Although different types of immune cells were increased in thoracic specimens of TAAD, macrophages are one of the most abundant types [27, 34, 38, 39]. Macrophages found in the thoracic aortic wall are mainly derived from two main sources. Most macrophages in the aneurysmal thoracic aortic wall derive from circulating monocytes, which are produced from bone barrow and mobilized from peripheral reservoirs such as the spleen [40, 41, 42]. Transplantation of bone marrow cells expressing green fluorescent protein revealed that a proportion of macrophages in the aortic adventitia originated from bone marrow-derived monocytes [43]. Besides circulating monocytes, aortic macrophages might also develop from embryonic precursors and early postnatal circulating monocytes. This group of macrophages are known as tissue-resident macrophages, which are independent from bone-marrow progenitors and are self-maintained and self-developed during adulthood [41, 42]. Updated investigations have added evidence that tissue-resident macrophages might also originate from multipotent stem cells, which was identified in the media and adventitia of TAAD thoracic aorta expressing macrophage marker CD68 [44, 45].

Macrophages have been classified into several different phenotypes over the past two decades, mainly including M1 and M2 macrophages [46]. M1 macrophages are also known as pro-inflammatory macrophages with production of proteolytic enzymes and pro-inflammatory cytokines. In comparison, M2 macrophages have an anti-inflammatory role through promoting ECM remodeling and tissue repair.

Single-cell RNA sequencing described a systemic landscape of immune cells in TAAD. Macrophages were classified into ‘M1 like’, ‘M2 like’, ‘IFN-response’, ‘remodeling’, and ‘proliferating’ subgroups [27, 47, 48] (Fig. 3). ‘M1 like’ macrophages were identified by IL-1B, TNF and NFKB1, while ‘M2 like’ macrophages were characterized by MERTK, MRC1, STAB1 and CD163. ‘M1 like’ macrophages are further divided into three subgroups including ‘M1 like 1’, ‘M1 like 2’ and ‘M1 like 3’. ‘M1 like 1’ macrophages expressed several cytokine genes, such as CCL3L1, CCL4L2, CCL4 and TNF, associated with the inflammatory response. ‘M1 like 2’ macrophages expressed EREG, AREG, TIMP1 and VCAN which are involved in tissue remodeling and the inflammatory response. ‘M1 like 3’ macrophages expressed ETS1, RUNX3 and genes encoding major histocompatibility complex (MHC) class I molecules. This result indicated that ‘M1 like 3’ macrophages could present antigens to CD8${}^{+}$ T lymphocytes. Similarly, ‘M2 like’ macrophages were classified into two subgroups including ‘M2 like 1’ and ‘M2 like 2’. ‘M2 like 1’ macrophages expressed PDK4, STAB1, TXNIP and MAF, involved in glucose metabolism, anti-inflammation, and phagocytosis. ‘M2 like 2’ macrophages expressed C1QA, C1QB, C1QC and RAB13. ‘IFN-response’ macrophages expressed interferon-induced genes, including IFI44L, ISG15, IFIT1 and IFITM3. ‘Remodeling’ macrophages shared characteristics with SMCs and fibroblasts including IGFBP7, ADIRF, DSTN, TPM2, MGP, and MYL9. In addition, protease genes including ADAMTS1, MMP-2 and CTSF were highly expressed, associated with tissue remodeling. Finally, ‘proliferating’ macrophages are closely associated with proliferation with high expression of microtubule-related genes TUBB, TUBA1B and STMN, histone-related genes H2AFZ, HMGB2 and HMGN2 and cyclin-dependent kinases regulatory subunit 1 CKS1B.

Fig. 3.

Origin and classification of macrophages in TAAD thoracic aortic wall. Macrophages within thoracic aortic walls originate mainly from two sources. One main origin of macrophages is circulating monocytes. Another source is embryonic precursors and early postnatal circulating monocytes. Macrophages from human aneurysmal thoracic aortas are classified into eight subclusters including M1 like 1, M1 like 2, M1 like 3, M2 like 1, M2 like 2, IFN-response, remodeling, and proliferating macrophage. Macrophages of mice aneurysmal thoracic aorta induced by BAPN are divided into three subgroups including resident-like, antigen-presenting and pro-inflammatory macrophage. TAAD, thoracic aortic aneurysm and dissection; IFN, interferon; TNF, tumor necrosis factor; BAPN, $\beta$-aminopropionitrile; MHC, major histocompatibility complex; TAAD, thoracic aortic aneurysm and dissection.

