Lymphatic remodelling in response to lymphatic injury in the hind limbs of sheep

Abstract

Contractile activity in the lymphatic vasculature is essential for maintaining fluid balance within organs and tissues. However, the mechanisms by which collecting lymphatics adapt to changes in fluid load and how these adaptations influence lymphatic contractile activity are unknown. Here we report a model of lymphatic injury based on the ligation of one of two parallel lymphatic vessels in the hind limb of sheep and the evaluation of structural and functional changes in the intact, remodelling lymphatic vessel over a 42-day period. We show that the remodelled lymphatic vessel displayed increasing intrinsic contractile frequency, force generation and vessel compliance, as well as decreasing flow-mediated contractile inhibition via the enzyme endothelial nitric oxide synthase. A computational model of a chain of lymphatic contractile segments incorporating these adaptations predicted increases in the flow-generation capacity of the remodelled vessel at the expense of normal mitochondrial function and elevated oxidative stress within the lymphatic muscle. Our findings may inform interventions for mitigating lymphatic muscle fatigue in patients with dysfunctional lymphatics.

Figure 1:. Sheep lymphatic anatomy and lymphatic remodeling after surgery

(A) Schematic of the hindlimb lymphatic and venous circulation. (B) 3D MRI image of the lymphatic anatomy of the hind limbs was conducted one day pre-surgery and 42 days after surgery rotated so that both of the limbs can be visualized. Pre-surgical MRI shows the two primary collecting lymphatic vessels that drain the hind limbs. These two vessels run in parallel and both drain to the popliteal lymph node. At 42 days post-surgery the caudal vessel is absent from the MR images on the wounded leg. This is a representative image set, MRI images were taken from three sheep. (C) NIR imaging allowed visualization of network remodeling due to the lymphatic ligation. Pre-surgical images show both the cranial and caudal collecting lymphatics intact. Immediately after ligation of the caudal collecting lymphatic, an absence of fluorophore transport confirmed successful ligation and excision of the vessel. This is a representative image set, images were taken from all five sheep. (D) The diameter of the cranial vessel on the control and wounded leg was measured distal to the wound site using the MR images before and 42 days after surgery (n=3). Values were normalized by the pre-surgical average diameter of the respective vessel. Mean ± standard error. Statistical significance was determined with a paired two-sided Student’s t-test.

Figure 2:. Lymphatic function assessed via NIR imaging

NIR lymphatic imaging was used to non-invasively evaluate lymphatic functional changes in response to the lymphatic injury. NIR imaging data was collected from the intact vessels on both the control (n=5) and wounded (n=5) legs. (A) Packet frequency, the frequency of fluorescent pulsations within the vessel, exhibited significant differences between intact vessels on the control and wounded legs at day 28 post-surgery. (B) The percent difference between the packet frequencies of the control and wounded leg depict a decline due to the surgery. (C) Packet transport, the integral of the packet frequency signal over time, was used to evaluate fluorescence transport driven by intrinsic lymphatic contractions. Significant differences in packet transport were observed at day 28. (D) Effective pumping pressure (Peff), was used to evaluate the maximal pressure generated by the collecting vessel to overcome occlusion. Significant differences in Peff were observed at day 28. Mean ± standard error. (*, significant difference between control and wounded evaluated using 1-way ANOVA with post-hoc Dunnett’s test for multiple comparisons, p < 0.05; ¥, significant difference between baseline value at Day 0, p < 0.05, exact p-values between comparisons are as noted directly on the figure).

Figure 3:. Isolated vessel contractile response to varying transmural pressure

Intact lymphatic vessels were isolated from the control and wounded legs to evaluate changes in the stretch response due to the surgery. Vessel segments were cannulated and exposed to a range of transmural pressures. (A) Wire myography was used to evaluate the contractile force of multiple segments cut from a single control and remodeled vessel (non-pressurized) and exhibited significant increases in the contractile force of the remodeled vessel from the wounded leg (n=4 control, n=4 remodeled). (B) Vessel segments from the wounded leg exhibited higher contraction frequencies than control segments (n=4 control, n=4 wounded). (C) The amplitude of lymphatic contractions did not significantly differ between the control and wounded groups (n=4 control, n=4 wounded). (D) Lymphatic tone was not significantly different in vessel segments collected from the wounded leg (n=4 control, n=4 wounded). Mean ± standard error are plotted and open circles represent individual data points. Significant difference between control and wounded groups was evaluated at their respective transmural gradient using 2-way ANOVA with post-hoc Sidak correction for multiple comparisons. In addition to this, 1-way ANOVA with post-hoc Dunnett’s correction for multiple comparisons was used to determine significant differences between baseline transmural pressure and the applied transmural pressure for control or wounded groups respectively (*, significant difference between control and wounded, p < 0.05; ***, significant difference between control and wounded, p < 0.001, †, significant difference between control and control baseline at 1.5 cmH₂O transmural pressure, p < 0.05; ‡, significant difference between wounded and wounded baseline at 1.5 cmH₂O transmural pressure, p < 0.05, exact p-values between comparisons are as noted directly on the figure, all n-values are independent experiments/animals)

