Decreased potential for lymphatic vessel generation is a hallmark of early diagnosed arterial hypertension and can be reversed by treatment with angiotensin converting enzyme inhibitors

Abstract

Arterial hypertension is the main risk factor for cardiovascular diseases. It is characterized by inflammation and impaired generation of blood microvasculature leading to organ damage. While the immune system is an important actor involved in both phenomena, we hypothesized that hypertensive patients show alterations in numbers of regulatory T cells (Tregs) and proangiogenic CD31 + CXCR4+ T cell subsets, which lead to alterations in serum cytokine pattern and decreased generation of blood and lymphatic vessels. Sixteen recently diagnosed hypertensive patients before treatment implementation were enrolled and followed-up for 2 years. Control group comprised of 14 healthy individuals. Each patient underwent echocardiography, eye fundus examination, optical coherence tomography, impedance cardiography, 24-hour blood pressure monitoring and central arterial pressure waveform analysis. Th and Treg cells double positive for CD31 and CXCR4, as well as subsets of naive and memory Tregs were analyzed with flow cytometry. The patients’ serum cytokine pattern was defined with Luminex and ELISA. Functional in vivo assay enabled evaluation of pro-angiogenic and pro-lymphangiogenic activity of the patients’ sera before and after hypotensive treatment implementation. Serum of recently diagnosed hypertensive patients showed decreased potential for generation of new lymphatic vessels in animal model. Two years after treatment implementation, the lymphangiogenic potential of the patients’ sera was restored only in patients treated with angiotensin converting enzyme inhibitors (ACEI). Increased capacity for lymphatic vessel formation was associated with improved renal function (decreased microalbuminuria, increased GFR). Increased lymphangiogenic potential of the serum in hypertensive patients was associated with its cytokine profile, notably increased concentrations of VEGF-C and MDC (CCL22) and decreased levels of MIP-1α and MIP-1β. Therapy with ACEI prevented also loss of CD31 + CXCR4+ Th and CD31 + CXCR4+ Treg cells which correlated with better renal function and lower stiffness of vessel wall. Lower numbers of all Tregs and shift from Tcm to Tem phenotype in Treg population were a hallmark of fast progression of microangiopathies in hypertensive patients not treated with ACEI. In conclusion, our data indicate that early diagnosed arterial hypertension is characterized by decreased potential for lymphatic vessel generation . However, it can be reversed by ACEI implementation leading to improved renal function, lower stiffness of vessel wall and preservation of anti-inflammatory and proangiogenic T cell subsets. The current study sheds new light on interactions between immune system and lymphangiogenesis in hypertension. Therapy with ACEI may prevent or significantly delay development of vascular complications in hypertension.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-40323-x.

Introduction

Arterial hypertension belongs to the most frequent chronic diseases with global prevalence of over 30%¹. It is a well-known risk factor for cardiovascular complications, stroke, vision loss and renal insufficiency, leading to death of 8.5 million people worldwide². Therefore, early detection and improved treatment of hypertension are of great importance. The disease is characterized by reduction in density of arterioles and capillaries called microvascular rarefaction³. Rarefaction significantly increases peripheral resistance, which results in elevated blood pressure and hypertension-induced organ damage⁴. Constant balance between de novo angiogenesis and microvascular regression is hampered in hypertensive patients and contributes to microvascular rarefaction. However, still not much is known about the mechanisms that stay behind this process and how they change over time. Moreover, there is very scarce data available regarding early diagnosed hypertension before hypotensive drug administration, which are of great importance for understanding pathogenesis of the disease.

It was found that both leukocytes and endothelium actively modulate the disease outcome, as they are important sources of factors involved in generation of new blood and lymphatic vessels. Machnik and colleagues reported that tissue hypertonicity activates tonicity-responsive enhancer binding protein (TonEBP) that leads to upregulation of vascular endothelial growth factor C (VEGF-C) in monocytes⁵. Then, monocyte-derived VEGF-C drives lymphangiogenesis. This way monocytes act as onsite controllers of interstitial volume and contribute to blood pressure homeostasis⁶,. Simultaneously, we have found that CD4 + T cells from adolescents with type 1 diabetes produce significantly more VEGF-A, another member of VEGF family, than the cells from healthy individuals⁷. Interestingly, the pro-angiogenic activity of T cells was subset dependent. CD4 + effector memory T cells (Tem) were found to be the main source of VEGF-A among conventional CD4 + T cells. Moreover, Hur et al. described a subset of T cells that promoted formation and repair of blood vessels in vitro and in vivo⁸. These pro-angiogenic lymphocytes were found to express platelet endothelial cell adhesion molecule (CD31) and CXCR4, a receptor for stromal-derived factor 1 (SDF-1). Interestingly, these pro-angiogenic lymphocytes were demonstrated to be indispensable for endothelial progenitor cell (EPC) colony formation and early differentiation. Further analysis of clinical samples revealed that their frequency in peripheral blood reflects pro-angiogenic potential of human EPCs in vitro⁹. In addition, Hill et al. showed an inverse correlation between the number of EPC colonies and cardiovascular risk factors¹⁰.

Impact of immune cells on angiogenesis and clinical outcome of hypertension is in strong association with inflammation. Elevated serum levels of proinflammatory mediators are a hallmark of arterial hypertension and development of its complications¹¹. Indeed, excessive secretion of proinflammatory cytokines by immune cells is a commonly accepted risk factor for the generation of atherosclerotic plaque in the vessel wall^12,13. Recently, Loperena et al. showed that increased endothelial stretch promoted monocyte differentiation into CD14+/CD16 + + intermediate phenotype. Moreover, the cells acquired CD209 (DC-SIGN) a marker of activation and showed increased transcription of interleukin (IL)−6, IL-1β, IL-23, CCL4 (C-C motif chemokine ligand) 4, and tumour necrosis factor α (TNFα)¹⁴. Interestingly, it was shown that variety of proinflammatory cytokines, including listed above IL-6^15–17, IL-1β¹⁸and TNFα^19–24 are involved in the initiation and progression of hypertension²⁵. Simultaneously it is well known that excessive inflammatory response leads to endothelial injury and development of complications in hypertensive patients^14,25.

Prominent role of T cells in vessel formation and deleterious impact of inflammation in hypertension suggest that immunoregulatory cell populations might play a significant role in pathogenesis and progression of arterial hypertension and other circulatory system diseases. It was reported previously that patients suffering from cardiovascular disease (including acute myocardial infarction, stable and unstable angina) have reduced numbers of CD4 + CD25+FOXP3 + regulatory T cells (Tregs) as compared with healthy individuals²⁶. In addition, an imbalance between anti-inflammatory Tregs and proinflammatory Th-17 cells was reported in individuals with unstable carotid artery plaque, favouring the latter subsets²⁷. Nevertheless, knowledge regarding Tregs in hypertension in human is still limited. In addition, there is lack of reports if CD31 and CXCR4 positive cells can be also identified within Treg population. Therefore, in the current study we decided to test if CD31 + CXCR4+ Treg subset exists, and if yes- what proportion of total Treg pole they constitute and how they are connected with hypertension, angiogenesis and patient clinical status. We found this topic interesting as most of the previous studies on Tregs and hypertension studied very limited Treg phenotype and used angiotensin II induced animal model of the disease. In general the link between hypertension, its complications and Tregs was reported in rodent studies²⁸. For example, hypertension in mice was associated with decreased frequency of Tregs in peripheral blood even before onset of the disease. Tregs were reported to prevent endothelial dysfunction in coronary arterioles in hypertensive animals. Interestingly, their adoptive transfer was proven to have therapeutic effect (e.g. prevention of vascular stiffness) due to their anti-inflammatory activities^29–31. However, only recently Gackowska et al. reported a reduced numbers of naive Tregs in hypertensive children³². Taking into account physiological role of Tregs, these observations are not surprising. Tregs prevent excessive inflammation and control all cells of the innate and adaptive immunity. Therefore, in the current study we hypothesized that patients with recently diagnosed hypertension are characterized by lower potential for blood and lymphatic vessel generation due to alterations in serum cytokine pattern and frequency of CD31 + CXCR4+ T cell subsets. We have also hypothesized that serum cytokine pattern and numbers of studied T cell subsets are sensitive indicators of the blood pressure control. To verify this finding we analysed proportions of various CD4+ T cell subsets, including thymus derived CD4+CD25HighFoxP3+ Tregs (Tregs), CD4+CD31+CXCR4+ T cells (CD31+CXCR4+Th) and CD31+CXCX4+ Tregs (CD4+CD25HighFoxP3+CD31+CXCR4+) in the individuals with recently diagnosed hypertension before treatment implementation. The patients were followed-up for 2 years and tested annually. Simultaneously, the in-depth analysis of the patients’ serum cytokine pattern and functional in vivo studies enabled deep characterization of pro-angiogenic and pro-lymphangiogenic activity of the patients’ sera before and after hypotensive drug implementation that led to blood pressure control.

