Active lymphatic pumping is a major component of lymph transport. Active pumping is achieved by the spontaneous contractions of collecting lymphatics, which propel lymph centrally and act in concert with 1-way lymphatic valves, spaced at regular intervals, to prevent backflow (Scallan et al., 2016). Efficient lymph transport also depends on a low level of collecting lymphatic permeability, which normally is comparable to that of venules (Scallan & Huxley, 2010).

Rodent models of cardiovascular disease, including atherosclerosis, metabolic syndrome, obesity and advanced aging, have uncovered various types of lymphatic vessel dysfunction that occur in association with these diseases, as summarized in recent reviews (de Godoy, 2019; Jiang et al., 2019; Azhar et al., 2020; Ho & Srinivasan, 2020; Kataru et al., 2020; Norden & Kume, 2020). Lymphatic dysfunction may include a reduction in the frequency of spontaneous contractions, impaired contractile strength of lymphatic muscle cells (LMCs), increased permeability of the collecting vessel wall to water and solutes, and increased backflow through lymphatic valves, which normally are highly competent. Not all of the diseases listed above are characterized by every facet of lymphatic dysfunction, but many exhibit multiple components. Lymphatic dysfunction might not only be a consequence of these diseases, it could also contribute to their progression and/or manifestation. Precedent for this idea is provided by the observations by Rossitto et al. (Rossitto et al., 2020b) that even subtle amounts of edema can interfere with normal organ function (Rossitto et al., 2020a) and that subclinical edema with an underlying lymphatic component is present in patients and animal models of heart failure with preserved ejection fraction. Systemic subclinical lymphedema has also been noted in patients with a basal metabolic index (BMI) > 50 kg/m² (de Godoy, 2019).

Previously, one of our laboratories provided in vivo evidence for lymphatic abnormalities in apolipoprotein E knockout (ApoE ^−/− ) mice fed a high fat diet (HFD) (Lim et al., 2009). This is a well-established mouse model of cardiovascular disease (Ilyas et al., 2022), in which a diet rich in fat and cholesterol on an ApoE deficient background accelerates the development of hypercholesteremia, atherosclerotic plaque development (Zhao et al., 2020) and inflammation of the skin and other tissues (Feingold et al., 1995; van Ree et al., 1995). Lim et al. (Lim et al., 2009) showed that ApoE^−/− mice fed a diet rich in fat and cholesterol exhibited evidence of swelling in the tail and foot pads, pointing to possible disruption in lymphatic drainage. Histological examination revealed fluid pockets and increased lipid deposition in the dermis as well as increased infiltration of CD45⁺ leukocytes. Collecting lymphatic vessels in the ear skin of these mice exhibited evidence of hyperpermeability, based on the extravasation and/or reflux of Evans Blue dye after uptake from the interstitium. Two subsequent studies showed that the transport of macromolecules, such as fluorescein isothiocyanate dextran, in the skin (Lim et al., 2013) and aortic lymphatic vessels (Yeo et al., 2020) was significantly reduced in ApoE ^−/− mice compared to ApoE ^+/+ mice. Collecting vessels in skin also failed to downregulate Lyve1 upon maturation and showed reduced smooth muscle cell coverage, compared to ApoE ^+/+ controls (Lim et al., 2009). Collectively, these findings were suggestive of altered collecting vessel function.

In the present study, we directly investigated various aspects of lymphatic collecting vessel function in ApoE-deficient mice on a HFD. Vessels were excised from ApoE ^+/+ or ApoE ^−/− mice fed a HFD and studied under defined conditions ex vivo, in which we could quantify possible changes in lymphatic contractile strength, spontaneous contraction frequency, calculated and measured lymph pump output, secondary valve function, and collecting vessel permeability to albumin. Our results show that all of these aspects of lymphatic vessel function are compromised in ApoE ^−/− mice, compared to ApoE ^+/+ control mice, fed a HFD. These findings extend those of other groups (Blum et al., 2014; Garcia Nores et al., 2016; Nitti et al., 2016; Castorena-Gonzalez, 2022) suggesting that significant dysfunction of lymphatic collecting vessels develops in association with a high fat diet, and extend our previous in vivo observations (Lim et al., 2009; Lim et al., 2013; Yeo et al., 2020), showing that lymphatic dysfunction is further exacerbated by ApoE deficiency. The combination of ApoE deficiency and HFD negatively impacts both lymphatic smooth muscle (LMC) and lymphatic endothelial cell (LEC) function, which may result in impaired removal of cholesterol by the lymphatic system from the skin and the arterial wall (Lim et al., 2013; Martel et al., 2013; Martel & Randolph, 2013) and thereby exacerbate the progression and/or severity of atherosclerosis (Yeo KP et al., 2020).