Figure 4:. Remodeled vessels exhibit impaired flow-mediated dilation

Intact lymphatic vessels were isolated from the control and wounded legs to evaluate changes due to remodeling on the flow-mediated contractile response. Vessel segments were cannulated and exposed to a range of axial pressure gradients in order to vary flow through the vessel while holding transmural pressure constant. 100uM LNAME was used to inhibit nitric oxide production. (A) The frequency of lymphatic contraction was significantly higher for isolated vessels collected from the wounded leg when compared to control when flow was imposed with an axial pressure gradient of 1 or 3 cmH₂O across the vessel. However, this statistical significance of this difference between the control and remodeled vessel was lost when the flow was elevated by raising the axial pressure to 5 cmH₂O or when vessels were treated with L-NAME (n=3 control, n=5 wounded). (B) There were no significant differences in contraction amplitude (given as a percent of the systolic/diastolic diameter) between the control and remodeled vessel as a response to flow or L-NAME (n=3 control, n=5 wounded) (C) There were no significant changes to the tone as a result of flow or L-NAME application (n=3 control, n=5 wounded). (D) The vessel from the wounded leg did exhibit a response to flow, however this response occurred at a much higher shear stress than for the control vessel indicating a lower sensitivity to shear (n=1 control, n=1 wounded). Significant difference between control and wounded groups was evaluated at their respective axial pressure gradient using 2-way ANOVA with post-hoc Sidak correction for multiple comparisons. In addition to this, 1-way ANOVA with post-hoc Dunnett’s correction for multiple comparisons was used to determine significant differences between baseline axial pressure and the applied axial pressure for control or wounded groups respectively. Mean ± standard error are plotted and open circles represent individual data points (*, significant difference between control and wounded, p < 0.05, exact p-values between comparisons are as noted directly on the figure)

Figure 5:. Proteomic analysis of isolated lymphatic muscle cells

Comparative protein content profiling of control (n=5 biological replicates) vs remodeled (n=5 biological replicates) lymphatic muscle cells together with the molecular and cellular pathways predicted by IPA. (A) Un-biased, non-clustered heat map generated in “perSPECtives” using the normalized weighted spectral counts highlights the change in the protein content profiles in the “wounded” relative to the “control” lymph muscle cells. (B) The main canonical pathways affected by the top 105 proteins with higher than twofold expression level in the “wounded” vs “control” samples, subjected to analysis by the IPA algorithm. The % of total proteins increased- (red color) or decreased (green color) in each pathway together with the corresponding log (-p value) are displayed. (C) Protein networks significantly regulated (p< 0.05) in the “wounded” relative to the “control” lymph muscle cells are displayed as IPA predicted diseases and cellular functions using the z score (down-regulated corresponding to z<-1.5 and blue color; up-regulated, corresponding to z>+1.5 and orange color). The size of each rectangle is proportional with the value of the z score associated to a specific pathway. Increased organismal injury, cell death, small molecules and lipids metabolism and cellular movement are the main diseases and cellular pathways characterizing the “wounded” system. Only proteins which passed a selected significance statistical threshold (ANOVA, p < 0.05 and FDR < 1% for protein and peptide expression) and had been identified with at least two-fold differential expression across all samples are shown in the representative heat map.

Figure 6:. Structural and mechanical properties of isolated lymphatic vessels

Vessels from the wounded leg displayed structural and mechanical changes caused by the ligation surgery. (A) Vessel cross sections from the wounded leg (left column- transverse, right column- axial) exhibited atrophy of the musculature surrounding the vessel compared to control. Additionally, vessel segments from the wounded leg appear to qualitatively have an increase in matrix surrounding the vessel (n=2 control, n=2 wounded). (B) Biaxial testing was conducted on isolated vessel segments at various axial stretches and transmural pressures. Remodeled vessels exhibited a decrease in vessel stiffness compared to control. The biaxial data was fit to a constitutive model to calculate material parameters that capture pressure-diameter behavior (details in theory).

Figure 7:. Computational simulation of lymphatic chain performance

Isolated vessel function and mechanical properties were utilized to inform a computational model, which was used to elucidate how the structural and functional changes caused by the surgery would affect intrinsically driven flow. (A) Mechanical parameters coupled with ex-vivo isolated vessel frequencies were used to simulate predicted lymph flow rates as a function of pressure generation. Our computational model suggests that vessel segments from the wounded leg may exceed the flow rate of control by increasing contraction frequency and contractile force. (B) Pressure-volume curves for these simulations. (C) Mechanical parameters coupled with in-vivo contraction frequencies assessed via NIR imaging at day 42. Our computational model suggests that, in an in-vivo context, vessel segments from the wounded leg may have decreased intrinsically driven lymph flow rates. (D) Pressure-volume curves for these simulations. (E) Biaxial testing data was used to inform a computational model that predicts active and passive wall stress as a function of circumferential stretch. Testing data suggests an increase in the passive and active compliance of the vessels from the wounded leg. (F) The active tension parameter from the computational model of the remodeled vessel was adjusted ± 3% to determine the predicted effect of the tension change on pump function.