Materials and methods

Patients

Sixteen patients with recently diagnosed arterial hypertension before hypotensive drug implementation were recruited into the study. The control group comprised of 14 age, gender and BMI matched healthy individuals. The exclusion criteria were: coexisting autoimmune disease, history of neoplasm, smoking and treatment with drugs that influence angiogenesis and/or lymphangiogenesis. Patients were followed-up for 2 years and examined three times in annual intervals (HT1, HT1 and HT3 control points). The study was conducted in accordance with the guidelines of the Declaration of Helsinki and under the protocol approved by the Independent Bioethics Commission for Research of the Medical University of Gdańsk (agreement no. NKEBN/172/2010). All individuals gave written informed consent and all experiments were performed in accordance with relevant guidelines and regulations.

Hypertension was diagnosed when SBP was ≥ 140 mmHg or DBP was ≥ 90 mmHg. During the visits the following parameters were examined: BMI, heart rate, systolic (SBP) and diastolic blood pressure (DBP). Each patient underwent echocardiography and abdomen ultrasound with renal Doppler examination, eye fundus examination and optical coherence tomography (OCT; evaluation of hypertensive retinopathy), impedance cardiography (Niccomo Cardioscreen Gmbh), annual ambulatory 24-hour blood pressure monitoring (ABPM; Schiller) and central arterial pressure waveform analysis (SphygmoCor). In all patients SBP and DBP were found to decrease 10–20% at night as compared to daytime values. The rule was observed during the entire follow-up (at HT1, HT2 and HT3 control points) indicating low risk of cardiovascular diseases and kidney insufficiency what was probably associated with early stage of hypertension. The Table 1 presents mean results of daytime readings during ambulatory 24-hour blood pressure monitoring to show a complete picture of pressure changes and drug effects during patients’ normal activity. The laboratory tests were carried out in the Central Laboratory at the University Clinical Center in Gdansk where the following parameters were measured in the patients’ blood: creatinine, C-reactive protein (CRP), glucose (Glc), platelets (PLT), total cholesterol, triglycerides (TAG), high-density lipoprotein (HDL) and low-density lipoprotein (LDL). In addition, an individualized monitoring schedule was established for each patient to assess serum levels of potassium (K⁺), sodium (Na⁺), calcium (Ca²⁺), and magnesium (Mg²⁺) ions. No electrolyte imbalances were observed during the follow-up period. Regular evaluation of glomerular filtration rate (GFR) and microalbuminuria (hallmark of renal microangiopathy) were also performed. Clinical characteristics of the studied groups is presented in Table 1. After examination at the HT1 control point the hypotensive drugs were implemented according to the guidelines of the European Society of Hypertension (ESH)¹. The hypotensive drugs applied, including combination therapy, are listed in Table S1. Based on the treatment applied the patients were stratified at HT3 control point into two groups ACEI (treated with ACEI in mono or combination therapy) and Other (various hypotensive drugs with exception of ACEI and angiotensin II receptor blockers; Table S1). For some analyses the patients at HT3 control point were also divided into Angiopathy_1 and Angiopathy_2 subgroups that differed in severity of angiopathy. The angiopathy stage was assign according to grade of hypertensive retinopathy and renal disfunction.

Table 1.

Clinical characteristics of the studied groups. The table presents clinical characteristics of the healthy individuals (Control) and hypertensive patients at 3 different control points during 2-year follow-up (HT1, HT2 and HT3). Data are presented as medians with minimum and maximum values within the brackets. Differences between the groups were calculated with MW test. Statistically significant differences (p < 0.05) between healthy individuals (Control) and hypertensive patients before the treatment implementation (HT1) are marked with *. ** and € refer to statistically significant differences between the patients at HT1 vs. HT3 and HT1 vs. HT2 control points, respectively (MW test).

n	HT1	HT2	HT3	Control
	16	10	16	14
Sex (female/male)	9/7	4/6	9/7	8/6
Age [years]	59.5 (44–75)	63 (45–76)	61.5 (46–77)	60 (51–76)
Angiopathy	13/16&	10/10&	15/16&	N/A
Severity of angiopathy (1/2)	11/2	6/4	6/9	N/A
Progress of angiopathy	N/A	4/10&	8/16&	N/A
Presence of microalbuminuria	12/16	5/9&	4/12&	N/A
Microalbuminuria ¥	10 (0–80)	8.2 (0.8–28.1)	4.2 (1.5–46.9)**	3.0 (1.1–5.4)*
BMI	27.06 (24.4–31.4)	27.3 (22.3–34.7)	26.7 (24.8–34.5)	25.4 (21.87–31.22)
Creatinine [mg/dl]	0.88 (0.66–1.23)	0.79 (0.6–0.92)	0.74 (0.53–1.13)**	0.64 (0.52–0.87)*
GFR [ml/min/1.73^2]	79.9 (60–90)	90 (77.3–90)	90 (69.2–90)**	90 (74.7–90)*
CRP [mg/l]	1.68 (0.51–6.2)	1.66 (0.35–17)	1.91 (0.3–11.97)	0.8 (0- 4.6)
Platelets [x10^9/l]	250 (160–376)	232.5 (150–326)	253 (161–478)	248 (158–317)
Total cholesterol [mg/dl]	222 (182–299)	218 (141–246)	218 (155–275)	219 (139–286)
Triglycerides [mg/dl]	86 (36–224)	110 (57–245)	120 (47–192)	69 (47–158)*
HDL [mg/dl]	57 (45–92)	69 (47–100)	55 (42–87)	74 (52–112)*
LDL [mg/dl]	139 (102–208)	122 (59–153)	131.5 (75–183)	126 (66–190)
SBP [mmHg]§	150. (122–173)	132 (111–164)€	131.5 (130–166)**	127.5 (107–139)*
DBP [mmHg] §	89.5 (72–111)	81 (73–107) €	85 (76–96)**	76 (66–88)*
HR [beats/min] §	62 (45–81)	65 (51–88)	67 (54–92)	65 (62–90)
SVR [dynes/ seconds/cm− 5]	1183.2 (853.2–1534.8.2.8)	1239.31 (1049.81–1533)	1161.41 (975.12–1744.96.12.96)	1225.5 (722–1490.49.49)
CAI	26 (15–37)	26 (21–36)	27 (20–33)	24.5 (7–44)
CI	3.33 (2.59–4.12)	3.42 (2.89–4.24)	3.25 (2.53–4.83)	3.59 (2.68–4.85)

&- refers to the number of patients who were examined for the parameter at the given control point; BMI= body mass index, GFR = glomerular filtration rate, CRP = C-reactive protein, HDL= high-density lipoprotein, LDL = low-density lipoprotein, SBP= systolic blood pressure, DBP = diastolic blood pressure, HR= heart rate, SVR= systemic vessel resistance, CAI= central augmentation index, CI= cardiac index, N/A- nonapplicable, ¥ microalbuminuria is expressed as albumin to creatinine ratio (ACR) in mg of albumin per g of creatinine, § refers to mean values of daytime readings during ambulatory 24-hour blood pressure monitoring Progress of angiopathy in hypertensive patients was evaluated in relation to the examination at HT1 control point. In terms of severity of angiopathy 1 and 2 refer to low and high grade of angiopathy, respectively.

Analysis of cytokines and growth factors in serum

Serum was collected from patients and healthy controls into tubes with clot accelerator and centrifuged at 2000 g for 10 min. Then serum was aliquoted into three sterile tubes and stored at −80 °C until analysis, but not longer than 12 months. Concentrations of the following mediators: soluble CD40 ligand (sCD40L), epidermal growth factor (EGF), eotaxin/CCL11, fibroblast growth factor (FGF-2), FMS-like tyrosine kinase 3 ligand (Flt-3 L), fractalkine, granulocyte colony- stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), growth-regulating oncogene α (GRO- α/CXCL1), interferon alpha-2 (IFN-α2), IFN-γ, interleukin 1 α (IL −1α), IL-1β, IL-1 receptor antagonist (IL-1RA), IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8/CXCL8, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-15, IL-17 A, interferon gamma-induced protein 10 (IP-10/CXCL10), monocyte chemoattractant protein 1 (MCP-1/CCL2), monocyte-chemotactic protein 3 (MCP-3/CCL7), macrophage derived chemokine (MDC/CCL22), macrophage inflammatory protein-1 alpha (MIP-1α/CCL3), MIP-1β/CCL4, transforming growth factor α (TGF-α), tumour necrosis factor α (TNF-α), and TNF-β were measured with Milliplex MAP Human Cytokine/Chemokine Magnetic Bead Panel (HCYTMAG-60 K-PX38, Millipore, St. Charles, USA) and analyzed with Luminex (MAGPIX, Merck Millipore) according to the manufacturer instructions. In addition, levels of vascular endothelial growth factor A (VEGF-A), VEGF-C, VEGF-D, platelet derived growth factor PDGF-AB/BB, PDGF-AA, angiopoietin-2, bone morphogenic protein-9 (BMP-9), soluble endoglin (sCD105/sEng), endothelin, follistatin, heparin-binding EGF-like growth factor (HB-EGF), regulated upon activation normal T-cell expressed and secreted (RANTES/CCL5), hepatocyte growth factor (HGF), leptin, placental growth factor (PLGF) and monokine induced by IFNγ (MIG/CXCL9) were measured with ELISA.