Popliteal collectors from ApoE ^−/− mice typically exhibited blunted contraction amplitudes compared to ApoE ^+/+ vessels (with both strains fed a HFD). Representative examples of spontaneous contractions as a function of pressure are shown in Figures 1A,B for the two strains. Smaller contraction amplitudes are apparent for the ApoE ^−/− vessel at all pressures studied as well as a lower frequency at pressure = .5 cmH₂O (Figure 1B). Because the absolute amplitude is in part determined by vessel size, which can vary between animals and strains, the summary data in panel C are expressed as amplitude normalized to the maximum passive diameter. This analysis confirms that the normalized amplitudes of ApoE ^−/− vessels were ∼60% of that for ApoE ^+/+ vessels, and these differences were statistically significant at all pressures between .5 and 10 cmH₂O. In contrast, there were no significant differences in tone between ApoE ^+/+ and ApoE ^−/− vessels at any pressure (Figure 1D). ApoE ^−/− vessels had a significantly lower frequency of spontaneous contractions at each the two lowest pressures, .5 and 1 cmH₂O, but comparable frequencies to ApoE ^+/+ vessels at all other pressures (Figure 1E). Calculated flow (FPF) for ApoE ^−/− vessels was 60%–80% of that for ApoE ^+/+ vessels at all pressures, with the differences being significant at about half those pressures (Figure 1F). It should be noted that one vessel was excluded from this analysis because it appeared to be an outlier: 2 of 9 ApoE ^−/− vessels had no contractions at either of the two lowest pressures and 4 of 9 ApoE ^−/− vessels had contraction frequencies <.5 per min at the two lowest pressures, but 1 vessel had a 3x higher frequency and 3x higher tone than the average for ApoE ^+/+ vessels; for that reason it was considered an outlier. Inclusion of that vessel eliminated any statistical differences between ApoE ^+/+ and ApoE ^−/− vessels for tone or frequency but did not change the differences for normalized amplitude or FPF. These results suggest an impairment in contractile function at all pressures in ApoE ^−/− vessels and impairment in the pacemaking mechanism at the two lowest pressures. The failure of the ApoE ^−/− vessels to produce spontaneous contractions at pressures <2 cmH₂O is a pattern that we have observed in other models of contractile dysfunction (Zawieja et al., 2019; Davis et al., 2020).

Next, we conducted tests of lymphatic valve function using methods described previously (Davis et al., 2011; Sabine et al., 2018). For these tests, collecting segments containing only a single valve were studied in Ca²⁺-free Krebs solution to eliminate spontaneous contractions that would interfere with the tests. A diagram illustrating the experimental set-up is provided in Figure 2A. During cannulation and cleaning, the vessel segment was rotated so that the valve leaflets were oriented as in this image. A representative recording of a back-leak test for an ApoE ^+/+ vessel is shown in Figure 2B. Initially, with Pin = Pout = .5 cmH₂O, the valve was open and the local pressure upstream from the valve measured with a servo-null micropipette (Psn) rose as the Pout ramp began until Pout reached ∼1.8 cmH₂O when the valve snapped closed (arrow). After that time Psn stayed at a low level even as Pout reached 10 cmH₂O. The ApoE ^−/− valve in Figure 2C also closed almost immediately after the start of the Pout ramp (arrow) but once Pout reached ∼3 cmH₂O, Psn began to increase until it reached a value of 3.5 cmH₂O at Pout = 10 cmH₂O. The concomitant rise in D upstream from the valve during the Pout ramp confirms that the ApoE ^−/− valve was leaky. The pressure back-leak (Psn-Pin) at Pout = 10 cmH₂O was 0 cmH₂O for the ApoE ^+/+ valve and 3.0 for the ApoE ^−/− valve.