Macrophage subgroups are also identified in the thoracic aorta of BAPN administrated TAAD mice. A total of three macrophage subgroups were identified including ‘resident-like’, ‘antigen-presenting’ and ‘pro-inflammatory’ subsets [34]. In detail, ‘resident-like’ macrophages highly expressed Lyve1, Cx3cr1, F13a1, Lyve1, and Gas6. This group of macrophages were thought to be located residentially in the adventitia and may develop and self-renew independently of CCR2-mediated monocyte recruitment [49, 50]. ‘Antigen-presenting’ macrophages highly expressed Cd74, MHC II genes. This may indicate that this macrophage subcluster is involved in presenting antigens. ‘Pro-inflammatory’ macrophages highly expressed pro-inflammatory chemokines and M1-polarized macrophage markers. This subcluster is similar to M1 macrophages and may infiltrate and drive inflammatory responses. Subclusters, classifications and definitions of these new macrophage subgroups might provide novel insights into understanding the pathogenesis of TAAD.

Since macrophages play such a pivotal role in the pathogenesis of TAAD, understanding the regulatory mechanism for macrophages might provide novel insights into macrophage-based therapy for TAAD [68]. Over the decades, a series of regulators of macrophages have been identified as being closely associated with the pathogenesis of TAAD. These regulators are mainly classified into three categories. The first type of regulators affects macrophage accumulation and infiltration, which initiate further biological process. The remaining two types of regulators influence macrophage functions through modulating pro-inflammatory phenotype and MMPs expression (Table 1, Ref. [11, 29, 32, 34, 35, 63, 65, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93]).

Previous investigations have uncovered some mechanisms that are involved in macrophage accumulation, infiltration, driving the inflammatory response. Macrophage infiltration was more significant in TAAD patients with atherosclerosis [94]. Elevated plasma levels of LDL cholesterol promoted Ang II-induced TAAD through enhancing macrophage infiltration [95]. Nicotine free base promoted TAAD progression in SMC specific TGFBR2 knockout mice, whose effect was sensitized by BAPN co-stimulation [96]. These results might partly explain why atherosclerosis, hypercholesterolemia and tobacco use might serve as established risk factors for TAAD [97]. However, these studies only provided observations on this phenomenon and did not provide in-depth mechanisms of how macrophages differentiate and infiltrate into thoracic aortic walls.

As mentioned above, circulating monocytes are the main sources of thoracic aortic macrophages. They might recruit and infiltrate into aortic walls through chemokine/chemokine-receptor pathways and selectins [63, 98, 99, 100, 101, 102]. The CCL2/CCR2 axis is confirmed as a critical manner of the chemokine/chemokine-receptor pathway that participate in macrophage accumulation and infiltration into aortic walls. Global and myeloid specific deficiency of CCR2 decreased macrophage recruitment, inhibited inflammatory cytokines, and reduced Ang II-induced aortic dissection in mice [63]. Similarly, global and bone-marrow-derived cell-specific CCL2 deficiency protected mice from elastase-induced aneurysms [69, 70]. Besides CCL2 and CCR2, genetic knockout of IL-1$\beta$, IL-1R, IL-6, SMAD3, ADAMTS-4 or CCN4 attenuated thoracic aortic dilation through inhibition of macrophage recruitment [29, 35, 71, 72, 73]. Updated knowledge has identified that epigenetic mechanisms may also participate in the regulation of macrophage infiltration [103]. Upregulation of miR-146a and miR-21, as well as the downregulation of miR-29b, miR-29c and miR-27b are closely associated with aortic inflammation and macrophage infiltration [74]. These miRNAs might be under the control of mesenchymal stem cells as immunomodulators of aortic inflammation through regulation of proinflammatory cytokines [45, 104, 105, 106].

Exogenous administration of specific drugs could also reduce the pathologic changes in different mouse models of TAAD through reducing macrophage infiltration and inflammation. Pretreatment of IL-1R antagonist anakinra reduced macrophage infiltration and attenuated TAAD formation induced by elastase [29]. Similarly, calcium channel blocker azelnidipine reduced BAPN-induced TAAD through anti-inflammatory effects [107]. Several investigations focus on in-depth mechanisms of how macrophage infiltration is reduced. The glycolytic enzyme pyruvate kinase M2 activator TEPP-46 markedly attenuated the progression of TAAD induced by BAPN through inhibition of macrophage infiltration associated with the NOD-like receptor family and NLRP3 inflammasome [55].

When macrophages accumulated and infiltrate, the pro-inflammatory phenotype and associated inflammatory response are essential to participate in the pathological process of TAAD. Hypertension is the most important risk factor for TAAD [108, 109]. A previous study has shown that hypertensive TAAD patients tended to have more pro-inflammatory macrophages in the adventitia and media of thoracic aorta [110]. This result may indicate that hypertension might promote a pro-inflammatory phenotype of macrophages, with underlying mechanisms needing further investigation.