In vivo evaluation of angio- and lymphangiogenic potential of the patients’ sera

Analysis of angio- and lymphangiogenic potential of sera collected from the studied groups was assessed in vivo in serum-induced cutaneous angiogenesis assays (SIA) as it was described before³³with own modifications. Briefly, 100 serum samples were assessed in 300 BALB/c healthy, wild type male mice (provided by Charles River Laboratories) at age of 6–8-weeks. The animals were randomised for intradermal (i.d.) injection with serum samples collected from healthy subjects and patients with hypertension just after the diagnosis (HT1) and serum samples derived from the same hypertensive individuals 1 (HT2) and 2 years later (HT3). Serum of each person was tested in 3 different and randomly chosen animals to exclude individual differences between the mice. The animals received 3 injections of tested serum in left flank (50 µl/injection) and 3 injections of 0.9% NaCl solution (50 µl/injection) to the right flank (to exclude effect of injection itself on generation of new vessels). After 5 days mice were euthanized in CO2 chamber with mixture of CO2 and O2 to fulfil the objective of rapid unconsciousness and death with minimal distress to the animals. The differences in blood and lymphatic vessel formation were assessed post mortem. For this purpose skin sections were collected with 3 biopsy punches of 6 mm diameter each. Then, the samples were fixed with formaldehyde and paraffin embedded, cut and stained with haematoxylin and eosin (H&E) according to standard procedure. Afterwards, blood and lymphatic vessels were identified with anti-CD31 (DIA-310, Dianova) and anti-podoplanin (AB109059, Abcam) antibodies, respectively. Each time the number of vessels formed in response to the control injection of NaCl was subtracted from the number of vessels formed in mice after serum injection. Animals’ allocation to the experimental groups was blinded. The team who analysed histological samples known the codes of the animals, but had no data on the health status of the donors of the tested serum samples. Human serum injections were well tolerated by the animals and no visible inflammatory response after the procedure was observed. Animal studies were approved by the Local Ethical Committee for Animal Studies in Gdańsk (agreement no. 15/2010). All procedures were performed in accordance with the relevant guidelines and regulations. The study is reported in accordance with ARRIVE standards (https://arriveguidelines.org) to maximize the quality and reliability of the research.

Flow cytometry

Peripheral blood mononuclear cells (PBMCs) were obtained by Ficoll gradient centrifugation from blood collected into sterile heparinized tubes. The isolated cells were stained with 2 panels of the antibodies (Ab) obtained from the following manufacturers: 1). BioLegend: anti-CD4 Pacific Blue (344620), anti-VEGFR-2 Alexa Fluor 488 (359914), anti- CD31 APC-Cy-7 (303120), anti- Nrp-1 PE-Cy7 (354508), anti- CXCR4 PerCP-Cy5.5 (306516), anti- CD127 PerCP-Cy5.5 (351322); 2) eBioscience: anti- FoxP3 APC (17–4777-42), 3) Abcam: primary anti-Galectin-1 (ab25138) with secondary Ab conjugated with Alexa Fluor 488 (ab181448); 3) BD: anti-CD25 PE (555432), anti-CD45RA PE-Cy7 (337186); 4) Invitrogen: anti-CD62L APC-AF750 (MHCD62L27), and analyzed immediately with flow cytometer (LSRFortessa, BD Biosciences). Conventional CD4 + T cells and thymus derived natural Tregs were identified as CD4 + CD25-/LowFoxP3- and CD4 + CD25HighFoxP3+ cells, respectively³⁴. Tregs were gated into naive (Tn), central memory (Tcm), effector memory (Tem) and effector memory cells that re-express CD45RA (Temra) according to the following phenotypes: CD45RA+CD62L+, CD45RA-CD62L+, CD45RA-CD62L-, and CD45RA+CD62L-, respectively³⁴. Cells double positive for CD31 and CXCR4 were considered as proangiogenic subsets according to the previous reports⁸.

Statistical analysis

Results were analysed with GraphPad Prism (v10.1.1). The data were not normally distributed as confirmed with Shapiro-Wilk’s test. Therefore, to compare the differences between hypertensive patients at different time-points (HT1, HT2, HT3) we used Kruskal-Wallis (KW) test with post hoc analysis (Dunn’s test). In cases were only two groups were compared (e.g. HT1 vs. controls; HT1 vs. HT3) Mann-Whitney’s U test (MW) was used. Correlations between studied parameters were evaluated with Spearman’s rank correlation (SC). Values of p < 0.05 were considered significant. Heat maps were created with ClustVis tool available online at http://biit.cs.ut.ee/clustvis/#mathematics.

Results

Two years after the treatment implementation a significant increase in lymphangiogenic potential of serum and improved renal function are observed in hypertensive patients treated with ACEI

As microvascular rarefaction was reported to be a hallmark of arterial hypertension²we hypothesized that serum of hypertensive patients has lower proangiogenic activity. To verify this hypothesis we have performed serum-induced angiogenesis (SIA) assays using in vivo model, as described in the Materials and methods section. Intra dermal injection of human sera induced generation of new vessels in the animal skin. The effect was more pronounced for serum than for the control injections with NaCl and generation of new blood vessels was macroscopically visible (Fig. 1A). Then formaldehyde fixed and paraffine embedded samples where cut and stained with HE to visualize cellular and tissue morphology (Fig. 1B). After staining with anti-CD31 and anti-podoplanin antibodies (Abs) the numbers of blood and lymphatic vessels, respectively were counted. No statistically significant differences were found for numbers of blood vessels generated in in vivo model by sera collected from healthy and hypertensive patients at all control points (Fig. 1C). Interestingly, our in vivo studies showed that lymphangiogenic potential of sera collected from early diagnosed hypertensive patients is significantly decreased as compared with healthy individuals (Fig. 1D, MW, p = 0.01). However, 2-year treatment of hypertension substantially increased serum ability to generate new lymphatic vessels (HT1 vs. HT3; Fig. 1D, KW test with post hoc analysis, p = 0.02). Representative micrographs of animal skin samples stained with ICH for CD31 and podoplanin that visualize blood and lymphatic vessels, respectively are also presented in the Fig. 1E. Patients’ stratification according to the treatment applied revealed that this effect was only induced in the patients who were on mono or combination therapy with ACEI. Sera of ACEI treated patients (ACEI group) at HT3 control point generated significantly higher numbers of lymphatic vessels in animal model then sera collected from individuals being on different therapy (Other group; ACEI vs. Other; Fig. 1D, M-W, p = 0.008) and sera of healthy subjects (ACEI vs. Control; Fig. 1D, M-W, p = 0.04). Two-year treatment with hypotensive drugs was associated with improved renal function that was reflected by decreased microalbuminuria (Fig. 1F, M-W, p = 0.04), increased GFR (Fig. 1G; M-W; p = 0.03), and decreased serum creatinine values (Fig. 1H; M-W; p = 0.02). Noteworthy, the most significant improvement in kidney function was observed in ACEI treated group where no microalbuminuria was detected 2 years after the therapy implementation (Fig. 1F; ACEI vs. Other, M-W, p = 0.013). Most of the patients experienced improvement in arterial blood pressure control over time. In the studied cohort 8/10 (80%) and 14/16 (87.5%) patients had well controlled blood pressure (SBP < 140 mmHg and DBP < 90 mmHg) at HT2 and HT3 control point, respectively. Median SBP and DBP values at HT1 control point (before the treatment assignment) were 150 and 89.5 mmHg, while 1 (HT2) and 2 years (HT3) after therapy implementation they dropped to 132 and 81 mmHg, and 131.5 and 85 mmHg, respectively (Table 1). Interestingly, no statistically significant differences for SBP and DBP were found between patients treated for 2 years with ACEI and other types of hypotensive drugs. In ACEI treated group median decrease in SBP and DBP between HT1 and HT3 control points was − 18.5 and − 7.5 mmHg, respectively, while in the patients treated with different types of hypotensive drugs the drop in SBP and DBP was − 15 and − 4.5 mmHg, respectively.

Fig. 1.

Increased capacity for lymphatic vessel formation and improved renal function in hypertensive patients 2 years after ACEI implementation. (A) Representative macroscopic view of animal skin after NaCl (control; left panel) and hypertensive patient serum (right panel) injection. (B) Representative light micrograph of H&E stained mouse skin samples at the place of NaCl (control; left panel) and hypertensive patient serum (right panel) injection. The scale bar reflects 500 μm. (C) The graphs depict numbers of blood vessels generated after injection of serum samples derived from healthy individuals (Control, n = 9), early diagnosed hypertensive patients before the treatment implementation (HT1, n = 8), and then 1 (HT2, n = 9) and 2 (HT3, n = 12) years later. Differences between the groups were calculated with K-W with post hoc tests. (D) The left graph depicts numbers of lymphatic vessels generated after injection of serum samples derived from healthy individuals (Control, n = 11), early diagnosed hypertensive patients before the treatment implementation (HT1, n = 13), and then 1 (HT2, n = 9) and 2 (HT3, n = 12) years later. Differences between the groups were calculated with K-W with post hoc tests. The right graph presents numbers of lymphatic vessels generated after injection of serum samples derived from healthy individuals (Control, n = 11) and hypertensive patients at HT3 control point treated with ACEI (ACEI, n = 7) and other hypotensive drugs (Other, n = 5). Differences between Control vs. ACEI and Control vs. Other were calculated with MW. (E) Representative micrographs of animal skin samples stained with ICH for CD31 (blood vessels) and podoplanin (lymphatic vessels) after i.d. injection of NaCl and serum collected from healthy and hypertensive donors during the indicated visits. (F) Upper panel depicts microalbuminuria in hypertensive patients at HT1 vs. HT3 control points. Differences between the groups were calculated with M-W. Lower panel depicts microalbuminuria in hypertensive patients treated with ACEI and other drugs at HT3 control point. Differences between the groups were calculated with M-W. (G) GFR in hypertensive patients at HT1 vs. HT3 control points. Differences between the groups were calculated with M-W. (H) Serum creatinine in hypertensive patients at HT1 vs. HT3 control points. Differences between the groups were calculated with M-W. For all graphs p < 0.05 was considered statistically significant and p values are given in the figure. In the boxplots median is indicated by the symbol within the box, lower and upper bounds of the box correspond with the 25th and 75th percentiles. The lower and upper whiskers indicate minimum and maximum values, respectively. Microalbuminuria is expressed as albumin/creatinine ratio (ACR) in mg of albumin per g of creatinine.