Summary data sets for the two genotypes are shown in Figure 2D. To analyze the changes in Psn during Pout ramps, the Psn recordings were aligned, binned in .5 cmH₂O Pout intervals and an average value ±SEM for each bin was obtained using a LabView program. At Pout = 10 cmH₂O, ApoE ^+/+ valves showed only very slight back-leak (.1 cmH₂O), but back-leak was much more severe for ApoE ^−/− valves and the differences were statistically significant at Pout levels ≥4 cmH₂O.

Next, we tested for possible changes in the ability of valve leaflets to close at fixed levels of an adverse pressure gradient. The experimental set-up was the same as shown in Figure 2. Representative recordings of valve closure tests are shown in Figures 3A,B. With Pin held at .5 cmH₂O, the ApoE ^+/+ valve closed when the Pout ramp reached ∼1.2 cmH₂O (Figure 3A, arrow). Under similar conditions, the ApoE ^−/− valve did not close until Pout reached ∼18 cmH₂O (Figure 3B, arrow).

Summary data for the valve tests are shown in Figures 3C,D. The relationship between the ΔP for closure (i.e. the difference between Pout–Pin at the point of valve closure) and Pin, which was used to set the baseline diameter for the test, is shown in Figure 3C. ΔP for closure values were significantly higher for ApoE ^−/− valves at all pressures except the lowest one. The individual values of ΔP for closure for each valve as a function of normalized diameter are shown in Figure 3D, and are described by a curvilinear function as for other genotypes (Davis et al., 2011). ApoE ^+/+ valves closed at ΔP < 1 cmH₂O when D values were ∼70% of Dmax and required somewhat higher pressures in the range of 2–15 cmH₂O when D was maximal (set by increasing Pin to 10 cmH₂O). The value of ΔP at Dmax is somewhat higher than for other “control” mouse genotypes fed a normal diet (Sabine et al., 2015; Lapinski et al., 2017; Chen et al., 2020; Scallan et al., 2021) and it is not clear if this reflects the influence of the HFD per se. The behavior of about half the ApoE ^−/− valves was similar in that 5 of 11 ApoE ^−/− valves showed a relatively low ΔP for closure even at Dmax. However, 4 of 11 ApoE ^−/− valves closed at low D values but never closed at higher D values and 2 of 11 ApoE ^−/− valves never closed at any diameter tested, even under an imposed adverse ΔP = 30 cmH₂O. As explained in previous publications, we propose that an upward shift in the ΔP for closure vs. D/Dmax relationship reflects an increase in valve leaflet stiffness. Thus, the 3 different types of behavior for the ApoE ^−/− valves likely represent one population of ∼normal valves, one group with increased stiffness that retards closure under physiological conditions, and another group that is intermediate, possibly in transition from normal to increased stiffness.

Because FPF is only an estimate of the pumping ability of a collecting lymphatic vessel and is typically determined under conditions when Pin and Pout are equal, we performed additional tests to directly assess the ability of 2-valve lymphangions to pump against an adverse pressure gradient. Such gradients are generated in lymphatic networks by the intrinsic pumping of each branching order of collectors and are exacerbated by the imposition of gravitational loads (Olszewski, 1977; Olszewski & Engeset, 1980). Indeed, this is the reason why human lymphedema most often occurs in dependent extremities.

