal cases present with disseminated intravascular coagulation (DIC) in comparison to only 1% of the surviving population .
Viral infection of type II epithelial cells induces alveolar macrophage activation , resulting in the secretion of pro-inflammatory cytokines. These pro-inflammatory cytokines recruit highly toxic neutrophils and activated platelets into the alveolar space, which contribute to ARDS, the cytokine storm, and systemic sepsis. Alveolar macrophages phagocytose alveolar debris and produce cytokines, chemokines, and growth factors involved in the repair of lung tissue and likely play a role in SARS-coronavirus-induced pulmonary fibrosis ; however, more work is needed to distinguish the effects of pro-inflammatory insult from that of growth factors alone.
SARS-CoV and SARS-CoV-2 Infection and Lung Fibrosis
Increased TGF-β-activity is suspected to arise, in part, from a positive-feedback loop-induced influx of TGF-β-producing immune cells and the upregulation of highly coagulative and fibrinolytic pathways apt for TGF-β activation . Unfortunately, these conditions are ripe for producing rapid and massive edema and fibrosis that seeks to remodel but ultimately blocks pulmonary airways in COVD-19. TGF-β has been described as the cornerstone of tissue inflammation and increased lung collagen deposition in viral-induced pneumonia of aged individuals and is thought to be responsible for chronic lung pathology and fibrotic outcomes of viral pneumonia . Pulmonary fibrosis is seen with SARS infection as early as 2 to 3 weeks  and persists 9 months after hospitalization in 21% of patients . In relation to COVID-19, TGF-β signaling pathways were identified as an integrative pathway in the disease by miRNA analyses [23, 24]. Hence, a comprehensive treatment plan for COVID-19-induced tissue damage necessitates addressing the underlying factors of viral-induced fibrosis and biological mediators, including TGF-β.
The protective role of the ACE2 receptor in lung fibrosis is impaired by SARS-CoV-2 downregulation of ACE2. The protective role of ACE2 is evident in bleomycin models of fibrosis in which recombinant human ACE2 diminishes lung collagen amassment, while ACE2-specific small interfering RNA (siRNA) treatment increases lung collagen accrual in mouse bleomycin models [25,26,27]. Additionally, lung biopsies of idiopathic pulmonary fibrosis (IPF) display a 92% reduction of ACE2 mRNA and a 74% ACE2 enzymatic activity . The protective role of ACE2 occurs through negative regulation of ANG-II. This is significant, as ANG-II is suspected to engage in an autocrine feedback loop to stimulate TGF-β production  while a decrease in ANG-II activity is accompanied by a decrease in TGF-β levels [30, 31]. This bolsters the case for SARS-CoV-2 direct growth factor-induced fibrosis, rather than the simply indirect tissue damage from the hyperinflammatory insult of COVID-19.
Moreover, Pang et al. identified elevated serum levels of TGF-β1 during the early phase of SARS infection . SARS-CoV infection introduces high levels of TGF-β as observed in alveolar epithelial cells, bronchial epithelial cells, monocytes, and macrophages [33, 34] (Fig. 1). The SARS-CoV N protein can associate with the TGF-β-associated SMAD3 transcription factor to interfere with SMAD3/4 complex and can enhance the interaction between SMAD3 with downstream target genes . Additionally, overexpressing the N protein in lung epithelial cells and fibroblasts enhances the TGF-β-induced expression of PAI-1 and collagen I independently of SMAD4 . Interestingly, the SARS-CoV N protein shares little homology with analogous structures of other known coronaviruses  but shares 90.52% homology with SARS-CoV-2 . More work is needed to determine if the SARS-CoV-2 N protein similarly enhances the pro-fibrotic activity of lung fibroblasts independently of TGF-β but presents a promising avenue of investigation for direct intervention in SARS-CoV-2-induced fibrosis.
CXC Chemokines in Betacoronavirus Infections
C-X-C chemokines are named for the four highly conserved cysteine residues on the NH2 terminus, with the first two cysteine residues being separated by a variable, “X,” amino acid residue [38, 39]. The function of CXC chemokines in angiogenesis can be largely accredited to the presence or absence of an “ELR” (Glu-Leu-Arg) motif adjacent to the first cysteine residue . ELR+ CXC chemokine ligands, such as CXCL1, CXCL2, CXCL3, CXCL5, CXCL6, CXCL7, and CXCL8, promote angiogenesis. All ELR+ chemokines signal only through CXCR2, except for CXCL6 and CXCL8, which signal through both CXCR1 and CXCR2 .