Improved lymphangiogenic potential of serum of patients treated with ACEI is associated with increased levels of VEGF-C and MDC and decreased concentrations of MIP-1α and MIP-1β

In aim to understand mechanism of altered lymphangiogenesis in hypertension and identify the crucial regulators of this process we measured 52 cytokines and growth factors in the same serum samples that were used in in vivo assay. The mediators were measured with Luminex and ELISA methods. Among 52 analyzed factors, 36 were detected in the studied material. The following 16 cytokines: Flt-3 L, GM-CSF, IL-1α, IL-1β, IL-1RA, IL-2, IL-5, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-15, MCP-3, PLGF and TNF-β were below the detection limit in majority of the samples (data not shown). As compared with healthy individuals, recently diagnosed hypertensive patients had significantly higher serum levels of EGF (p = 0.008), VEGF-D (p = 0.03), HB-EGF (p = 0.03), TNF-α (p = 0.04), MCP-1 (p = 0.02) and MIG (p = 0.01; Fig. 2A; Table 2). Simultaneously, the following mediators were decreased in sera of HT1 group as compared with the controls: VEGF-C (K-W with post hoc analysis; p = 0.03), IP-10 (K-W with post hoc analysis; p = 0.02), MDC (K-W with post hoc analysis; p = 0.01) and BMP-9 (K-W; p = 0.02; Fig. 2A; Table 2).

Fig. 2.

Alterations in serum cytokine profile during 2 years after hypotensive treatment implementation and its impact on lymphangiogenesis. (A) The heat map shows median serum cytokine levels in healthy (Control, n = 14) and hypertensive patients before the treatment implementation (HT1, n = 16), and then 1 (HT2, n = 10) and 2 (HT3, n = 16) years later. (B) Left panel depicts negative correlations between serum levels of MIP-1α and MIP-1β in hypertensive patients 2 years after treatment implementation (HT3 control point) and numbers of lymphatic vessels induced by these serum samples in the in vivo model (SC; R and p values are given). Middle panel shows differences in MIP-1α and MIP-1β levels in sera of the same patients at HT1, HT2 and HT3 control points (KW with post hoc tests, p values are given). Right panel presents concentration of MIP-1α and MIP-1β in sera of hypertensive patients treated with ACEI (ACEI) and other drugs (Other) at HT3 control point (M-W, p values are given). (C) Left panel depicts positive correlations between serum levels of VEGF-C and MDC in hypertensive patients 2 years after treatment implementation (HT3 control point) and numbers of lymphatic vessels induced by these serum samples in the in vivo model (SC; R and p values are given). Middle panel shows differences in VEGF-C and MDC levels in sera of the same patients at HT1, HT2 and HT3 control points (KW with post hoc tests, p values are given). Right panel depicts concentration of VEGF -C and MDC in sera of hypertensive patients treated with ACEI (ACEI) and other drugs (Other) at HT3 control point (M-W, p values are given). For all analyses p < 0.05 was considered statistically significant. Noteworthy, not all sera from HT3 control point were tested in animal model, therefore differences in sample numbers are visible between correlations and box plots, where all collected samples were used for cytokine level analysis. In the boxplots median is indicated by the symbol within the box, lower and upper bounds of the box correspond with the 25th and 75th percentiles. The lower and upper whiskers indicate minimum and maximum values, respectively.

Table 2.

Serum cytokine pattern in healthy individuals and hypertensive patients during 2- year follow-up. The table provides median cytokine concentrations, with min and max values in brackets for healthy individuals (Control, n = 14) at single time point, and for hypertensive patients at three time points: after diagnosis of hypertension (HT1, n = 16), one (HT2, n = 10) and two years later (HT3, n = 16). Cytokine concentrations are given in [pg/ml]. Differences between control and hypertensive patients at the day of diagnosis (Control vs. HT1) and between HT1 and HT3 control points were calculated with Mann-Whitney U test. P values are given for these comparisons. Statistically significant differences (p < 0.05) for comparisons between control vs. HT1 and HT vs. HT3 are marked with * and **, respectively.

Cytokine	Control median (min-max)	HT1 median (min-max)	HT2 median (min-max)	HT3 median (min-max)	p Control vs. HT1	p HT1 vs. HT3
Endoglin	820.61 (0–1664.00.00)	716.50 (523.03–906.85.03.85)	849.48 (486.02–1018.00)	899.17 (543.18–122)**	0.22	0.003
MDC	929.37 (377.63–1949.00)	537.21 (384.170–933.57.170.57)*	732.29 (267.47–1171.00)	777.46 (516.65–1050) **	0.01	6 × 10− 4
VEGF-A	227.42 (0–938.88.88)	185.89 (17.930–419.9)	353.77 (103.44–772.98.44.98)	411.70 (27.34–10.91) **	0.26	1 × 10− 5
BMP-9	63.83 (6.82–110.12.82.12)	41.47 (18.760–81.02)*	44.47 (4.50–62.06.50.06)	42.67 (6.29–82.33)	0.02	0.94
Eotaxin	177.12 (29.65–331.54.65.54)	140.05 (56.940–240.33.940.33)	143.05 (52.83–291.14.83.14)	129.35 (0–293.89.89)	0.29	0.54
EGF	74.55 (0–181.26.26)	111.27 (59.900–170.18.900.18)*	86.96 (36.22–145.16.22.16)	75.70 (53.82–155.16.82.16)**	0.008	0.006
IL-6	2.68 (0–19.94.94)	3.24 (0.330–9.14)	2.18 (0–8.39.39)	0.31 (0–3.23.23)**	0.29	3 × 10− 4
Angiopoietin-2	935.88 (0–1266.00.00)	902.51 (372.6–1430)	1165.00 (394.08–1908.00)	1028.0 (419.73–2346)	0.62	0.23
PDGF-AB/BB	425.36 (175.34–1535.00)	393.92 (0–920.43.43)	578.85 (109.02–866.34.02.34)	498.42 (192.34–1130)	0.68	0.20
HGF	117.74 (0–331.09.09)	155.68 (100.08–304.05.08.05)	152.10 (41.36–279.82.36.82)	176.83 (86.48–386.6)	0.22	0.21
Leptin	6009.50 (0–22186.00.00)	5652 (1169–15288)	4397.00 (2106.00–5935.00.00.00)	5672 (1963–26951)	0.98	0.86
IFN-γ	4.62 (0–75.32.32)	3.21 (0–28.6.6)	1.95 (1.49–35.80)	2.69 (0 −27.72)	0.51	0.97
GRO	1010.43 (498.82–1888.00)	1125 (438.72–6112)	1085.00 (386.62–1902.00)	843.85 (0–5256)	0.57	0.30
MIP-1	22.84 (0–54.05.05)	27.77 (0–112.2.2)	6.57 (0–46.56.56)	4.92 (0–55.52.52)**	0.62	0.001
MIP-1β	33.04 (7.07–78.71)	36.69 (5.72–115.34.72.34)	20.70 (5.72–101.04.72.04)	18.75 (0–87.88.88)**	0.98	0.04
TNF-α	12.16 (1.38–24.84)	19.74 (7.84–50.76)*	8.35 (2.41–17.18)	8.59 (0–48.08.08)**	0.049	2 × 10− 4
VEGF-C	594.79 (310.42–996.03.42.03)	453.87 (46.41–712.93.41.93)*	454.92 (18.43–921.52.43.52)	775.92 (449.37–1045)**	0.03	6 × 10− 4
VEGF-D	72.35 (0–131.35.35)	123.93 (3.54–364.46.54.46)*	62.14 (0–215.68.68)	108.75 (0–220.37.37	0.03	0.46
IL-7	0.00 (0–7.61.61)	2.46 (0–31.8.8)	1.23 (0–29.99.99)	3.87 (0–22.05.05)	0.51	0.83
RANTES	775.29 (254.66–3140.00)	1035 (0–2344)	891.72 (357.80–1525.00)	667.95 (123.31–2493)**	0.39	0.04
Follistatin	143.67 (0–260.36.36)	167.27 (74.14–284.79.14.79)	176.07 (48.13–260.36.13.36)	152.84 (88.07–261.91.07.91)	0.26	0.51
MIG	139.15 (72.91–328.93.91.93)	243.16 (99.33–820.91.33.91)*	154.58 (58.77–492.56.77.56)	65.87 (43.42–137.67.42.67)**	0.01	3 × 10− 6
IP-10	305.19 (199.71–1146.00)	244.87 (143.71–334.7)*	361.77 (203.15–559.47.15.47)	303.01 (0–578.43.43)**	0.02	0.01
TGF-α	3.41 (0–34.98.98)	2.82 (0–17.84.84)	1.32 (0–8.21.21)	2.86 (0–11.6.6)	0.59	0.65
IFN-α2	9.86 (0–120.74.74)	0 (0–83.72.72)	5.08 (0–77.00)	0 (0–117.2.2)	0.33	1.00
G-CSF	26.28 (0–1036.00.00)	10.54 (0–201.08.08)	10.54 (3.99–31.48)	14 (0–341.42.42)	0.29	0.97
Fractalkine	22.08 (0–173.35.35)	0 (0–141.39.39)	0.00 (0–37.97.97)	4.9 (0–118.41.41)	0.28	0.51
sCD40L	17377.00 (8037.00–17377.00.00.00)	16,094 (0–17377)	16798.50 (3624.00–17377.00.00.00)	17,377 (0–17377)	0.31	0.45
IL-4	2.01 (0–46.57.57)	0 (0–71.62.62)	4.73 (0–38.47.47)	0 (0–34.5.5)	0.49	0.65
FGF-2	45.62 (0–125.48.48)	22.25 (0–120.03.03)	20.26 (0–125.48.48)	25.53 (0–172.84.84)	0.25	0.92
PDGF-AA	118.29 (60.67–275.41.67.41)	135.49 (0–177.82.82)	126.62 (68.02–167.93.02.93)	110.92 (46.75–183.25.75.25)	0.57	0.35
PLGF	1.07 (0–4.91.91)	1.85 (0–26.49.49)	3.05 (0–22.26.26)	1.85 (0–39.23.23)	0.18	0.73
IL-17α	5.74 (0–29.94.94)	3.19 (0–59.4.4)	0.00 (0–80.60.60)	0 (0–51.11.11)	0.49	0.41
IL-8	52.60 (0–682.08.08)	86.6 (0.27–551.2)	18.81 (0–69.70.70)	12.05 (4.25–48.28)**	0.54	0.04
MCP-1	425.08 (203.28–870.49.28.49)	557.93 (341.85–1270)*	465.39 (211.12–983.69.12.69)	367.02 (0–675.17.17)**	0.02	6 × 10− 4
HB-EGF	42.19 (0–86.00)	53.54 (31.740–141.19.740.19)*	67.90 (33.11–104.11.11.11)	49.56 (13.46–94.61)	0.03	0.08