For pump tests, a 2-valve segment (i.e., a complete lymphangion) was used. Pin was held constant at a low value while Pout (which is equivalent to afterload under these conditions) was raised ramp-wise (typically to 8–10 cmH₂O), as shown in Figure 4A. Diameter was measured on the upstream side of the outflow valve and playback of the recorded video was used to determine the outflow valve position (either open or closed) in each video frame. Once the Pout ramp began, the outflow valve was closed except for a brief period of time at the peak of each lymphatic systole when the valve transiently opened [as the internal pressure spike exceeded Pout (Davis et al., 2011)]. As Pout continued to rise, the ability of the pump to eject its contents was exceeded at some point, as evident by the failure of the output valve to open during a contraction. The pump limit, P_limit, was defined as the value of Pout (minus Pin) at the time of last successful ejection. In the representative examples shown in Figure 4, P_limit for ApoE ^+/+ vessel (panel A) was 6 cmH₂O (after subtraction of Pin) and P_limit for the ApoE ^−/− vessel was 2.2 cmH₂O (panel B). The pump test data are summarized in Figure 4D and show that popliteal collectors from ApoE ^−/− mice have a significantly lower pump strength (1.2 vs. 4.8 cmH₂O) than comparable vessels from ApoE ^+/+ mice, in agreement with the lower calculated pump flows for ApoE ^−/− vessels in Figure 1F.

A complicating factor in this protocol was the tendency for the outflow valve to sometimes “lock” into the open position during the Pout ramp; it can be forced into this position if the inflow valve ceases to open in diastole. This phenomenon is described in a previous publication (Bertram et al., 2017) and occurs if the respective local adverse pressure gradients across the two valves favor closure of the inflow valve rather than the outflow valve. In such cases, we determined P_limit from the previous (successful) ejection prior to outflow valve “lock”. This likely underestimated the actual P_limit, but a minimal estimate of P_limit could at least be made. An example of this behavior is shown in Figure 4A (on the contraction after the arrow). As shown in this case, output valve “lock” was also accompanied by a sudden increase in D (coinciding with the vertical line in panel A); prior to that time D declined slightly with rising Pout (Davis et al., 2009; Davis et al., 2012b). Valve “lock” did not occur in the ApoE ^−/− vessel; in that case D increased gradually prior to ejection failure, probably due to back-leak (Figure 4B).

Next, we measured the permeability of collecting lymphatics to albumin. Previous observations in ApoE ^−/− mice showed that Evans Blue, after injection into the interstitium of the ear and uptake into initial lymphatics, appeared to leak from the collecting vessels draining the ear (Lim et al., 2009). We confirmed those results in the hindlimb, as shown in Figures 5A,B. The injection of Evans blue into the foot resulted in its subsequent appearance in the two main popliteal lymphatics in the calf. In ApoE ^+/+ mice, the borders of these vessels were sharply defined, whereas in ApoE ^−/− mice focal spots of dye leakage from the popliteal collectors were evident in ApoE ^−/− mice (arrows). The near-uniform spacing suggests that the leakage was possibly occurring at valve sites. Because subsequent analysis of dye leakage would have been highly subjective, we quantified the permeability of popliteal collectors using ex vivo methods similar to those previously described in detail (Scallan et al., 2015).

For permeability measurements, a collecting lymphatic was dissected from the mesentery, taking advantage of the lack of strong spontaneous contractions of lymphatics in this tissue bed. After cannulation and cleaning, a vessel was perfused through a ‘theta’ micropipette that contained two identical solutions, one of which contained albumin tagged with a fluorophore. Initially, a washout solution was perfused to obtain a baseline recording on the photometer. Switching to the other half of the micropipette selectively perfused the fluorescent albumin without a change in intraluminal pressure. Over time, fluorescent albumin moved across the lymphatic wall. The photometer voltage increased rapidly after the solution change, followed by a slow increase due to albumin movement across the vessel wall. Permeability (Ps) could be determined from the slope of the steady-state portion of the recording, which was fit using linear regression. Measurement of mesenteric collecting lymphatic Ps revealed a loss of barrier function in ApoE ^−/− vessels (109 × 10^–7 cm/s; n = 15) compared with ApoE ^+/+ controls (21 × 10^–7 cm/s; n = 20) that was statistically significant (Figure 5C). Further analysis of the data revealed that there were two populations of ApoE ^−/− vessels, one with normal permeabilities (ApoE ^−/− Low) and one with significantly elevated permeabilities (ApoE ^−/− High). The reasons for these differences are not clear but suggest that some ApoE ^−/− vessels may be more susceptible to damage by whatever combination of factors is mediating the hyperpermeability.