A pro-inflammatory phenotype of macrophages in TAAD is mainly regulated through affecting production of inflammatory factors and the capability of migration and invasion of macrophages [111]. Several pro-inflammatory regulators have been identified over the past two decades. Single-nuclear RNA sequencing and genome-wide association studies have identified LRP1 as a key potential regulator of inflammation in macrophage of TAAD [75]. In experimental investigations, overexpression of core circadian clock gene BMAL1 induced a pro-inflammatory response in cultured macrophages [76]. TGF-$\beta$ induced severe inflammatory response through enhancing macrophage invasion [77]. Silencing of ANGPTL8 and AT1R decreased inflammatory factors in macrophages including IL-1$\beta$, IL-6, MCP-1 and TNF-$\alpha$ [11, 78]. In vivo, myeloid-specific knockout of TGF-$\beta$ or NEU1 reduced macrophage pro-inflammatory functions and ameliorated TAAD induced by BAPN [79, 80]. ADAMTS1-deficient macrophages exhibited low activity of the inflammatory response through abrogated migration capacity in BAPN mice [81].

Meanwhile, some molecules demonstrated anti-inflammatory effects on macrophages in TAAD. Functional silencing or knockout of SR-A1 or SOCS3 may aggregate TAAD in murine models. In detail, SR-A1 deficiency aggravated BAPN induced TAAD in mice through TYRO3 mediated efferocytosis and inflammatory cascades in macrophages [82]. Macrophage-specific deletion of SOCS3 exaggerated TAAD through M1-dominant differentiation of macrophages via acute enhancement of STAT3 activation [83, 84]. Conversely, restoration of RGS1 reduces Ang II-induced TAAD through inhibiting macrophage chemotaxis and desensitizes chemokine receptor signaling [85].

Notably, metabolic reprogramming is required for the proper polarization and function of activated macrophages [112]. Similar to the Warburg effect of tumor cells, pro-inflammatory M1 pro-inflammatory macrophages increase glucose consumption, lactate release and decreased oxygen consumption rate. However, M2 anti-inflammatory macrophages are characterized by the employment of oxidative glucose metabolism pathways [113]. Besides, fatty acid, vitamin and iron metabolisms are also closely associated with macrophage polarization [114]. Using untargeted metabolomics of clinical TAAD specimens and BAPN-induced mice model, C18-ceramide was identified to be increased through the de novo synthesis pathway and promoted macrophage pro-inflammatory phenotype through the NLRP3-caspase 1 pathway [86]. Similarly, succinate, tryptophan, kynurenine, quinolinic acid and kynurenine-to-tryptophan ratio were also identified to be increased in TAAD [87, 88]. Knockdown of the key synthetic enzyme of succinate OGDH or kynurenine pathway enzyme kynureninase could reduce the expression of inflammatory factors in macrophages.

Regulation of the pro-inflammatory phenotype in macrophages through pharmacological intervention has been proven effective in attenuating TAAD in vivo. Targeting pro-inflammatory Il1rn${}^{\mathrm{+}}$/Trem1${}^{\mathrm{+}}$ macrophage subpopulations through mLR12 could significantly reduce thoracic aortic rupture rate in BAPN-administrated mice [34]. Dexamethasone treatment suppressed NF-$\kappa$B signaling pathway in macrophages and further reduced the inflammatory response, immune cell infiltration and incidence of TAAD in BAPN mice [115]. Selective mineralocorticoid receptor antagonist eplerenone protected mice from BAPN-induced TAAD through decreasing TNF$\alpha$ and IL-6 in macrophages [116]. An in-depth understanding of pro-inflammatory macrophages might broaden our horizon on anti-inflammatory therapy for TAAD.

ECM degradation is another critical mechanism involved in TAAD development, which has been reported to be under the control of specific molecules. For example, macrophage-derived legumain is essential for ECM degradation. Macrophage-specific deletion of legumain (LGMN) alleviated BAPN-induced thoracic aortic dilation, aneurysm, and dissection in mice [89]. Deficiency of urokinase-generated plasmin protected against media destruction and aneurysm formation by reducing the activation of pro-MMPs [90]. Inhibition of endogenous ceramide synthesis in macrophages by myriocin, attenuated BAPN-induced TAAD in mice through reducing macrophage inflammation and expression of MMPs [86].

Specific MMPs in macrophages could be directly regulated, including MMP-2, MMP-9, and MMP-12. macrophage-derived HIF-1$\alpha$ activation triggered ECM degradation and elastic plate breakage through increasing ADAM17 to induce MMP-2 and MMP-9 expression [91]. In vitro, MCP-1 and IL-6 enriched conditioned medium induced differentiation of monocytes into macrophages and expression of MMP-9 [63]. Cytosolic DNA from damaged aortic SMCs induced MMP-9 expression in macrophages through the STING pathway and its target IRF3 [65]. STING deficiency protected against thoracic aortic aneurysm and dissection through reducing MMP-9 in macrophages. Selective NLRP3 inhibitor MCC950 prevented TAAD through reducing MMP-9 expression and activation in macrophage via the NLRP3-caspase-1 inflammasome [92]. Another study confirmed an IL-3/IL-3R$\beta$/MMP-12 axis in macrophage during the progression of TAAD. IL-3 deficiency in macrophages diminished JNK and ERK1/2 dependent AP-1 pathways, thus decreasing expression of MMP-12 [32]. Epigenetic mechanisms also participate in the regulation of MMPs in macrophages, such as miR-320 [93].