In addition, we observed substantial changes in the serum cytokine profile in hypertensive patients during the follow-up. The most pronounced differences were found between the day of diagnosis (HT1) and the last control point (HT3; Fig. 2A; Table 2). The differences referred to decreased concentrations of proinflammatory mediators (Table 2): IL-6 (M-W; p = 0.0003), TNF-α (M-W; p = 0.0002), RANTES (M-W; p = 0.04), MIP-1α (M-W; p = 0.001), MIP-1β (M-W; p = 0.04), MIG (M-W; p = 0.00002), IL-8 (M-W; p = 0.04) and MCP-1 (M-W; p = 0.0006). In addition, a significant decrease in EGF (M-W; p = 0.006) was observed and tendency for decrease in HB-EGF (M-W; p = 0.08; Fig. 2A; Table 2). Simultaneously, 2 years after diagnosis of hypertension we found a significant increase in serum levels of endoglin (HT1 vs. HT3; M-W; p = 0.003, Table 2), MDC (M-W; p = 0.0009; Table 2), VEGF-A (M-W, p = 0.00001; Table 2), VEGF-C (M-W, p = 0.0006; Table 2) and IP-10 (M-W, p = 0.01; Table 2).

Interestingly, among these cytokines only 4 correlated with lymphangiogenesis. The strongest prolymphangiogenic potential showed sera with the lowest concentrations of MIP-1α (SC; R= −0.61, p = 0.03; Fig. 2B) and MIP-1β (SC; R= −0.76, p = 0.005; Fig. 2B) and the highest levels of VEGF-C (SC; R = 0.57, p = 0.04; Fig. 2B) and MDC (SC; R = 0.72, p = 0.01, Fig. 2C). As comparted with the day of diagnosis (HT1 control point), two-year therapy with hypotensive drugs (HT3 control point) resulted in decline in serum concentrations of MIP-1α and MIP-1β (HT3 vs. HT1, K-W, p = 0.002 and p = 0.01, respectively; Fig. 2B) and increased levels of VEGF-C and MDC (HT3 vs. HT1, K-W with post hoc analysis, p = 0.008 and p = 0.03; respectively; Fig. 2C). Noteworthy, when the patients were stratified again at the HT3 control point into ACEI and Other subgroups according to the therapy applied, we observed that the most significant changes took place in the individuals treated with ACEI (Fig. 2B-C). Serum concentrations of MIP-1α, which correlated negatively with lymphangiogesis were significantly lower in the patients treated with ACEI at HT3 control point, than in the individuals on different therapy (M-W, p = 0.007; Fig. 2B). Simultaneously, levels of VEGF-C that is a known stimulator of lymphangiogenesis^5–7and was associated with increased generation of lymphatic vessels in our study were ≈ 40% higher in the ACEI group (median VEGF-C concentration 837 pg/ml), than in the other patients (median VEGF-C concentration 529 pg/ml; M-W, p = 0.0006; Fig. 2C) at HT3 control point. Noteworthy, serum concentrations of VEGF-C at time of diagnosis (HT1) were lower in the cohort assign to ACEI treatment, than in the individuals who received other types of hypotensive medications (median VEGF-C levels 447.99 pg/ml vs. 570.87 pg/ml, respectively). Therefore, the ACEI treated patients experienced significant increase in serum VEGF-C levels over time (HT1 vs. HT3, MW p=3 × 10^− 4), while the difference did not reach statistical significance in case of the individuals on different therapy (HT1 vs. HT3, MW p = 0.48). These data suggest VEGF-C and MIP-1α were the key factors responsible for ACEI induced lymphangiogenesis in our patients.

Higher numbers of CD31+CXCR4+ Th cells are associated with improved renal function and are preserved by ACEI treatment

Frequency of CD31+CXCR4+ Th cells was analysed with flow cytometry. The gating strategy is presented in the Fig. 3A. Two years after hypotensive drug implementation (HT3), a significant decrease in frequency of CD31+CXC4+ cells among CD4+ T cell population was observed, as compared with the time of diagnosis (HT3 vs. HT1; K-W with post hoc analysis, p = 0.04) and HT2 control time point (HT3 vs. HT2; K-W with post hoc analysis, p = 0.03; Fig. 3B). Interestingly, stratification of the patients according to the treatment applied revealed that the loss of this proangiogenic population of Th cells was not observed in the patients treated with ACEI. The ACEI subgroup was characterized by the same numbers of CD31+CXCR4+ Th cells as the control cohort. Simultaneously, patients treated with different drugs than ACEI showed significantly lower numbers of CD31+CXCR4+ Th cells, than healthy individuals (M-W, p = 0.002; Fig. 3B) and ACEI subgroup (M-W, p = 0.01; Fig. 3B).

Fig. 3.

Numbers of CD31+CXCR4+ Th cells in peripheral blood during 2 years of hypertension duration and their association with renal function. (A) The upper panel depicts gating strategy for analysis of CD31+CXCR4+ Th cells that included cell conglomerate exclusion via FSC-H vs. FSC-A and SSC-H vs. SSC-A gating. (B) Left panel presents frequency of CD31+CXCR4+ cells within CD4+ T cell population in healthy individuals (Control, n = 13) and early diagnosed hypertensive patients before the treatment implementation (HT1, n = 16), and then 1 (HT2, n = 8) and 2 (HT3, n = 12) years later (KW with post hoc tests, p values are given). (C) Frequency of CD31+CXCR4+ Th cells in peripheral blood was in negative correlation with serum endoglin levels (SC; R and p values are given). (D) Serum concentration of endoglin was found to increase 2 years after hypertension diagnosis (upper panel, HT1 vs. HT3; MW; p value is given). However, 2-year treatment with ACEI was associated with lower increase of serum endoglin concentration as compared with therapy with other drugs (lower panel, M-W, p values are given, results for the control group are shown as reference). (E) Numbers of CD31+CXCR4+ Th cells were in positive correlation with GFR, and (F) in negative correlation with microalbuminuria and (G) serum creatinine levels (SC; R and p values are given). C, E-G present data for HT3 control point (2 years after diagnosis of hypertension; n = 11). In the boxplots median is indicated by the symbol within the box, lower and upper bounds of the box correspond with the 25th and 75th percentiles. The lower and upper whiskers indicate minimum and maximum values, respectively. For all analyses p < 0.05 was considered statistically significant and p values are given in the figure. Microalbuminuria is expressed as albumin/creatinine ratio (ACR) in mg of albumin per g of creatinine.