Subsequent imaging of the valve regions of popliteal collectors in situ revealed that many valves in ApoE ^−/− mice also had reduced LMC coverage in their sinus regions. Three examples are shown in Figure 6 along with three examples from ApoE ^+/+ vessels, in which the green fluorescence signal denotes SMA staining to mark the locations of LMCs, with unusually large gaps occurring at valve sites; valves at those sites were identified based on CD31 staining (not shown).

Despite existing evidence that hypercholesterolemia impairs lymphatic drainage, it was unknown whether functional alterations in collecting lymphatics contributed to this impairment. Through comprehensive analyses of collecting lymphatic function using vessels from hypercholesterolemic ApoE ^−/− mice, we demonstrate here that hypercholesterolemia significantly compromises the contractile, valve and barrier functions of collecting lymphatics.

Hypercholesterolemia is the major risk factor for atherosclerotic disease and is also a common clinical feature in diabetes, obesity and some autoimmune diseases. Obesity relates to the accumulation of excess body fat, as determined by the measurement of BMI, and is also a major risk factor for cardiovascular diseases. Although obesity is frequently associated with hypercholesterolemia and the combination of both conditions worsens cardiovascular risk status, many obese patients exhibit a normal blood lipid profile (Miettinen, 1971). Likewise, many hypercholesterolemic patients (Guerrero-Romero & Rodriguez-Moran, 2006) as well as ApoE ^−/− mice (Schreyer et al., 2002; Chiba et al., 2003; Hofmann et al., 2008) are not obese. However, obvious similarities between obesity and hypercholesterolemia include the deposition of lipids in extravascular tissues in the form of white adipose expansion or foam cells and the association of lipids with impaired lymphatic drainage (Khan et al., 2022).

A number of studies using mouse models of obesity induced by HFD have reported a clear association between obesity and impaired lymph transport (Weitman et al., 2013; Blum et al., 2014; Savetsky et al., 2014; Nitti et al., 2016). The impairment of lymphatic flow is in part due to vessel wall dilation and in part to reduced contractile activity of collecting lymphatics (Blum et al., 2014). The degree of lymphatic dysfunction appears to be related to the composition of the high fat diet. In obesity induced by feeding mice a Western diet, impairment of lymphatic flow was associated only with collecting lymphatic valve dysfunction and not impaired contraction amplitude (Castorena-Gonzalez, 2022); this is in contrast to obesity induced by HFD, which is characterized by both contractile and (apparent) valve dysfunction (Blum et al., 2014; Nitti et al., 2016). These differences in phenotype may be explained by differences in the percentage of fat present in the high fat and Western diets (60% in high fat diet versus 40% in Western diet). Like the effect of adipose tissue expansion on collecting lymphatics, hypercholesterolemia in ApoE ^−/− mice reduced the contractile function of collecting vessels (Figures 1, 4) and compromised valve function. Lymphatic valves in ApoE ^−/− mice were characterized by increased back-leak (Figure 2) and an increase in ΔP for closure (Figure 3), which we infer reflects an increase in leaflet stiffness, as supported by modelling studies (Bertram, 2020); however, actual stiffness measurements (e.g., reflected as a shift in the biaxial stress-strain relationship of the leaflets) are not presently feasible in these small structures. Notably, valve dysfunction in the Western diet-induced obesity model was only observed in mesenteric and not in popliteal collecting lymphatics (Castorena-Gonzalez, 2022), in contrast to the hyperpermeability of mesenteric collectors and both valve and contractile dysfunction in popliteal collectors of ApoE ^−/− mice fed a HFD. Moreover, the levels of valve dysfunction in ApoE ^−/− mice (Figures 2, 3) were greater than those in mice fed only a Western diet [see Figure 2 in (Castorena-Gonzalez, 2022)]. Although the levels of cholesterol increased in Western diet fed mice from 80 mg/dl to 120 mg/dl, these levels remain rather low compared to those in hypercholesterolemic ApoE ^−/− mice which reach 1,000–2000 mg/dl. These comparisons further support the idea that diet composition may differentially affect collecting vessel function and appropriate caution should be used when comparing results from diet-induced obesity and hypercholesterolemic mouse models.