The ELR− chemokine family consists mainly of IFN-γ-inducible chemokines: CXCL9, CXCL10, and CXCL11. ELR− chemokines activate the CXCR3 receptor through autocrine and paracrine signaling . These CXC chemokines share approximately 40% homogeny in sequence but possess different binding affinities with CXCR3  and are expressed in distinct temporal and spatial patterns. Generally, CXCL9 has the lowest affinity for CXCR3, while CXCL11 has the highest . CXCL10 can out-compete CXCL4, but not CCL2, CCL5, CXCL8, or MIP-1α for endothelial cell surface binding .
Chemokines and cytokines influence viral disease progression by modulating viral infection, orchestrating immune cell chemoattraction , and polarizing the growth and reduction of the microvasculature and the repair of viral-induced tissue injury. Precedence of this has been set by analysis of the previous coronavirus infections. Experts have examined the severe acute respiratory syndrome (SARS) associated coronavirus (SARS-CoV) for insight on SARS-CoV-2. Hunag et al. concluded that IFN-γ, IL-18, TGF-β, IL-6, CXCL19, CCL2, CXCL9, and IL-8, but not TNF-α, IL-2, IL-4, IL-10, IL-13, or TNFRI, are highly elevated in the acute phase of SARS infection . Another group found CXCL9, CXL10, and IL-8 to have been highly correlated with adverse outcomes in SARS patients . CXCL10 and CCL2 are upregulated in both severe and non-severe SARS patients compared to healthy controls. The upregulation of CXCL10 during the late phase of illness can significantly distinguish severe from non-severe SARS patients in late stages . Additionally, markedly high CXCL10 levels are sustained in lung and lymphoid tissue for the duration of SARS illness .
Chemokine signaling pathways (hsa04062) were found to be among the most significantly involved pathways in SARS infection by KEGG analysis . Protein-protein interaction (PPI) networks of SARS infected cells revealed CXCL10 to be one of the top eight genes with the most interactions of the 180 genes determined to be differentially expressed in PBMCs between SARS patients and healthy controls by microarray. CXCL10 has been identified as a promising therapeutic target for SARS  and was found to be among the top 10 signal transducers most closely linked to heart failure in coronavirus infections through differential gene screening analysis .
On the other hand, middle east respiratory syndrome (MERS) infections seem to have very different methods of transcription regulation than that of the related coronavirus, SARS-CoV-2. Microarray analysis and comparison of genes differentially regulated by MERS-CoV and SARS-CoV-2 (127 and 50, respectively), found only eight genes, including CXCL5, CXCL6, and CXCL8, to be modulated by the two coronaviruses in the same manner, either up- or downregulated . Hence, alongside insight from previous coronavirus outbreaks, the roles of chemokines have proven critical in understanding COVID-19 pathogenesis but warrant further investigation.
Chemokines in the Diagnosis and Prognosis of COVID-19
Evidence suggests that chemokines may have diagnostic and prognostic significance in COVID-19. CXCL5, CXCL9, and CXCL10 are highly upregulated in the plasma of COVID-19 patients and have been suggested as potential biomarkers for disease and development of the “cytokine storm” . Investigators have found a significant increase in the expression of genes encoding CXCL1, CXCL3, CXCL6, CXCL18, and CXCL17 in patients with moderate COVID-19 compared to healthy controls. Another group found CXCL8, CXCL10, and CXCL13 to be upregulated in COVID-19 patients and were the only three CXC chemokines differentially regulated against healthy controls . On the other hand, severe COVID-19 patients displayed higher plasma levels of CXLC9, CXCL10, and CXCL11 than moderate COVID-19 patients . Another study found bronchoalveolar cells of severely ill COVID-19 patients to express increased hypoxia-induced factor 1-alpha (HIF1α) and the downstream targets CXCL8, CXCR1, CXCR2, and CXCR4 by scRNASeq .