Frequency of CD31+CXCR4+ cells among Th population was in negative correlation with endoglin serum levels (SC; R=−0.72, p = 0.01; Fig. 3C). Noteworthy, endoglin concentration was found to increase with time of hypertension duration (HT3 vs. HT1; M-W; p = 0.005; Fig. 3D, upper panel). However, stratification of the patients at the HT3 control point, revealed that serum endoglin levels did not increase in the patients treated with ACEI but only in the individuals who were on different therapy (ACEI vs. Other; M-W; p = 0.03; Fig. 3D, lower panel).

Higher numbers of Th CD31+CXCR4+ cells were associated with increased GFR (SC; R = 0.66, p = 0.03; Fig. 3E), decreased microalbuminuria (SC; R= −0.71, p = 0.01; Fig. 3F) and decreased serum creatinine levels (SC; R= −0.60, p = 0.04; Fig. 3G), which are hallmarks of improved renal function.

Higher numbers of CD31+CXCR4+ Tregs are associated with lower stiffness of vessel wall and their numbers can be preserved over time by ACEI therapy

CD31+CXCR4+ Treg numbers were analysed with flow cytometry. The gating strategy used for their identification is presented in the Fig. 4A. Two years after hypertension diagnosis a significant decrease in frequency of CD31+CXCR4+ Tregs was observed (HT3 vs. HT 1; K-W with post hoc analysis, p = 0.01, Fig. 4B, left panel). However, the decrease was significantly lower in the individuals treated with ACEI, than in the patients who were on different therapy (ACEI vs. Other; M-W, p = 0.008, Fig. 4B, right panel). Interestingly, we found that lower numbers of CD31+CXCR4+ Tregs were associated with higher values of systemic vessel resistance (SVR; M-W; p = 0.007; Fig. 4C, left panel), which reflects negative changes (increased stiffness) in the arterioles. Stratification of the patients according to the therapy applied, revealed that individuals treated with ACEI for 2 years after diagnosis had lower values of SVR than those treated with different drugs (M-W; p = 0.01; Fig. 4C, right panel).

Fig. 4.

Alterations in numbers of CD31+CXCR4+ Tregs during 2 years of hypertension duration and their association with systemic vessel resistance. (A) The upper panel depicts gating strategy for analysis of CD31+CXCR4+ Treg cells that included cell conglomerate exclusion via FSC-H vs. FSC-A and SSC-H vs. SSC-A gating. Tregs were identified as CD4+CD25HighFoxP3+ lymphocytes and then studied for CD31 and CXCR4 expression. (B) The left panel depicts frequency of CD31+CXCR4+ cells within Treg cell population in healthy individuals (Control, n = 9) and early diagnosed hypertensive patients before the treatment implementation (HT1, n = 16), and then 1 (HT2, n = 8) and 2 (HT3, n = 12) years later (KW with post hoc tests, p values are given). The right panel shows differences in CD31+CXCR4+ Treg numbers between hypertensive patients treated for 2 years (HT3 control point) with ACEI (ACEI, n = 6) or other types of hypotensive drugs (Other, n = 6; MW, p value is given). (C) Left panel depicts association between values of systemic vessel resistance (SVR, n = 10) and numbers of CD31+CXCR4+ Tregs (Angio Tregs_H and Angio_Tregs_L stands for higher and lower frequency of CD31+CXCR4+ Tregs, respectively; MW; p value is given). Right panel shows SVR values in hypertensive patients 2 years after treatment with ACEI (ACEI, n = 7) and other types of hypotensive drugs (Other, n = 5; MW; p value is given). In the boxplots median is indicated by the symbol within the box, lower and upper bounds of the box correspond with the 25th and 75th percentiles. The lower and upper whiskers indicate minimum and maximum values, respectively. For all analyses p < 0.05 was considered statistically significant and p values are given in the figure.

Faster progression of microangiopathies in hypertension is associated with lower numbers of Tregs and shift from Tcm to Tem phenotype in Treg population

As Treg cells (CD4+CD25HighFoxP3+) can ameliorate cardio-vascular diseases (CVD) by suppressing excessive inflammation, we have also monitored their numbers in hypertensive patients. Figure 5A depicts the gating strategy for identification of Tregs and their subsets according to antigen experience. Naive (Tn), central memory (Tcm), effector memory (Tem) and effector memory Tregs that re-expressed CD45RA (Temra) were identified according to the following phenotype: CD45RA+CD62L+, CD45RA-CD62L+, CD45RA-CD62L-, and CD45RA+CD62L-, respectively.

Fig. 5.

Association of central memory and effector memory Treg subsets with hypertension complications. (A) The upper panel depicts gating strategy for analysis of naive and memory populations of Tregs. Each time only single cells were analyzed due to FSC-H vs. FSC-A and SSC-H vs. SSC-A gating strategy. Tregs were identified as CD4+CD25HighFoxP3+ lymphocytes. The naive (Tn), central memory (Tcm), effector memory (Tem) and effector memory T cells that re-expressed CD45RA (Temra) were identified within Treg population according to the following phenotype: CD45RA+CD62L+, CD45RA-CD62L+, CD45RA-CD62L- and CD45RA+CD62L-, respectively. (B) Left panel depicts frequency of Tregs (CD25HighFoxP3+) within peripheral blood CD4+ T cells in healthy individuals (Control, n = 10) and early diagnosed hypertensive patients before the treatment implementation (HT1, n = 12), and then 1 (HT2, n = 8) and 2 (HT3, n = 11) years later (KW with post hoc tests). Right panel shows numbers of Tregs in hypertensive patients after 2- year therapy with ACEI (ACEI, n = 6) and other types of hypotensive drugs (Other, n = 5). (C) Two years after hypertension diagnosis patients with more advanced angiopathy (Aniopathy_2, n = 6) showed decreased numbers of peripheral blood Tregs (MW; p value is given; Angiopathy_1 = less advanced angiopathy, n = 5). (D) Treg numbers were in negative correlation with serum VEGF-A levels (SC, n = 11; R and p values are given). (E) Left panel depicts serum concentration of VEGF-A in hypertensive individuals at the day of diagnosis, before hypotensive drug administration (HT1, n = 16) and 2 years after treatment implementation (HT3, n = 16; MW). Right panel shows serum VEGF-A levels in hypertensive patients 2 years after treatment implementation stratified into ACEI (ACEI, n = 8) and other drug (Other, n = 8) treated groups (MW, p value is given). (F) Systemic vessel resistance (SVR, n = 12) was in positive and (G) negative correlation with frequency of effector memory (Tem) and central memory (Tcm) Treg subsets, respectively (SC; R and p values are given). (H) Higher numbers of Tem Tregs (n = 12) were associated with higher microalbuminuria (SC; R and p values are given) and (I) progression of angiopathy (Angio_progress) 2 years after hypertension diagnosis (MW; p value is given; Angio_stable= lack of angiopathy progression during 2 years after hypertension diagnosis). (J) Higher ratio of Tcm/Tem subsets within Treg population was observed in patients with non-progressing angiopathy (Angio_stable, n = 6) as compared with the patients whose angiopathy progressed during 2 years after hypertension diagnosis (Angio_progress, n = 6; MW; p value is given). In the boxplots median is indicated by the symbol within the box, lower and upper bounds of the box correspond with the 25th and 75th percentiles. The lower and upper whiskers indicate minimum and maximum values, respectively. For all analyses p < 0.05 was considered statistically significant and p values are given in the figure.

Two-year follow-up revealed tendency towards Treg loss with time of hypertension duration (Fig. 5B, left panel). However, the trend was less visible in the individuals treated with ACEI (Fig. 5B, right panel). When the patients at HT3 control point were divided into 2 subgroups according to stage of angiopathy, we found that more advanced stages of angiopathy were associated with lower Treg numbers (≤ 2.5% of Tregs within CD4+ T cell population; M-W; p = 0.004; Fig. 5C). Grade of hypertensive retinopathy and renal disfunction were taken into account while assigning the angiopathy stage. In addition, frequency of Tregs was in negative correlation with VEGF-A (SC; R= −0.61; p = 0.04; Fig. 5D) which was found to significantly increase with time of hypertension duration (HT1 vs. HT3; M-W; p = 0.00004; Fig. 5E, left panel) with the exception of the patients treated with ACEI (ACEI vs. Other; M-W; p = 0.00001; Fig. 5E, right panel). Interestingly, antigen experience of Tregs was associated with SVR. Frequency of effector memory (Tem) Tregs correlated positively with SVR (SC; R = 0.71; p = 0.01; Fig. 5F), while higher numbers of central memory (Tcm) Tregs were associated with lower values of this parameter (SC; R= −0.66; p = 0.02; Fig. 5G). Tem Treg numbers were also in positive correlation with microalbuminuria (SC; R = 0.68; p = 0.02; Fig. 5H). Interestingly, patients whose angiopathy progressed during 2-year follow-up had higher numbers of Tem Tregs (HT1 vs. HT3; M-W; p = 0.04; Fig. 5I) and lower Tcm to Tem ratio in Treg population (HT1 vs. HT3; M-W; p = 0.004; Fig. 5J).