The differences between the phenotypes of mice on a HFD vs. Western diet also raise the issue of the influence of perivascular fat on lymphatic function. Although there were no obvious differences in the amount of fat associated with ApoE ^+/+ and ApoE ^−/− popliteal vessels in vivo, as indicated by the images in Figure 5A, we cannot state whether there were quantitative differences in vessel-associated fat between the two strains. However, in all of our ex vivo functional tests, the fat was removed from the lymphatic vessel in order to allow internal diameter measurement and/or visualization of the valve(s). Thus, any acute effects of perivascular fat were eliminated. Vessels in which all perivascular fat remained intact would not have be amenable to tests of valve function, pump function or permeability. Although studies such as Harvey et al. (Harvey et al., 2005) and Cao et al. (Cao et al., 2021) have demonstrated an association between fat accumulation and lymphatic collector permeability, to our knowledge, there are no studies looking at the influence of perivascular vessel adipose tissue (PVAT) on lymphatic contractile function or valve function using techniques such as those routinely used in arteries to show the modulatory effects of PVAT (Agabiti-Rosei et al., 2018; Hillock-Watling & Gotlieb, 2022). This remains an area for further investigation.

In the present study, we also detected an elevated permeability of collecting vessels in hypercholesterolemic ApoE ^−/− mice, as previously noted for collecting vessels of diabetic mice (Scallan et al., 2015). Although our analysis revealed that hypercholesterolemia significantly altered the barrier function of collecting lymphatics, hyperpermeability was only characteristic of half the ApoE ^−/− vessels (Figure 5). Although the permeability of this cohort of vessels was elevated 5-fold, that elevation was rather modest compared to the 130-fold increase in permeability measured previously in db/db lymphatic vessels (Scallan et al., 2015), and the other half of the ApoE ^−/− vessels had normal permeability to albumin. The reasons why only some vessels were susceptible remain unknown. Lymphatic hyperpermeability has been previously demonstrated in another HFD mouse model (Cao et al., 2021) as well as in Apelin ^−/− mice (Sawane et al., 2013), and in both cases the hyperpermeability was reversed to a large extent by COX2 inhibition, suggesting that activation of COX2 may be a common mechanism underlying elevated lymphatic permeability. Whether COX2 activation underlies the permeability changes of ApoE ^−/− vessels in the present study remains to be determined.

Although further study is needed to elucidate the mechanisms by which hypercholesterolemia impairs collecting lymphatic function, several factors may underlie the altered function of LECs and LMCs in this context. The first possibility is that collecting vessels may accumulate cholesterol, which in turn affects their function, as observed in arteries of ApoE ^−/− mice. This might be addressed by additional analyses of vessels by light and electron microscopy. Like obesity, hypercholesterolemia is associated with increased oxidative stress and the systemic and/or local production of reactive oxygen species (ROS) and nitric oxide (NO). Various ROS molecules, such as H₂O₂, are known to impair contractile activity (Zawieja et al., 1991; Zawieja & Davis, 1993), in part through the activation of ATP-sensitive K⁺ (K_ATP) channels in LMCs (Davis et al., 2022); increased K⁺ efflux through this channel leads to LMC hyperpolarization and the slowing of spontaneous contractions (Davis et al., 2022). It is possible that K_ATP channel activation contributes to the lower frequency of spontaneous contraction in ApoE ^−/− vessels (Figure 1E), as was previously shown for mesenteric lymphatic collectors from rats fed a high fructose diet (Zawieja et al., 2016), but that possibility remains to be tested. NO production may also inhibit lymphatic contractions, but the effects of NO on lymphatic contractile function are complex. Low NO concentrations, such as those associated with pulsatile flow from spontaneous contractions, enhance diastolic filling and contraction amplitude (Gasheva et al., 2006), but higher NO concentrations, such as those associated with iNOS activation and cytokine production (Liao et al., 2011; Torrisi et al., 2016) inhibit both contraction frequency and amplitude (Scallan & Davis, 2013; Rehal & von der Weid, 2015; Chen et al., 2017; Kim et al., 2021; Davis et al., 2022). The contributions of ROS and NO to the impaired contractile responses of ApoE ^−/− vessels shown in Figures 1, 4 remain to be determined.