There is also data to suggest that CXC chemokine levels may be variable in the disease progression of young versus old patients. Interestingly, high CXCL9 plasma levels were found in COVID-19 infected adults, but not children . However, another study found that CXCL9, CXCL10, CXCL11, and CXCR3 displayed reduced expression in older adults by shotgun RNA-sequencing profiles of nasopharyngeal swabs . Angioni et al. determined CXCL8, IL-10, IL-15, IL-27, and TNF-α immune profiles to correlate with older age, longer hospitalization, and disease severity . However, another study found that high IL-6, CXCL-8, and TNF-α levels were associated with adverse outcomes of COVID-19 independently of demographics and comorbidities  and mirrored those of critically ill ARDS and sepsis patients . It is important to note that disease time course plays a role in the fluctuation of chemokine and cytokine levels along disease progression. For instance, IFN-λ3, IL-6, CXCL10, and CXCL9 levels are observed to rapidly increase before decreasing after the onset of severe pneumonia . Additionally, although VEGF and CXCL10 levels appear to decrease throughout hospitalization gradually, levels remain significantly high in the severe and critical patients for the duration of hospitalization . More work is needed to establish demographic differences in cytokines profiles over the course of disease to harness the full power of chemokines as distinguishable biomarkers for actionable targeting.
Chemokines in COVID-19 Disease Pathogenesis
Aydemir et al. utilized computational modeling to predict COVID-19 miRNA targets and constructed an integrative pathway network analysis of the 40 identified SARS-CoV-2 miRNAs and their genetic targets . Further, KEGG analysis revealed angiogenesis and inflammation mediated by chemokine and cytokine signaling to be among the top eight differentially regulated pathways of SARS-CoV-2 infection. Additionally, investigators found CXCL1, CXCL9, CXCL10, CXCL11, and CXCL16 to be main targets of SARS-CoV-2 miRNA regulation of gene targets associated with cytotoxins and gene regulation . Additionally, RNASeq analysis and KEGG enrichment of differentially expressed genes in the PBMC COVID-19 patients vs. healthy controls found “cytokine-cytokine receptor interaction” and “viral protein interaction with cytokine and cytokine receptor” related pathways to be two of the most differentially regulated . Interestingly, this study found the expression of CXCL10 and CXCL8 to be significantly downregulated and upregulated in COVID-19 PMBCs, respectively.
CXC chemokines play a large role in the dynamic microenvironment and cellular composition of COVID-19 infections. Specifically, SARS-CoV-2-infected lungs overexpress neutrophil chemoattractants CXCL1, CXCL2, CXCL3, CXCL5, CXCL8, and CCL20 . Neutrophils, which are involved in early-stage anti-viral defense, are known to cause cytotoxic lung inflammation after lysis and high neutrophil-to-lymphocyte ratios are associated with poor outcomes of patients with COVID-19 [65, 66]. COVID patients additionally displayed higher activation status of resident and non-resident macrophages. Non-resident macrophages showed an upregulation of CCL2, CCL3, CCL20, CXCL1, CXCL3, IL1β, IL8, IL18, and TNF while resident macrophages were characterized by CCL2, CCL3, and CXCL10 expression . CD69high and CXCR3low mucosa-associated invariant T (MAIT) cell counts are associated with poor outcomes by unsupervised analyses . Better outcomes were identified in patients with significantly higher frequency macrophages of the CD3+ CD4+ CD45RO+ CXCR3+ subsets, higher counts of CD14+ CD11C+ HLA-DR+ subset dendritic cells, and a lower neutrophil count . One study found CXCL10 to be the core gene in a PPI network mapping the 29 immune-related differentially regulated genes in COVID-19 patients and could be correlated with the CD4+, CD8+, monocyte, and DC-related immune signatures [70, 71].
The cytokine storm is the leading cause of death in patients infected with COVID-19. The cytokine storm involves hyper-activated T lymphocytes and the release of pro-inflammatory cytokines that enhance vascular permeability and plasma leakage, resulting in injury to pulmonary tissue, ARDS, and multi-organ failure. These pro-inflammatory mediators include IFN-γ, IL-1RA, IL-6, IL-10, IL-19, CCL2, CCL7, CXCL2, CXCL9, CXCL10, CXCL5, ENRAGE, and poly (ADP-ribose) polymerase 1 . Increased levels of IL-1β, IFN-γ, CXCL10, and CCL2 suggest that T-helper-1 (Th1) cell function is consistent with that of MERS and SARS infections.