Discussion

In the current study we present results of 2-year follow-up of early diagnosed hypertensive patients who were examined for angio- and lymphangiogenic potential of their sera and were tested for alterations in a pattern of 52 cytokines in the context of their clinical status. We show that serum of early diagnosed hypertensive patients before hypotensive drug administration is characterized by decreased lymphangiogenic potential. Nevertheless, 2- year treatment of the disease was found to induce substantial changes in serum cytokine pattern of the patients and improved lymphangiogenic potential of their sera with the most pronounced effect in the individuals treated with ACEI. Higher lymphangiogenic potential was associated with improved renal function expressed as lower microalbuminuria and serum creatinine with concomitant increase in GFR. In addition two years after diagnosis of hypertension a decrease in frequency of proangiogenic CD31+CXCR4+ Th and Treg cell subsets was observed with the exception of individuals treated with ACEI. Finally, we found that shift from Tcm to Tem subsets within Treg population reflects deleterious vascular alterations in hypertensive patients.

Reduced density of arterioles and capillaries is a hallmark of hypertension. However, it is unknown how this process is regulated and if it changes over time. Moreover, it is unclear if microvascular rarefaction is a cause or consequence of hypertension². We previously showed that arterial hypertension is characterized by cytokine profile that suppresses both angio- and lymphangiogenesis (increased concentrations of endostatin and decreased levels of bFGF and angiogenin), supporting the hypothesis that hypertension is associated with impaired generation of new vessels^4,35. With the current study we performed an in vivo serum-induced cutaneous angiogenesis assay (SIA) to verify if impaired angio- and/or lymphangiogenesis might be associated with development of hypertension. To understand if alterations in new vessel formation are cause or consequence of hypertension, we recruited hypertensive patients just after the diagnosis (control point HT1) and before any treatment implementation. The patients had a comprehensive examination of their clinical status (including Doppler renal assessment, OCT, impedance cardiography, central arterial pressure waveform analysis and analysis of basic biochemical parameter of blood and urine) in aim to correlate it’s alterations with cytokine profile, proportions of T cell subsets and potential for new vessel formation. Interestingly, serum of early diagnosed hypertensive patients showed a significant decline in generation of new lymphatic vessels, as compared to healthy individuals, while proangiogenic activity of the sera was not affected. Decreased lymphangiogenesis at the day of hypertension diagnosis and before any treatment administration, suggests that insufficient generation of new lymphatic vessels may contribute to pathogenesis of arterial hypertension. In addition, these data shade new light on lymphatic circulation also in the context of pathogenesis of renal complications of hypertension. Interestingly, our study revealed that the decreased lymphangiogenic potential of the serum in hypertensive patients is reversible and can be restored to normal or even higher levels after 2- year treatment with ACEI as mono or combination therapy. We found only one report regarding impact of ACEI on lymphangiogenesis³⁶. These observations were contradictory to ours. The authors showed that tumour bearing mice treated with ACEI had lower cancer expression of VEGF-C and diminished lymphangiogenesis, resulting in smaller tumour size. However the study was performed in mouse model and the design of the study differs significantly from ours. The group measured tumour lymphangiogenesis in 15 mice treated with ACEI for 21 days. While in our study the impact of ACEI on lymphangiogenesis was observed not earlier than 2 years after the treatment implementation. At the control time point HT2 (1 year after ACEI introduction) the differences in prolymphangiogenic potential of the patients’ sera were still not visible in our cohort. Therefore, the discrepancies between the studies result not only from the differences between the species, but mostly from the significant difference in duration of the follow-up.

According to our data ACEI induced lymphangiogenesis was associated with increased serum levels of VEGF-C and MDC (CCL22). These cytokines, seem to be the direct trigger for increased generation of new lymphatic vessels in the current study. Our data are in line with the observations of Kimura and colleagues who showed that macrophage derived VEGF‑C and MDC (CCL22) are responsible for induction of lymphangiogenesis in cancer³⁷. Nevertheless, among these two cytokines VEGF-C is known as the main chemoattractant for lymphatic endothelial cells and critical mediator for lymphangiogenesis and capillary stabilization^5,6,38–40. VEGF-C was also identified as an osmosensitive, hypertonicity-driven regulator of blood pressure. Machnik and colleagues showed that increased tissue tonicity activated a tonicity-responsive enhancer binding protein (TonEBP) that led to upregulation of VEGF-C in monocytes. Then, monocyte-derived VEGF-C triggered hyperplasia and increased density of the lymphcapillary network, lowering blood vessel wall tonicity and arterial blood pressure. Depletion of mononuclear phagocyte system (MPS-cells) or blocking VEGF-C signalling, augmented interstitial hypertonic volume retention, decreased endothelial nitric oxide synthase expression and prevented hyperplasia of the lymphatic capillary network, leading to elevated blood pressure^5,6. Interestingly, in our study the median levels of VEGF-C were strikingly higher in the patients treated for 2 years with ACEI (837 pg/ml), than in the individuals on different types of hypotensive drugs (529 pg/ml), despite initial (HT1) VEGF-C levels were lower in the patients designated for ACEI treatment. These observations indicate that therapy with ACEI leads to VEGF-C upregulation that in turn induces lymphangiogenesis. There is lack of reports on ACEI impact on serum levels of VEGF-C in human. There is only one study reporting correlation between ACEI treatment and different member of VEGF family- a VEGF-A. Zheng and colleagues reported that treatment with ACEI protects from microvascular diabetic complications via downregulation of VEGF-A levels in vitreous fluid and retina of diabetic rats. Interestingly, this effect was independent on anti-hypertensive action of the drug⁴¹. In our study, we observed increase in VEGF-A levels over time in hypertensive individuals. Nevertheless, when the patients were stratified according to the medications applied, it turned out that cohort treated with ACEI had significantly lower serum levels of VEGF-A, than individuals on different type of drugs. Therefore, our data also suggest that ACEI downregulates VEGF-A. In fact downregulation of VEGF-A and simultaneous upregulation of VEGF-C after 2-year treatment with ACEI in our patients is not surprising. VEGF-A is not an isoform of VEGF-C. Despite both proteins regulate neovascularization, they are encoded by distinct genes, they impact generation of different vessel type and thus exert different biological effects^42–46. Observations of Zheng and colleagues are in line with our study also in terms of beneficial impact of ACEI on microvasculature⁴¹. In consequence Zheng et al. reported retinoprotective effect of ACEI in diabetes, while we observed renoprotective action of these drugs in hypertension, expressed as improved GFR and lower proteinuria.

Our observations also shed new light on lymphangiogenesis in the context of kidney function in hypertension. It is well known that hypertension, notably when uncontrolled, leads to kidney damage². The current study indicates that mild alterations in renal function are usually present at the diagnosis of hypertension. Hypertensive patients at HT1 control point had already significantly lower GFR, as well as higher serum creatinine and urine albumin levels, than healthy individuals. Nevertheless, these changes were still reversible after implementation of hypotensive drugs. Two-year therapy improved blood pressure control and resulted in decreased serum creatinine levels and improved GFR in patients regardless the type of hypotensive drug implemented. However, only individuals treated with ACEI experienced simultaneously a significant decrease in microalbuminuria. These observations are of great importance, as suppression of glomerular albumin leak is a key factor in prevention of nephropathy⁴⁷.

Renoprotective effect of ACEI has been reported previously but was interpreted as a result of the fall in filtration pressure⁴⁸. However, if this hypothesis was correct, than a decrease in microalbuminuria would be observed in all patients with improved blood pressure control regardless the type of hypotensive drug implemented. While in our study the antiproteinuric effect and increased prolymphangiogenic potential of serum characterized only the cohort treated with ACEI. This way our observations link enhanced lymphangiogenesis with improved renal function, providing new insight into the mechanism of action of ACEI. At this point our data correspond with the study of Baranwal et al. who showed in mouse model that expanded renal lymphatics improve post-injury recovery of kidney⁴⁹. In addition, the authors reported that after initial inflammatory response, the acute injury phase was followed by kidney infiltration with CD4 + T cells. Thus, a significant increase in CD4/CD8 T cell ratio accompanied the lymphangiogenesis. Noteworthy, animals with higher lymphangiogenesis had reduced collagen deposition in kidneys and improved functional recovery over time. Unfortunately, this team did not characterize phenotype of CD4 + T cells⁴⁹. Therefore, we don’t know the proportions of CD31+CXCR4+, and CD25HighFoxP3 + CD31+CXCR4 + T cells within the CD4+ T cell (Th) population in their study. Nevertheless, in light of our data and the previous reports of the others, we may hypothesize that increased ratio of CD4/CD8 cells during enhanced lymphangiogenesis resulted from mobilization of CD4+CD31+CXCR4+ T cells and Tregs expressing CD31 and CXCR4.