The increased oxidative environment in hypercholesterolemia is also known to participate in the modification of low-density lipoproteins. These modified lipoproteins are then taken up by macrophages through scavenger receptors, which in turn become foam cells. Interestingly, two recent studies revealed the expression of scavenger receptors for modified low-density lipoprotein including MSR-1 (SR-A1), MARCO (SR-A6), and CD36 on lymphatic endothelial cells (Berendam et al., 2019; Fujimoto et al., 2020). This raises the possibility that LEC function, including the regulation of permeability, is altered through the uptake of these modified low-density lipoproteins.

Hypercholesterolemia may also compromise the function of collecting lymphatics by altering their structure. This idea is consistent with our observations of reduced LMC coverage in the valve sinus regions of collecting vessels (Figure 6) as well as relatively high levels of LYVE-1 expression (Lim et al., 2009). LYVE-1 is normally downregulated in LECs of mature collecting lymphatics (compared to lymphatic capillaries), suggesting that either complete vessel maturation is retarded or a reversion to a less differentiated phenotype in ApoE ^−/− collectors is triggered. The loss of SMC coverage in the valve regions may explain some of the contractile amplitude impairment, but that seems unlikely to be the complete explanation as the LMC loss was highly localized. Interestingly, at least two other conditions are also associated with LMC loss from collecting lymphatics and impaired contractions, including overproduction of the cytokine TNFα (Kenney et al., 2022a) and exposure to bacterial toxins (Jones et al., 2018). A common theme underlying LMC loss in both the cases, and possibly in ApoE ^−/− mice on HFD, may be elevated cytokine and/or ROS production (Kenney et al., 2022b). LMC loss at valve sinuses could theoretically impact permeability by disrupting LEC-LMC interactions or by increasing the fragility of the sinuses so that they become more susceptible to pressure overload. Decreased FOXC2 expression in LECs of ApoE ^−/− mice (Lim et al., 2013) may also contribute to the observed valve dysfunction in these mice as FOXC2 deficiency has been shown to compromise valve function by enhancing back-leak and increasing leaflet stiffness, making it more difficult for the valves to close (Sabine et al., 2015; Castorena-Gonzalez et al., 2020). An increase in leaflet stiffness would point to possible changes in the extracellular matrix in the valve leaflets and/or the vessel wall, similar to the increased arterial stiffness known to be associated with collagen accumulation in the remodelled arterial wall (Bakker et al., 2005; Martinez-Lemus et al., 2011). Consistent with this hypothesis, obesity-associated fibrosis has been proposed to contribute to reduced lymph transport in a mouse model of secondary lymphedema (Savetsky et al., 2014).

Given the role of lymphatic vessels in immunity (Oliver et al., 2020), the dysfunction of lymphatics observed in hypercholesterolemic mice is expected to affect the immune response. Indeed, poor lymphatic drainage in these mice may compromise the transport of antigens, trafficking of immune cells, antigen presentation and downstream T-cell activation. This is in line with previous studies from our labs and others revealing impaired dendritic cell migration (Lim et al., 2009; Lim et al., 2013) and T-cell egress via lymphatics in ApoE ^−/− mice (Tay et al., 2019) as well as higher susceptibility to bacterial, viral or fungal infections in these mice (Moazed et al., 1997; Netea et al., 1997; de Bont et al., 1999; Ludewig et al., 2001; Martens et al., 2008).

Finally, lymphatic dysfunction in ApoE ^−/− mice may also facilitate the progression of atherosclerosis by interfering with the efficiency of reverse cholesterol transport, an atheroprotective mechanism dependent on lymphatic transport (Lim et al., 2013; Martel et al., 2013). This conclusion is supported by our previous data demonstrating that the ligation of lymphatic vessels draining the aorta promotes the accumulation of cholesterol in the arterial wall (Lim et al., 2013; Martel et al., 2013). Lymphatic drainage may also aid in the clearance of inflammatory mediators that contribute to the progression of atherosclerosis. Moreover, we recently presented evidence that the efficient lymphatic drainage of the aorta is required for the therapeutic effect of the cholesterol-lowering drug, ezetimibe (Yeo et al., 2020). Therefore, targeting lymphatic function would seem to be a promising approach to treat cardiometabolic diseases.