CD4+CD31+CXCR4+ T cells (CD31+CXCR4+ Th cells)have been linked previously with angiogenesis by Hur and colleagues. They have shown that this T cell subset secretes high levels of proangiogenic cytokines including VEGF, IL-8, and matrix metalloproteinases and are indispensable for endothelial progenitor cell (EPC) colony formation and early differentiation. In addition, their frequency in peripheral blood of healthy individuals was found to positively correlate with colony-forming unit endothelial cells (CFU-EPCs) and thus with increased angiogenesis⁸. However in our study we did not find statistically significant correlation between numbers of CD31+CXCR4+ Th cells and proangiogenic potential of the patients’ sera. Nevertheless, higher frequency of CD31+CXCR4+ Th cells was associated with improved function of renal capillaries that was expressed by lower serum creatinine, lower microalbuminuria, and higher GFR values. Interestingly 2 years after diagnosis of hypertension (HT3 control point) only the patients treated with ACEI did not experience a loss of CD31+CXCR4+ Th cell number. Interestingly, loss of CD31+CXCR4+ Th cells was associated with increased serum levels of soluble endoglin (sEnd). Recently, it was reported that endoglin is expressed on various immune cells, including Th and Treg populations^50,51. It is upregulated during TCR (T cell receptor) stimulation. The cross-linking of surface bound endoglin was shown to enhance CD4 + T-cell proliferation⁵⁰. While increased serum levels of sEng in the patients not treated with ACEI indicate accelerated shedding of this protein form the cell membrane. Therefore, we may hypothesise that elevated serum concentrations of sEng in the patients not treated with ACEI resulted from increased shedding of this protein from various cell types, including CD31+CXCR4+ Th cells. In consequence, low abundance of membrane anchored Eng in this group resulted in lower proliferation of CD31+CXCR4+ Th cells and subsequent loss of their numbers at HT3 control point. This process can have clinical consequences, because elevated concentration of sEng in serum has been reported previously in preeclampsia, hypercholesterolemia, and atherosclerosis^52–54. In addition Emeksiz and colleagues reported that increased serum levels of sEng are detectable prior to development of subclinical structural vascular alterations in diabetic adolescents. Therefore, serum sEng has been considered as a marker of vascular complications in diabetes⁵⁵. In accordance with these reports increasing serum concentrations of sEng in hypertensive patients may also prognosticate development/progression of vascular complications., Interestingly, in our study the cohort treated with ACEI did not experience the loss of CD31+CXCR4+ Th cells and had significantly lower serum concentrations of sEng, than the patients treated with different types of drugs. Noteworthy, ACEI treated group showed also the highest lymphangiogenic potential of serum and the most pronounced improvement in kidney function after therapy implementation. Therefore, these data suggest that ACEI prevent/delay increase in serum sEng, which is associated with increased numbers of CD31+CXCR4+ Th cells and improve renal function by protective effect on microvasculature.

We have found similar pattern for CD31+CXCR4+ Tregs. A significant decrease in frequency of this Treg subset was observed over time in hypertensive patients. However, frequency of CD31+CXCR4+ Tregs was still significantly higher in ACEI treated individuals, than in the patients being on different types of hypotensive drugs. Noteworthy, higher numbers of CD31+CXCR4+ Tregs were associated with lower systemic vessel resistance (SVR) that reflects better condition of vasculature. SVR is also known as total peripheral resistance (TPR) and refers to the amount of force exerted on circulating blood by the vasculature of the body. It plays a vital role in establishing blood pressure and its higher values reflect deleterious alterations in vasculature, like increased stiffness of vessel wall due to histological changes or decreased ability for relaxation^56–58. Therefore, increased SVR is an important determinant of cardiovascular risk^56–58. Taking our observations all together we may summarize that 2-year therapy with ACEI improves condition of vasculature and protects from or at least slows down development of vascular complications in hypertension, while CD31+CXCR4+ T cells are involved in this process.

In the next step we analysed numbers of all Tregs in hypertensive patients. The loss of entire cell subset was less pronounced than in terms of CD31+CXCR4+ population. However, a trend towards lower Treg numbers in peripheral blood was observed over time. These data are in line with previous reports of the other groups on various cardiovascular diseases (CVDs). It was shown that patients suffering from CVD (including acute myocardial infarction, stable and unstable angina) have reduced numbers of CD4+CD25+FOXP3+ Tregs as compared with healthy individuals²⁶. While in individuals with unstable carotid artery plaque an imbalance between anti-inflammatory Tregs and proinflammatory Th-17 cells was reported²⁷. In our study Treg loss did not reach statistical significance. This may result from relatively good clinical status of our patients. They still did not suffer from severe and acute cardiovascular complications. Nevertheless, 2 years after hypertension diagnosis (HT3 control point) 15/16 patients had already diagnosed angiopathy. Interestingly, higher Treg numbers were associated with milder angiopathy and treatment with ACEI. As one could expect, vasculoprotective effect of Tregs does not result only from their quantity but also from their quality. We found that vascular pathologies were associated more with the shift in proportion of Tcm to Tem subset, than with the loss of entire Treg population. Higher numbers of Tem Tregs were associated with higher microalbuminuria and increased SVR values that reflects increased vessel rigidity and/or constriction⁵⁶. In addition, patients who experienced progression of angiopathy during the 2-year follow-up, had higher proportions of Tem Tregs to Tcm Tregs, than the patients whose angiopathy did not progress. Inversely, numbers of Tcm Tregs correlated negatively with SVR. High ratio of Tcm to Tem Tregs was a hallmark of stable angiopathy over the follow-up period. These observations are in line with the previous studies which reported that chronic inflammation that is the main risk factor for microvascular complications⁵⁸induces premature senescence of T cells, accelerating accumulation of Tem cells^60,61.

Noteworthy, previous studies on Tregs in hypertension mostly used angiotensin II induced animal model of this disease. In general, the link between hypertension, its complications and Tregs was reported in rodent studies^28–31, but little is known about Treg role in hypertension in human. For example it was shown that in mice hypertension is associated with decreased frequency of Tregs in peripheral blood even before onset of the disease^29–31. However, we did not find differences in Treg numbers between healthy individuals and early diagnosed hypertensive patients. Tregs were reported to prevent endothelial dysfunction in coronary arterioles in hypertensive animals. Interestingly, their adoptive transfer was proven to have therapeutic effect (e.g. prevention of vascular stiffness) due to their anti-inflammatory activities^29–31. These data correspond with our study. We found that lower SVR values that reflect lower vessel stiffness were observed in patients with higher proportions of Tcm and lower numbers Tem Treg subsets. Our data suggest that rather quality, than quantity of Tregs has impact on vessel condition in human. Recently Gackowska et al. reported reduced numbers of naive Tregs in hypertensive children³², what further supports our hypothesis. These findings underline the importance of fine interplay between immune and cardiovascular systems in regulation of vascular homeostasis.

In our patients Treg loss was associated with continuous increase in VEGF-A serum levels. Noteworthy, VEGF-A exerts proangiogenic activity, but simultaneously it is involved in inflammatory process and progression of hypertensive retinopathy^7,62–64. Factors that lead to endothelial damage (including ischemia, hypoxia and inflammation) induce VEGF-A which promotes neovascularization and potentiates inflammation⁶². Therefore, VEGF-A is recognized as an early marker of vascular endothelial damage, a major vascular permeability factor and marker of essential hypertension^42,63. Noteworthy, in our study ACEI therapy prevented rapid increase in VEGF-A serum concentrations. The cohort treated with ACEI showed about 2-fold lower levels of VEGF-A at HT3 control point, than the individuals being on different type of hypotensive drug.

In summary, in the current study we show for the first time that serum of hypertensive patients at the day of diagnosis is characterized by decreased lymphangiogenic potential. Nevertheless, 2-year treatment with ACEI leads to alterations in serum cytokine pattern characterized by significant increase in VEGF-C and decrease in MIP-1α levels. In consequence, long term treatment with ACEI improves lymphangiogenic potential of the patients’ sera. This in turn leads to improved renal function expressed as increase in GFR and decrease in microalbuminuria and serum creatinine levels. Simultaneously, therapy with ACEI prevents loss of CD31+CXCR4+ Th and CD31+CXCR4+ Treg cells that is observed in hypertensive patients being on different drug regimen. Higher numbers of these T cell subsets are associated with better renal function and lower stiffness of vessel wall. While lower numbers of all Tregs and shift from Tcm to Tem phenotype in Treg population are hallmark of fast progression of microangiopathies in hypertension.

The current study sheds new light on interactions between immune system and generation of new lymphatic vessels in hypertension. In addition, we show that ACEI have strong renoprotective effect and their implementation may prevent or significantly delay development of vascular complications in hypertension.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

NMT, KT, JS and PT conceived the study. NMT, KT, PT, JS, GPS, WB and RP designed the study. TK, KP and DIG collected flow cytometry data. MF, BT, AM, PG, MRG, MP, PG, LG recruited and examined the patients, collected samples and clinical data of the patients. MZ, TK and NMT analyzed flow cytometry data. AŻ and GPS performed animal studies. RP and JK performed IHC and counted numbers of blood and lymphatic vessels in animal samples. MG, ZUW and NMT performed statistical analysis and the figures. NMT, ZUW, MG, TK wrote the manuscript with contribution of all the authors. NMT revised the manuscript with the contribution of all the authors. All the authors read and approved the final version of the manuscript.

Funding

The study was funded by National Science Centre, Poland (funding decision no. DEC-2012/07/B/NZ5/00017, granted to J.S.) and by the European Funds for Smart Economy 2021–2027 (FENG) Priority FENG.02 Innovation-friendly environment, Measure FENG.02.01 International Research Agendas in the frame of the project “Science for Welfare, Innovations and Forceful Therapies (SWIFT)” no. FENG.02.01-IP.05–0031/23. (granted to N.M.T.).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.